fix poetry lock

- Added Langfuse configuration to settings and environment variables.
- Introduced Langfuse client for fetching prompts in chat service.
- Updated ChatConfig to include Langfuse prompt name.
- Enhanced service logic to retrieve prompts from Langfuse, improving prompt management capabilities.

This integration allows for rapid runtime prompt updates and eventual analytics of the performance of the Co-Pilot system.

This commit aims to trial this service as a potential option.

.branchlet.json

@@ -1,37 +0,0 @@
-{
-  "worktreeCopyPatterns": [
-    ".env*",
-    ".vscode/**",
-    ".auth/**",
-    ".claude/**",
-    "autogpt_platform/.env*",
-    "autogpt_platform/backend/.env*",
-    "autogpt_platform/frontend/.env*",
-    "autogpt_platform/frontend/.auth/**",
-    "autogpt_platform/db/docker/.env*"
-  ],
-  "worktreeCopyIgnores": [
-    "**/node_modules/**",
-    "**/dist/**",
-    "**/.git/**",
-    "**/Thumbs.db",
-    "**/.DS_Store",
-    "**/.next/**",
-    "**/__pycache__/**",
-    "**/.ruff_cache/**",
-    "**/.pytest_cache/**",
-    "**/*.pyc",
-    "**/playwright-report/**",
-    "**/logs/**",
-    "**/site/**"
-  ],
-  "worktreePathTemplate": "$BASE_PATH.worktree",
-  "postCreateCmd": [
-    "cd autogpt_platform/autogpt_libs && poetry install",
-    "cd autogpt_platform/backend && poetry install && poetry run prisma generate",
-    "cd autogpt_platform/frontend && pnpm install",
-    "cd docs && pip install -r requirements.txt"
-  ],
-  "terminalCommand": "code .",
-  "deleteBranchWithWorktree": false
-}

.dockerignore

@@ -1,9 +1,6 @@
 # Ignore everything by default, selectively add things to context
 *
 
-# Documentation (for embeddings/search)
-!docs/
-
 # Platform - Libs
 !autogpt_platform/autogpt_libs/autogpt_libs/
 !autogpt_platform/autogpt_libs/pyproject.toml
@@ -19,7 +16,6 @@
 !autogpt_platform/backend/poetry.lock
 !autogpt_platform/backend/README.md
 !autogpt_platform/backend/.env
-!autogpt_platform/backend/gen_prisma_types_stub.py
 
 # Platform - Market
 !autogpt_platform/market/market/

.github/workflows/claude-dependabot.yml

@@ -74,7 +74,7 @@ jobs:
 
       - name: Generate Prisma Client
         working-directory: autogpt_platform/backend
-        run: poetry run prisma generate && poetry run gen-prisma-stub
+        run: poetry run prisma generate
 
       # Frontend Node.js/pnpm setup (mirrors platform-frontend-ci.yml)
       - name: Set up Node.js

.github/workflows/claude.yml

@@ -90,7 +90,7 @@ jobs:
 
       - name: Generate Prisma Client
         working-directory: autogpt_platform/backend
-        run: poetry run prisma generate && poetry run gen-prisma-stub
+        run: poetry run prisma generate
 
       # Frontend Node.js/pnpm setup (mirrors platform-frontend-ci.yml)
       - name: Set up Node.js

.github/workflows/copilot-setup-steps.yml

@@ -72,7 +72,7 @@ jobs:
 
       - name: Generate Prisma Client
         working-directory: autogpt_platform/backend
-        run: poetry run prisma generate && poetry run gen-prisma-stub
+        run: poetry run prisma generate
 
       # Frontend Node.js/pnpm setup (mirrors platform-frontend-ci.yml)
       - name: Set up Node.js
@@ -108,16 +108,6 @@ jobs:
       #   run: pnpm playwright install --with-deps chromium
 
       # Docker setup for development environment
-      - name: Free up disk space
-        run: |
-          # Remove large unused tools to free disk space for Docker builds
-          sudo rm -rf /usr/share/dotnet
-          sudo rm -rf /usr/local/lib/android
-          sudo rm -rf /opt/ghc
-          sudo rm -rf /opt/hostedtoolcache/CodeQL
-          sudo docker system prune -af
-          df -h
-
       - name: Set up Docker Buildx
         uses: docker/setup-buildx-action@v3
 

.github/workflows/platform-backend-ci.yml

@@ -134,7 +134,7 @@ jobs:
         run: poetry install
 
       - name: Generate Prisma Client
-        run: poetry run prisma generate && poetry run gen-prisma-stub
+        run: poetry run prisma generate
 
       - id: supabase
         name: Start Supabase
@@ -176,7 +176,7 @@ jobs:
           }
 
       - name: Run Database Migrations
-        run: poetry run prisma migrate deploy
+        run: poetry run prisma migrate dev --name updates
         env:
           DATABASE_URL: ${{ steps.supabase.outputs.DB_URL }}
           DIRECT_URL: ${{ steps.supabase.outputs.DB_URL }}

.github/workflows/platform-frontend-ci.yml

@@ -11,7 +11,6 @@ on:
       - ".github/workflows/platform-frontend-ci.yml"
       - "autogpt_platform/frontend/**"
   merge_group:
-  workflow_dispatch:
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.event_name == 'merge_group' && format('merge-queue-{0}', github.ref) || format('{0}-{1}', github.ref, github.event.pull_request.number || github.sha) }}
@@ -152,14 +151,6 @@ jobs:
         run: |
           cp ../.env.default ../.env
 
-      - name: Copy backend .env and set OpenAI API key
-        run: |
-          cp ../backend/.env.default ../backend/.env
-          echo "OPENAI_INTERNAL_API_KEY=${{ secrets.OPENAI_API_KEY }}" >> ../backend/.env
-        env:
-          # Used by E2E test data script to generate embeddings for approved store agents
-          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-
       - name: Set up Docker Buildx
         uses: docker/setup-buildx-action@v3
 
@@ -235,25 +226,13 @@ jobs:
 
       - name: Run Playwright tests
         run: pnpm test:no-build
-        continue-on-error: false
 
-      - name: Upload Playwright report
-        if: always()
+      - name: Upload Playwright artifacts
+        if: failure()
         uses: actions/upload-artifact@v4
         with:
           name: playwright-report
           path: playwright-report
-          if-no-files-found: ignore
-          retention-days: 3
-
-      - name: Upload Playwright test results
-        if: always()
-        uses: actions/upload-artifact@v4
-        with:
-          name: playwright-test-results
-          path: test-results
-          if-no-files-found: ignore
-          retention-days: 3
 
       - name: Print Final Docker Compose logs
         if: always()

autogpt_platform/.claude/settings.local.json

@@ -0,0 +1,9 @@
+{
+  "permissions": {
+    "allow": [
+      "Bash(ls:*)",
+      "WebFetch(domain:langfuse.com)",
+      "Bash(poetry install:*)"
+    ]
+  }
+}

autogpt_platform/Makefile

@@ -1,4 +1,4 @@
-.PHONY: start-core stop-core logs-core format lint migrate run-backend run-frontend load-store-agents
+.PHONY: start-core stop-core logs-core format lint migrate run-backend stop-backend run-frontend load-store-agents backfill-store-embeddings
 
 # Run just Supabase + Redis + RabbitMQ
 start-core:
@@ -6,14 +6,12 @@ start-core:
 
 # Stop core services
 stop-core:
-	docker compose stop 
+	docker compose stop deps
 
 reset-db:
-	docker compose stop db
 	rm -rf db/docker/volumes/db/data
 	cd backend && poetry run prisma migrate deploy
 	cd backend && poetry run prisma generate
-	cd backend && poetry run gen-prisma-stub
 	
 # View logs for core services
 logs-core:
@@ -35,9 +33,15 @@ init-env:
 migrate:
 	cd backend && poetry run prisma migrate deploy
 	cd backend && poetry run prisma generate
-	cd backend && poetry run gen-prisma-stub
 
-run-backend:
+stop-backend:
+	@echo "Stopping backend processes..."
+	@cd backend && poetry run cli stop 2>/dev/null || true
+	@echo "Killing any processes using backend ports..."
+	@lsof -ti:8001,8002,8003,8004,8005,8006,8007 | xargs kill -9 2>/dev/null || true
+	@echo "Backend stopped"
+
+run-backend: stop-backend
 	cd backend && poetry run app
 
 run-frontend:
@@ -49,6 +53,9 @@ test-data:
 load-store-agents:
 	cd backend && poetry run load-store-agents
 
+backfill-store-embeddings:
+	cd backend && poetry run python -m backend.api.features.store.backfill_embeddings
+
 help:
 	@echo "Usage: make <target>"
 	@echo "Targets:"
@@ -58,7 +65,9 @@ help:
 	@echo "  logs-core - Tail the logs for core services"
 	@echo "  format - Format & lint backend (Python) and frontend (TypeScript) code"
 	@echo "  migrate - Run backend database migrations"
-	@echo "  run-backend - Run the backend FastAPI server"
+	@echo "  stop-backend - Stop any running backend processes"
+	@echo "  run-backend - Run the backend FastAPI server (stops existing processes first)"
 	@echo "  run-frontend - Run the frontend Next.js development server"
 	@echo "  test-data - Run the test data creator"
 	@echo "  load-store-agents - Load store agents from agents/ folder into test database"
+	@echo "  backfill-store-embeddings - Generate embeddings for store agents that don't have them"

autogpt_platform/backend/.gitignore

@@ -18,4 +18,3 @@ load-tests/results/
 load-tests/*.json
 load-tests/*.log
 load-tests/node_modules/*
-migrations/*/rollback*.sql

autogpt_platform/backend/Dockerfile

@@ -48,8 +48,7 @@ RUN poetry install --no-ansi --no-root
 # Generate Prisma client
 COPY autogpt_platform/backend/schema.prisma ./
 COPY autogpt_platform/backend/backend/data/partial_types.py ./backend/data/partial_types.py
-COPY autogpt_platform/backend/gen_prisma_types_stub.py ./
-RUN poetry run prisma generate && poetry run gen-prisma-stub
+RUN poetry run prisma generate
 
 FROM debian:13-slim AS server_dependencies
 
@@ -100,7 +99,6 @@ COPY autogpt_platform/backend/migrations /app/autogpt_platform/backend/migration
 FROM server_dependencies AS server
 
 COPY autogpt_platform/backend /app/autogpt_platform/backend
-COPY docs /app/docs
 RUN poetry install --no-ansi --only-root
 
 ENV PORT=8000

autogpt_platform/backend/backend/api/external/fastapi_app.py

@@ -1,57 +1,21 @@
-"""
-External API Application
-
-This module defines the main FastAPI application for the external API,
-which mounts the v1 and v2 sub-applications.
-"""
-
 from fastapi import FastAPI
-from fastapi.responses import RedirectResponse
 
+from backend.api.middleware.security import SecurityHeadersMiddleware
 from backend.monitoring.instrumentation import instrument_fastapi
 
-from .v1.app import v1_app
-from .v2.app import v2_app
-
-DESCRIPTION = """
-The external API provides programmatic access to the AutoGPT Platform for building
-integrations, automations, and custom applications.
-
-### API Versions
-
-| Version             | End of Life | Path                   | Documentation |
-|---------------------|-------------|------------------------|---------------|
-| **v2**              |             | `/external-api/v2/...` | [v2 docs](v2/docs) |
-| **v1** (deprecated) | 2025-05-01  | `/external-api/v1/...` | [v1 docs](v1/docs) |
-
-**Recommendation**: New integrations should use v2.
-
-For authentication details and usage examples, see the
-[API Integration Guide](https://docs.agpt.co/platform/integrating/api-guide/).
-"""
+from .v1.routes import v1_router
 
 external_api = FastAPI(
-    title="AutoGPT Platform API",
-    summary="External API for AutoGPT Platform integrations",
-    description=DESCRIPTION,
-    version="2.0.0",
+    title="AutoGPT External API",
+    description="External API for AutoGPT integrations",
     docs_url="/docs",
-    redoc_url="/redoc",
+    version="1.0",
 )
 
+external_api.add_middleware(SecurityHeadersMiddleware)
+external_api.include_router(v1_router, prefix="/v1")
 
-@external_api.get("/", include_in_schema=False)
-async def root_redirect() -> RedirectResponse:
-    """Redirect root to API documentation."""
-    return RedirectResponse(url="/docs")
-
-
-# Mount versioned sub-applications
-# Each sub-app has its own /docs page at /v1/docs and /v2/docs
-external_api.mount("/v1", v1_app)
-external_api.mount("/v2", v2_app)
-
-# Add Prometheus instrumentation to the main app
+# Add Prometheus instrumentation
 instrument_fastapi(
     external_api,
     service_name="external-api",

autogpt_platform/backend/backend/api/external/v1/app.py

@@ -1,39 +0,0 @@
-"""
-V1 External API Application
-
-This module defines the FastAPI application for the v1 external API.
-"""
-
-from fastapi import FastAPI
-
-from backend.api.middleware.security import SecurityHeadersMiddleware
-
-from .routes import v1_router
-
-DESCRIPTION = """
-The v1 API provides access to core AutoGPT functionality for external integrations.
-
-For authentication details and usage examples, see the
-[API Integration Guide](https://docs.agpt.co/platform/integrating/api-guide/).
-"""
-
-v1_app = FastAPI(
-    title="AutoGPT Platform API",
-    summary="External API for AutoGPT Platform integrations (v1)",
-    description=DESCRIPTION,
-    version="1.0.0",
-    docs_url="/docs",
-    redoc_url="/redoc",
-    openapi_url="/openapi.json",
-    openapi_tags=[
-        {"name": "user", "description": "User information"},
-        {"name": "blocks", "description": "Block operations"},
-        {"name": "graphs", "description": "Graph execution"},
-        {"name": "store", "description": "Marketplace agents and creators"},
-        {"name": "integrations", "description": "OAuth credential management"},
-        {"name": "tools", "description": "AI assistant tools"},
-    ],
-)
-
-v1_app.add_middleware(SecurityHeadersMiddleware)
-v1_app.include_router(v1_router)

autogpt_platform/backend/backend/api/external/v1/tools.py

@@ -70,7 +70,7 @@ class RunAgentRequest(BaseModel):
     )
 
 
-def _create_ephemeral_session(user_id: str) -> ChatSession:
+def _create_ephemeral_session(user_id: str | None) -> ChatSession:
     """Create an ephemeral session for stateless API requests."""
     return ChatSession.new(user_id)
 

autogpt_platform/backend/backend/api/external/v2/__init__.py

@@ -1,9 +0,0 @@
-"""
-V2 External API
-
-This module provides the v2 external API for programmatic access to the AutoGPT Platform.
-"""
-
-from .routes import v2_router
-
-__all__ = ["v2_router"]

autogpt_platform/backend/backend/api/external/v2/app.py

@@ -1,82 +0,0 @@
-"""
-V2 External API Application
-
-This module defines the FastAPI application for the v2 external API.
-"""
-
-from fastapi import FastAPI
-
-from backend.api.middleware.security import SecurityHeadersMiddleware
-
-from .routes import v2_router
-
-DESCRIPTION = """
-The v2 API provides comprehensive access to the AutoGPT Platform for building
-integrations, automations, and custom applications.
-
-### Key Improvements over v1
-
-- **Consistent naming**: Uses `graph_id`/`graph_version` consistently
-- **Better pagination**: All list endpoints support pagination
-- **Comprehensive coverage**: Access to library, runs, schedules, credits, and more
-- **Human-in-the-loop**: Review and approve agent decisions via the API
-
-For authentication details and usage examples, see the
-[API Integration Guide](https://docs.agpt.co/platform/integrating/api-guide/).
-
-### Pagination
-
-List endpoints return paginated responses. Use `page` and `page_size` query
-parameters to navigate results. Maximum page size is 100 items.
-"""
-
-v2_app = FastAPI(
-    title="AutoGPT Platform External API",
-    summary="External API for AutoGPT Platform integrations (v2)",
-    description=DESCRIPTION,
-    version="2.0.0",
-    docs_url="/docs",
-    redoc_url="/redoc",
-    openapi_url="/openapi.json",
-    openapi_tags=[
-        {
-            "name": "graphs",
-            "description": "Create, update, and manage agent graphs",
-        },
-        {
-            "name": "schedules",
-            "description": "Manage scheduled graph executions",
-        },
-        {
-            "name": "blocks",
-            "description": "Discover available building blocks",
-        },
-        {
-            "name": "marketplace",
-            "description": "Browse agents and creators, manage submissions",
-        },
-        {
-            "name": "library",
-            "description": "Access your agent library and execute agents",
-        },
-        {
-            "name": "runs",
-            "description": "Monitor execution runs and human-in-the-loop reviews",
-        },
-        {
-            "name": "credits",
-            "description": "Check balance and view transaction history",
-        },
-        {
-            "name": "integrations",
-            "description": "Manage OAuth credentials for external services",
-        },
-        {
-            "name": "files",
-            "description": "Upload files for agent input",
-        },
-    ],
-)
-
-v2_app.add_middleware(SecurityHeadersMiddleware)
-v2_app.include_router(v2_router)

autogpt_platform/backend/backend/api/external/v2/blocks.py

@@ -1,140 +0,0 @@
-"""
-V2 External API - Blocks Endpoints
-
-Provides read-only access to available building blocks.
-"""
-
-import logging
-from typing import Any
-
-from fastapi import APIRouter, Response, Security
-from fastapi.concurrency import run_in_threadpool
-from prisma.enums import APIKeyPermission
-from pydantic import BaseModel, Field
-
-from backend.api.external.middleware import require_permission
-from backend.data.auth.base import APIAuthorizationInfo
-from backend.data.block import get_blocks
-from backend.util.cache import cached
-from backend.util.json import dumps
-
-logger = logging.getLogger(__name__)
-
-blocks_router = APIRouter()
-
-
-# ============================================================================
-# Models
-# ============================================================================
-
-
-class BlockCost(BaseModel):
-    """Cost information for a block."""
-
-    cost_type: str = Field(description="Type of cost (e.g., 'per_call', 'per_token')")
-    cost_filter: dict[str, Any] = Field(
-        default_factory=dict, description="Conditions for this cost"
-    )
-    cost_amount: int = Field(description="Cost amount in credits")
-
-
-class Block(BaseModel):
-    """A building block that can be used in graphs."""
-
-    id: str
-    name: str
-    description: str
-    categories: list[str] = Field(default_factory=list)
-    input_schema: dict[str, Any]
-    output_schema: dict[str, Any]
-    costs: list[BlockCost] = Field(default_factory=list)
-    disabled: bool = Field(default=False)
-
-
-class BlocksListResponse(BaseModel):
-    """Response for listing blocks."""
-
-    blocks: list[Block]
-    total_count: int
-
-
-# ============================================================================
-# Internal Functions
-# ============================================================================
-
-
-def _compute_blocks_sync() -> str:
-    """
-    Synchronous function to compute blocks data.
-    This does the heavy lifting: instantiate 226+ blocks, compute costs, serialize.
-    """
-    from backend.data.credit import get_block_cost
-
-    block_classes = get_blocks()
-    result = []
-
-    for block_class in block_classes.values():
-        block_instance = block_class()
-        if not block_instance.disabled:
-            costs = get_block_cost(block_instance)
-            # Convert BlockCost BaseModel objects to dictionaries
-            costs_dict = [
-                cost.model_dump() if isinstance(cost, BaseModel) else cost
-                for cost in costs
-            ]
-            result.append({**block_instance.to_dict(), "costs": costs_dict})
-
-    return dumps(result)
-
-
-@cached(ttl_seconds=3600)
-async def _get_cached_blocks() -> str:
-    """
-    Async cached function with thundering herd protection.
-    On cache miss: runs heavy work in thread pool
-    On cache hit: returns cached string immediately
-    """
-    return await run_in_threadpool(_compute_blocks_sync)
-
-
-# ============================================================================
-# Endpoints
-# ============================================================================
-
-
-@blocks_router.get(
-    path="",
-    summary="List available blocks",
-    responses={
-        200: {
-            "description": "List of available building blocks",
-            "content": {
-                "application/json": {
-                    "schema": {
-                        "items": {"additionalProperties": True, "type": "object"},
-                        "type": "array",
-                    }
-                }
-            },
-        }
-    },
-)
-async def list_blocks(
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_BLOCK)
-    ),
-) -> Response:
-    """
-    List all available building blocks that can be used in graphs.
-
-    Each block represents a specific capability (e.g., HTTP request, text processing,
-    AI completion, etc.) that can be connected in a graph to create an agent.
-
-    The response includes input/output schemas for each block, as well as
-    cost information for blocks that consume credits.
-    """
-    content = await _get_cached_blocks()
-    return Response(
-        content=content,
-        media_type="application/json",
-    )

autogpt_platform/backend/backend/api/external/v2/common.py

@@ -1,36 +0,0 @@
-"""
-Common utilities for V2 External API
-"""
-
-from typing import TypeVar
-
-from pydantic import BaseModel, Field
-
-# Constants for pagination
-MAX_PAGE_SIZE = 100
-DEFAULT_PAGE_SIZE = 20
-
-
-class PaginationParams(BaseModel):
-    """Common pagination parameters."""
-
-    page: int = Field(default=1, ge=1, description="Page number (1-indexed)")
-    page_size: int = Field(
-        default=DEFAULT_PAGE_SIZE,
-        ge=1,
-        le=MAX_PAGE_SIZE,
-        description=f"Number of items per page (max {MAX_PAGE_SIZE})",
-    )
-
-
-T = TypeVar("T")
-
-
-class PaginatedResponse(BaseModel):
-    """Generic paginated response wrapper."""
-
-    items: list
-    total_count: int = Field(description="Total number of items across all pages")
-    page: int = Field(description="Current page number (1-indexed)")
-    page_size: int = Field(description="Number of items per page")
-    total_pages: int = Field(description="Total number of pages")

autogpt_platform/backend/backend/api/external/v2/credits.py

@@ -1,141 +0,0 @@
-"""
-V2 External API - Credits Endpoints
-
-Provides access to credit balance and transaction history.
-"""
-
-import logging
-from datetime import datetime
-from typing import Optional
-
-from fastapi import APIRouter, Query, Security
-from prisma.enums import APIKeyPermission
-from pydantic import BaseModel, Field
-
-from backend.api.external.middleware import require_permission
-from backend.data.auth.base import APIAuthorizationInfo
-from backend.data.credit import get_user_credit_model
-
-from .common import DEFAULT_PAGE_SIZE, MAX_PAGE_SIZE
-
-logger = logging.getLogger(__name__)
-
-credits_router = APIRouter()
-
-
-# ============================================================================
-# Models
-# ============================================================================
-
-
-class CreditBalance(BaseModel):
-    """User's credit balance."""
-
-    balance: int = Field(description="Current credit balance")
-
-
-class CreditTransaction(BaseModel):
-    """A credit transaction."""
-
-    transaction_key: str
-    amount: int = Field(description="Transaction amount (positive or negative)")
-    type: str = Field(description="One of: TOP_UP, USAGE, GRANT, REFUND")
-    transaction_time: datetime
-    running_balance: Optional[int] = Field(
-        default=None, description="Balance after this transaction"
-    )
-    description: Optional[str] = None
-
-
-class CreditTransactionsResponse(BaseModel):
-    """Response for listing credit transactions."""
-
-    transactions: list[CreditTransaction]
-    total_count: int
-    page: int
-    page_size: int
-    total_pages: int
-
-
-# ============================================================================
-# Endpoints
-# ============================================================================
-
-
-@credits_router.get(
-    path="",
-    summary="Get credit balance",
-    response_model=CreditBalance,
-)
-async def get_balance(
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_CREDITS)
-    ),
-) -> CreditBalance:
-    """
-    Get the current credit balance for the authenticated user.
-    """
-    user_credit_model = await get_user_credit_model(auth.user_id)
-    balance = await user_credit_model.get_credits(auth.user_id)
-
-    return CreditBalance(balance=balance)
-
-
-@credits_router.get(
-    path="/transactions",
-    summary="Get transaction history",
-    response_model=CreditTransactionsResponse,
-)
-async def get_transactions(
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_CREDITS)
-    ),
-    page: int = Query(default=1, ge=1, description="Page number (1-indexed)"),
-    page_size: int = Query(
-        default=DEFAULT_PAGE_SIZE,
-        ge=1,
-        le=MAX_PAGE_SIZE,
-        description=f"Items per page (max {MAX_PAGE_SIZE})",
-    ),
-    transaction_type: Optional[str] = Query(
-        default=None,
-        description="Filter by transaction type (TOP_UP, USAGE, GRANT, REFUND)",
-    ),
-) -> CreditTransactionsResponse:
-    """
-    Get credit transaction history for the authenticated user.
-
-    Returns transactions sorted by most recent first.
-    """
-    user_credit_model = await get_user_credit_model(auth.user_id)
-
-    history = await user_credit_model.get_transaction_history(
-        user_id=auth.user_id,
-        transaction_count_limit=page_size,
-        transaction_type=transaction_type,
-    )
-
-    transactions = [
-        CreditTransaction(
-            transaction_key=t.transaction_key,
-            amount=t.amount,
-            type=t.transaction_type.value,
-            transaction_time=t.transaction_time,
-            running_balance=t.running_balance,
-            description=t.description,
-        )
-        for t in history.transactions
-    ]
-
-    # Note: The current credit module doesn't support true pagination,
-    # so we're returning what we have
-    total_count = len(transactions)
-    total_pages = 1  # Without true pagination support
-
-    return CreditTransactionsResponse(
-        transactions=transactions,
-        total_count=total_count,
-        page=page,
-        page_size=page_size,
-        total_pages=total_pages,
-    )

autogpt_platform/backend/backend/api/external/v2/files.py

@@ -1,132 +0,0 @@
-"""
-V2 External API - Files Endpoints
-
-Provides file upload functionality for agent inputs.
-"""
-
-import base64
-import logging
-
-from fastapi import APIRouter, File, HTTPException, Query, Security, UploadFile
-from prisma.enums import APIKeyPermission
-from pydantic import BaseModel, Field
-
-from backend.api.external.middleware import require_permission
-from backend.data.auth.base import APIAuthorizationInfo
-from backend.util.cloud_storage import get_cloud_storage_handler
-from backend.util.settings import Settings
-from backend.util.virus_scanner import scan_content_safe
-
-logger = logging.getLogger(__name__)
-settings = Settings()
-
-files_router = APIRouter()
-
-
-# ============================================================================
-# Models
-# ============================================================================
-
-
-class UploadFileResponse(BaseModel):
-    """Response after uploading a file."""
-
-    file_uri: str = Field(description="URI to reference the uploaded file in agents")
-    file_name: str
-    size: int = Field(description="File size in bytes")
-    content_type: str
-    expires_in_hours: int
-
-
-# ============================================================================
-# Endpoints
-# ============================================================================
-
-
-def _create_file_size_error(size_bytes: int, max_size_mb: int) -> HTTPException:
-    """Create standardized file size error response."""
-    return HTTPException(
-        status_code=400,
-        detail=f"File size ({size_bytes} bytes) exceeds the maximum allowed size of {max_size_mb}MB",
-    )
-
-
-@files_router.post(
-    path="/upload",
-    summary="Upload a file",
-    response_model=UploadFileResponse,
-)
-async def upload_file(
-    file: UploadFile = File(...),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.UPLOAD_FILES)
-    ),
-    provider: str = Query(
-        default="gcs", description="Storage provider (gcs, s3, azure)"
-    ),
-    expiration_hours: int = Query(
-        default=24, ge=1, le=48, description="Hours until file expires (1-48)"
-    ),
-) -> UploadFileResponse:
-    """
-    Upload a file to cloud storage for use with agents.
-
-    The returned `file_uri` can be used as input to agents that accept file inputs
-    (e.g., FileStoreBlock, AgentFileInputBlock).
-
-    Files are automatically scanned for viruses before storage.
-    """
-    # Check file size limit
-    max_size_mb = settings.config.upload_file_size_limit_mb
-    max_size_bytes = max_size_mb * 1024 * 1024
-
-    # Try to get file size from headers first
-    if hasattr(file, "size") and file.size is not None and file.size > max_size_bytes:
-        raise _create_file_size_error(file.size, max_size_mb)
-
-    # Read file content
-    content = await file.read()
-    content_size = len(content)
-
-    # Double-check file size after reading
-    if content_size > max_size_bytes:
-        raise _create_file_size_error(content_size, max_size_mb)
-
-    # Extract file info
-    file_name = file.filename or "uploaded_file"
-    content_type = file.content_type or "application/octet-stream"
-
-    # Virus scan the content
-    await scan_content_safe(content, filename=file_name)
-
-    # Check if cloud storage is configured
-    cloud_storage = await get_cloud_storage_handler()
-    if not cloud_storage.config.gcs_bucket_name:
-        # Fallback to base64 data URI when GCS is not configured
-        base64_content = base64.b64encode(content).decode("utf-8")
-        data_uri = f"data:{content_type};base64,{base64_content}"
-
-        return UploadFileResponse(
-            file_uri=data_uri,
-            file_name=file_name,
-            size=content_size,
-            content_type=content_type,
-            expires_in_hours=expiration_hours,
-        )
-
-    # Store in cloud storage
-    storage_path = await cloud_storage.store_file(
-        content=content,
-        filename=file_name,
-        provider=provider,
-        expiration_hours=expiration_hours,
-        user_id=auth.user_id,
-    )
-
-    return UploadFileResponse(
-        file_uri=storage_path,
-        file_name=file_name,
-        size=content_size,
-        content_type=content_type,
-        expires_in_hours=expiration_hours,
-    )

autogpt_platform/backend/backend/api/external/v2/graphs.py

@@ -1,445 +0,0 @@
-"""
-V2 External API - Graphs Endpoints
-
-Provides endpoints for managing agent graphs (CRUD operations).
-"""
-
-import logging
-
-from fastapi import APIRouter, HTTPException, Query, Security
-from prisma.enums import APIKeyPermission
-
-from backend.api.external.middleware import require_permission
-from backend.data import graph as graph_db
-from backend.data.auth.base import APIAuthorizationInfo
-from backend.integrations.webhooks.graph_lifecycle_hooks import (
-    on_graph_activate,
-    on_graph_deactivate,
-)
-
-from .common import DEFAULT_PAGE_SIZE, MAX_PAGE_SIZE
-from .models import (
-    CreateGraphRequest,
-    DeleteGraphResponse,
-    GraphDetails,
-    GraphLink,
-    GraphMeta,
-    GraphNode,
-    GraphSettings,
-    GraphsListResponse,
-    SetActiveVersionRequest,
-)
-
-logger = logging.getLogger(__name__)
-
-graphs_router = APIRouter()
-
-
-def _convert_graph_meta(graph: graph_db.GraphMeta) -> GraphMeta:
-    """Convert internal GraphMeta to v2 API model."""
-    return GraphMeta(
-        id=graph.id,
-        version=graph.version,
-        is_active=graph.is_active,
-        name=graph.name,
-        description=graph.description,
-        created_at=graph.created_at,
-        input_schema=graph.input_schema,
-        output_schema=graph.output_schema,
-    )
-
-
-def _convert_graph_details(graph: graph_db.GraphModel) -> GraphDetails:
-    """Convert internal GraphModel to v2 API GraphDetails model."""
-    return GraphDetails(
-        id=graph.id,
-        version=graph.version,
-        is_active=graph.is_active,
-        name=graph.name,
-        description=graph.description,
-        created_at=graph.created_at,
-        input_schema=graph.input_schema,
-        output_schema=graph.output_schema,
-        nodes=[
-            GraphNode(
-                id=node.id,
-                block_id=node.block_id,
-                input_default=node.input_default,
-                metadata=node.metadata,
-            )
-            for node in graph.nodes
-        ],
-        links=[
-            GraphLink(
-                id=link.id,
-                source_id=link.source_id,
-                sink_id=link.sink_id,
-                source_name=link.source_name,
-                sink_name=link.sink_name,
-                is_static=link.is_static,
-            )
-            for link in graph.links
-        ],
-        credentials_input_schema=graph.credentials_input_schema,
-    )
-
-
-@graphs_router.get(
-    path="",
-    summary="List user's graphs",
-    response_model=GraphsListResponse,
-)
-async def list_graphs(
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_GRAPH)
-    ),
-    page: int = Query(default=1, ge=1, description="Page number (1-indexed)"),
-    page_size: int = Query(
-        default=DEFAULT_PAGE_SIZE,
-        ge=1,
-        le=MAX_PAGE_SIZE,
-        description=f"Items per page (max {MAX_PAGE_SIZE})",
-    ),
-) -> GraphsListResponse:
-    """
-    List all graphs owned by the authenticated user.
-
-    Returns a paginated list of graph metadata (not full graph details).
-    """
-    graphs, pagination_info = await graph_db.list_graphs_paginated(
-        user_id=auth.user_id,
-        page=page,
-        page_size=page_size,
-        filter_by="active",
-    )
-    return GraphsListResponse(
-        graphs=[_convert_graph_meta(g) for g in graphs],
-        total_count=pagination_info.total_items,
-        page=pagination_info.current_page,
-        page_size=pagination_info.page_size,
-        total_pages=pagination_info.total_pages,
-    )
-
-
-@graphs_router.post(
-    path="",
-    summary="Create a new graph",
-    response_model=GraphDetails,
-)
-async def create_graph(
-    create_graph_request: CreateGraphRequest,
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.WRITE_GRAPH)
-    ),
-) -> GraphDetails:
-    """
-    Create a new agent graph.
-
-    The graph will be validated and assigned a new ID. It will automatically
-    be added to the user's library.
-    """
-    # Import here to avoid circular imports
-    from backend.api.features.library import db as library_db
-
-    # Convert v2 API Graph model to internal Graph model
-    internal_graph = graph_db.Graph(
-        id=create_graph_request.graph.id or "",
-        version=create_graph_request.graph.version,
-        is_active=create_graph_request.graph.is_active,
-        name=create_graph_request.graph.name,
-        description=create_graph_request.graph.description,
-        nodes=[
-            graph_db.Node(
-                id=node.id,
-                block_id=node.block_id,
-                input_default=node.input_default,
-                metadata=node.metadata,
-            )
-            for node in create_graph_request.graph.nodes
-        ],
-        links=[
-            graph_db.Link(
-                id=link.id,
-                source_id=link.source_id,
-                sink_id=link.sink_id,
-                source_name=link.source_name,
-                sink_name=link.sink_name,
-                is_static=link.is_static,
-            )
-            for link in create_graph_request.graph.links
-        ],
-    )
-
-    graph = graph_db.make_graph_model(internal_graph, auth.user_id)
-    graph.reassign_ids(user_id=auth.user_id, reassign_graph_id=True)
-    graph.validate_graph(for_run=False)
-
-    await graph_db.create_graph(graph, user_id=auth.user_id)
-    await library_db.create_library_agent(graph, user_id=auth.user_id)
-    activated_graph = await on_graph_activate(graph, user_id=auth.user_id)
-
-    return _convert_graph_details(activated_graph)
-
-
-@graphs_router.get(
-    path="/{graph_id}",
-    summary="Get graph details",
-    response_model=GraphDetails,
-)
-async def get_graph(
-    graph_id: str,
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_GRAPH)
-    ),
-    version: int | None = Query(
-        default=None,
-        description="Specific version to retrieve (default: active version)",
-    ),
-) -> GraphDetails:
-    """
-    Get detailed information about a specific graph.
-
-    By default returns the active version. Use the `version` query parameter
-    to retrieve a specific version.
-    """
-    graph = await graph_db.get_graph(
-        graph_id,
-        version,
-        user_id=auth.user_id,
-        include_subgraphs=True,
-    )
-    if not graph:
-        raise HTTPException(status_code=404, detail=f"Graph #{graph_id} not found.")
-    return _convert_graph_details(graph)
-
-
-@graphs_router.put(
-    path="/{graph_id}",
-    summary="Update graph (creates new version)",
-    response_model=GraphDetails,
-)
-async def update_graph(
-    graph_id: str,
-    graph_request: CreateGraphRequest,
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.WRITE_GRAPH)
-    ),
-) -> GraphDetails:
-    """
-    Update a graph by creating a new version.
-
-    This does not modify existing versions - it creates a new version with
-    the provided content. The new version becomes the active version.
-    """
-    # Import here to avoid circular imports
-    from backend.api.features.library import db as library_db
-
-    graph_data = graph_request.graph
-    if graph_data.id and graph_data.id != graph_id:
-        raise HTTPException(400, detail="Graph ID does not match ID in URI")
-
-    existing_versions = await graph_db.get_graph_all_versions(
-        graph_id, user_id=auth.user_id
-    )
-    if not existing_versions:
-        raise HTTPException(404, detail=f"Graph #{graph_id} not found")
-
-    latest_version_number = max(g.version for g in existing_versions)
-
-    # Convert v2 API Graph model to internal Graph model
-    internal_graph = graph_db.Graph(
-        id=graph_id,
-        version=latest_version_number + 1,
-        is_active=graph_data.is_active,
-        name=graph_data.name,
-        description=graph_data.description,
-        nodes=[
-            graph_db.Node(
-                id=node.id,
-                block_id=node.block_id,
-                input_default=node.input_default,
-                metadata=node.metadata,
-            )
-            for node in graph_data.nodes
-        ],
-        links=[
-            graph_db.Link(
-                id=link.id,
-                source_id=link.source_id,
-                sink_id=link.sink_id,
-                source_name=link.source_name,
-                sink_name=link.sink_name,
-                is_static=link.is_static,
-            )
-            for link in graph_data.links
-        ],
-    )
-
-    current_active_version = next((v for v in existing_versions if v.is_active), None)
-    graph = graph_db.make_graph_model(internal_graph, auth.user_id)
-    graph.reassign_ids(user_id=auth.user_id, reassign_graph_id=False)
-    graph.validate_graph(for_run=False)
-
-    new_graph_version = await graph_db.create_graph(graph, user_id=auth.user_id)
-
-    if new_graph_version.is_active:
-        await library_db.update_agent_version_in_library(
-            auth.user_id, new_graph_version.id, new_graph_version.version
-        )
-        new_graph_version = await on_graph_activate(
-            new_graph_version, user_id=auth.user_id
-        )
-        await graph_db.set_graph_active_version(
-            graph_id=graph_id, version=new_graph_version.version, user_id=auth.user_id
-        )
-        if current_active_version:
-            await on_graph_deactivate(current_active_version, user_id=auth.user_id)
-
-    new_graph_version_with_subgraphs = await graph_db.get_graph(
-        graph_id,
-        new_graph_version.version,
-        user_id=auth.user_id,
-        include_subgraphs=True,
-    )
-    assert new_graph_version_with_subgraphs
-    return _convert_graph_details(new_graph_version_with_subgraphs)
-
-
-@graphs_router.delete(
-    path="/{graph_id}",
-    summary="Delete graph permanently",
-    response_model=DeleteGraphResponse,
-)
-async def delete_graph(
-    graph_id: str,
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.WRITE_GRAPH)
-    ),
-) -> DeleteGraphResponse:
-    """
-    Permanently delete a graph and all its versions.
-
-    This action cannot be undone. All associated executions will remain
-    but will reference a deleted graph.
-    """
-    if active_version := await graph_db.get_graph(
-        graph_id=graph_id, version=None, user_id=auth.user_id
-    ):
-        await on_graph_deactivate(active_version, user_id=auth.user_id)
-
-    version_count = await graph_db.delete_graph(graph_id, user_id=auth.user_id)
-    return DeleteGraphResponse(version_count=version_count)
-
-
-@graphs_router.get(
-    path="/{graph_id}/versions",
-    summary="List all graph versions",
-    response_model=list[GraphDetails],
-)
-async def list_graph_versions(
-    graph_id: str,
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_GRAPH)
-    ),
-) -> list[GraphDetails]:
-    """
-    Get all versions of a specific graph.
-
-    Returns a list of all versions, with the active version marked.
-    """
-    graphs = await graph_db.get_graph_all_versions(graph_id, user_id=auth.user_id)
-    if not graphs:
-        raise HTTPException(status_code=404, detail=f"Graph #{graph_id} not found.")
-    return [_convert_graph_details(g) for g in graphs]
-
-
-@graphs_router.put(
-    path="/{graph_id}/versions/active",
-    summary="Set active graph version",
-)
-async def set_active_version(
-    graph_id: str,
-    request_body: SetActiveVersionRequest,
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.WRITE_GRAPH)
-    ),
-) -> None:
-    """
-    Set which version of a graph is the active version.
-
-    The active version is used when executing the graph without specifying
-    a version number.
-    """
-    # Import here to avoid circular imports
-    from backend.api.features.library import db as library_db
-
-    new_active_version = request_body.active_graph_version
-    new_active_graph = await graph_db.get_graph(
-        graph_id, new_active_version, user_id=auth.user_id
-    )
-    if not new_active_graph:
-        raise HTTPException(404, f"Graph #{graph_id} v{new_active_version} not found")
-
-    current_active_graph = await graph_db.get_graph(
-        graph_id=graph_id,
-        version=None,
-        user_id=auth.user_id,
-    )
-
-    await on_graph_activate(new_active_graph, user_id=auth.user_id)
-    await graph_db.set_graph_active_version(
-        graph_id=graph_id,
-        version=new_active_version,
-        user_id=auth.user_id,
-    )
-
-    await library_db.update_agent_version_in_library(
-        auth.user_id, new_active_graph.id, new_active_graph.version
-    )
-
-    if current_active_graph and current_active_graph.version != new_active_version:
-        await on_graph_deactivate(current_active_graph, user_id=auth.user_id)
-
-
-@graphs_router.patch(
-    path="/{graph_id}/settings",
-    summary="Update graph settings",
-    response_model=GraphSettings,
-)
-async def update_graph_settings(
-    graph_id: str,
-    settings: GraphSettings,
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.WRITE_GRAPH)
-    ),
-) -> GraphSettings:
-    """
-    Update settings for a graph.
-
-    Currently supports:
-    - human_in_the_loop_safe_mode: Enable/disable safe mode for human-in-the-loop blocks
-    """
-    # Import here to avoid circular imports
-    from backend.api.features.library import db as library_db
-    from backend.data.graph import GraphSettings as InternalGraphSettings
-
-    library_agent = await library_db.get_library_agent_by_graph_id(
-        graph_id=graph_id, user_id=auth.user_id
-    )
-    if not library_agent:
-        raise HTTPException(404, f"Graph #{graph_id} not found in user's library")
-
-    # Convert to internal model
-    internal_settings = InternalGraphSettings(
-        human_in_the_loop_safe_mode=settings.human_in_the_loop_safe_mode
-    )
-
-    updated_agent = await library_db.update_library_agent_settings(
-        user_id=auth.user_id,
-        agent_id=library_agent.id,
-        settings=internal_settings,
-    )
-
-    return GraphSettings(
-        human_in_the_loop_safe_mode=updated_agent.settings.human_in_the_loop_safe_mode
-    )

autogpt_platform/backend/backend/api/external/v2/integrations.py

@@ -1,271 +0,0 @@
-"""
-V2 External API - Integrations Endpoints
-
-Provides access to user's integration credentials.
-"""
-
-import logging
-from typing import Optional
-
-from fastapi import APIRouter, HTTPException, Path, Security
-from prisma.enums import APIKeyPermission
-from pydantic import BaseModel, Field
-
-from backend.api.external.middleware import require_permission
-from backend.api.features.library import db as library_db
-from backend.data import graph as graph_db
-from backend.data.auth.base import APIAuthorizationInfo
-from backend.data.model import Credentials, OAuth2Credentials
-from backend.integrations.creds_manager import IntegrationCredentialsManager
-
-logger = logging.getLogger(__name__)
-
-integrations_router = APIRouter()
-creds_manager = IntegrationCredentialsManager()
-
-
-# ============================================================================
-# Models
-# ============================================================================
-
-
-class Credential(BaseModel):
-    """A user's credential for an integration."""
-
-    id: str
-    provider: str = Field(description="Integration provider name")
-    title: Optional[str] = Field(
-        default=None, description="User-assigned title for this credential"
-    )
-    scopes: list[str] = Field(default_factory=list, description="Granted scopes")
-
-
-class CredentialsListResponse(BaseModel):
-    """Response for listing credentials."""
-
-    credentials: list[Credential]
-
-
-class CredentialRequirement(BaseModel):
-    """A credential requirement for a graph or agent."""
-
-    provider: str = Field(description="Required provider name")
-    required_scopes: list[str] = Field(
-        default_factory=list, description="Required scopes"
-    )
-    matching_credentials: list[Credential] = Field(
-        default_factory=list,
-        description="User's credentials that match this requirement",
-    )
-
-
-class CredentialRequirementsResponse(BaseModel):
-    """Response for listing credential requirements."""
-
-    requirements: list[CredentialRequirement]
-
-
-# ============================================================================
-# Conversion Functions
-# ============================================================================
-
-
-def _convert_credential(cred: Credentials) -> Credential:
-    """Convert internal credential to v2 API model."""
-    scopes: list[str] = []
-    if isinstance(cred, OAuth2Credentials):
-        scopes = cred.scopes or []
-
-    return Credential(
-        id=cred.id,
-        provider=cred.provider,
-        title=cred.title,
-        scopes=scopes,
-    )
-
-
-# ============================================================================
-# Endpoints
-# ============================================================================
-
-
-@integrations_router.get(
-    path="/credentials",
-    summary="List all credentials",
-    response_model=CredentialsListResponse,
-)
-async def list_credentials(
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_INTEGRATIONS)
-    ),
-) -> CredentialsListResponse:
-    """
-    List all integration credentials for the authenticated user.
-
-    This returns all OAuth credentials the user has connected, across
-    all integration providers.
-    """
-    credentials = await creds_manager.store.get_all_creds(auth.user_id)
-
-    return CredentialsListResponse(
-        credentials=[_convert_credential(c) for c in credentials]
-    )
-
-
-@integrations_router.get(
-    path="/credentials/{provider}",
-    summary="List credentials by provider",
-    response_model=CredentialsListResponse,
-)
-async def list_credentials_by_provider(
-    provider: str = Path(description="Provider name (e.g., 'github', 'google')"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_INTEGRATIONS)
-    ),
-) -> CredentialsListResponse:
-    """
-    List integration credentials for a specific provider.
-    """
-    all_credentials = await creds_manager.store.get_all_creds(auth.user_id)
-
-    # Filter by provider
-    filtered = [c for c in all_credentials if c.provider.lower() == provider.lower()]
-
-    return CredentialsListResponse(
-        credentials=[_convert_credential(c) for c in filtered]
-    )
-
-
-@integrations_router.get(
-    path="/graphs/{graph_id}/credentials",
-    summary="List credentials matching graph requirements",
-    response_model=CredentialRequirementsResponse,
-)
-async def list_graph_credential_requirements(
-    graph_id: str = Path(description="Graph ID"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_INTEGRATIONS)
-    ),
-) -> CredentialRequirementsResponse:
-    """
-    List credential requirements for a graph and matching user credentials.
-
-    This helps identify which credentials the user needs to provide
-    when executing a graph.
-    """
-    # Get the graph
-    graph = await graph_db.get_graph(
-        graph_id=graph_id,
-        version=None,  # Active version
-        user_id=auth.user_id,
-        include_subgraphs=True,
-    )
-    if not graph:
-        raise HTTPException(status_code=404, detail=f"Graph #{graph_id} not found")
-
-    # Get the credentials input schema which contains provider requirements
-    creds_schema = graph.credentials_input_schema
-    all_credentials = await creds_manager.store.get_all_creds(auth.user_id)
-
-    requirements = []
-    for field_name, field_schema in creds_schema.get("properties", {}).items():
-        # Extract provider from schema
-        # The schema structure varies, but typically has provider info
-        providers = []
-        if "anyOf" in field_schema:
-            for option in field_schema["anyOf"]:
-                if "provider" in option:
-                    providers.append(option["provider"])
-        elif "provider" in field_schema:
-            providers.append(field_schema["provider"])
-
-        for provider in providers:
-            # Find matching credentials
-            matching = [
-                _convert_credential(c)
-                for c in all_credentials
-                if c.provider.lower() == provider.lower()
-            ]
-
-            requirements.append(
-                CredentialRequirement(
-                    provider=provider,
-                    required_scopes=[],  # Would need to extract from schema
-                    matching_credentials=matching,
-                )
-            )
-
-    return CredentialRequirementsResponse(requirements=requirements)
-
-
-@integrations_router.get(
-    path="/library/{agent_id}/credentials",
-    summary="List credentials matching library agent requirements",
-    response_model=CredentialRequirementsResponse,
-)
-async def list_library_agent_credential_requirements(
-    agent_id: str = Path(description="Library agent ID"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_INTEGRATIONS)
-    ),
-) -> CredentialRequirementsResponse:
-    """
-    List credential requirements for a library agent and matching user credentials.
-
-    This helps identify which credentials the user needs to provide
-    when executing an agent from their library.
-    """
-    # Get the library agent
-    try:
-        library_agent = await library_db.get_library_agent(
-            id=agent_id,
-            user_id=auth.user_id,
-        )
-    except Exception:
-        raise HTTPException(status_code=404, detail=f"Agent #{agent_id} not found")
-
-    # Get the underlying graph
-    graph = await graph_db.get_graph(
-        graph_id=library_agent.graph_id,
-        version=library_agent.graph_version,
-        user_id=auth.user_id,
-        include_subgraphs=True,
-    )
-    if not graph:
-        raise HTTPException(
-            status_code=404,
-            detail=f"Graph for agent #{agent_id} not found",
-        )
-
-    # Get the credentials input schema
-    creds_schema = graph.credentials_input_schema
-    all_credentials = await creds_manager.store.get_all_creds(auth.user_id)
-
-    requirements = []
-    for field_name, field_schema in creds_schema.get("properties", {}).items():
-        # Extract provider from schema
-        providers = []
-        if "anyOf" in field_schema:
-            for option in field_schema["anyOf"]:
-                if "provider" in option:
-                    providers.append(option["provider"])
-        elif "provider" in field_schema:
-            providers.append(field_schema["provider"])
-
-        for provider in providers:
-            # Find matching credentials
-            matching = [
-                _convert_credential(c)
-                for c in all_credentials
-                if c.provider.lower() == provider.lower()
-            ]
-
-            requirements.append(
-                CredentialRequirement(
-                    provider=provider,
-                    required_scopes=[],
-                    matching_credentials=matching,
-                )
-            )
-
-    return CredentialRequirementsResponse(requirements=requirements)

autogpt_platform/backend/backend/api/external/v2/library.py

@@ -1,247 +0,0 @@
-"""
-V2 External API - Library Endpoints
-
-Provides access to the user's agent library and agent execution.
-"""
-
-import logging
-
-from fastapi import APIRouter, HTTPException, Path, Query, Security
-from prisma.enums import APIKeyPermission
-
-from backend.api.external.middleware import require_permission
-from backend.api.features.library import db as library_db
-from backend.api.features.library import model as library_model
-from backend.data import execution as execution_db
-from backend.data.auth.base import APIAuthorizationInfo
-from backend.data.credit import get_user_credit_model
-from backend.executor import utils as execution_utils
-
-from .common import DEFAULT_PAGE_SIZE, MAX_PAGE_SIZE
-from .models import (
-    ExecuteAgentRequest,
-    LibraryAgent,
-    LibraryAgentsResponse,
-    Run,
-    RunsListResponse,
-)
-
-logger = logging.getLogger(__name__)
-
-library_router = APIRouter()
-
-
-# ============================================================================
-# Conversion Functions
-# ============================================================================
-
-
-def _convert_library_agent(agent: library_model.LibraryAgent) -> LibraryAgent:
-    """Convert internal LibraryAgent to v2 API model."""
-    return LibraryAgent(
-        id=agent.id,
-        graph_id=agent.graph_id,
-        graph_version=agent.graph_version,
-        name=agent.name,
-        description=agent.description,
-        is_favorite=agent.is_favorite,
-        can_access_graph=agent.can_access_graph,
-        is_latest_version=agent.is_latest_version,
-        image_url=agent.image_url,
-        creator_name=agent.creator_name,
-        input_schema=agent.input_schema,
-        output_schema=agent.output_schema,
-        created_at=agent.created_at,
-        updated_at=agent.updated_at,
-    )
-
-
-def _convert_execution_to_run(exec: execution_db.GraphExecutionMeta) -> Run:
-    """Convert internal execution to v2 API Run model."""
-    return Run(
-        id=exec.id,
-        graph_id=exec.graph_id,
-        graph_version=exec.graph_version,
-        status=exec.status.value,
-        started_at=exec.started_at,
-        ended_at=exec.ended_at,
-        inputs=exec.inputs,
-        cost=exec.stats.cost if exec.stats else 0,
-        duration=exec.stats.duration if exec.stats else 0,
-        node_count=exec.stats.node_exec_count if exec.stats else 0,
-    )
-
-
-# ============================================================================
-# Endpoints
-# ============================================================================
-
-
-@library_router.get(
-    path="/agents",
-    summary="List library agents",
-    response_model=LibraryAgentsResponse,
-)
-async def list_library_agents(
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_LIBRARY)
-    ),
-    page: int = Query(default=1, ge=1, description="Page number (1-indexed)"),
-    page_size: int = Query(
-        default=DEFAULT_PAGE_SIZE,
-        ge=1,
-        le=MAX_PAGE_SIZE,
-        description=f"Items per page (max {MAX_PAGE_SIZE})",
-    ),
-) -> LibraryAgentsResponse:
-    """
-    List agents in the user's library.
-
-    The library contains agents the user has created or added from the marketplace.
-    """
-    result = await library_db.list_library_agents(
-        user_id=auth.user_id,
-        page=page,
-        page_size=page_size,
-    )
-
-    return LibraryAgentsResponse(
-        agents=[_convert_library_agent(a) for a in result.agents],
-        total_count=result.pagination.total_items,
-        page=result.pagination.current_page,
-        page_size=result.pagination.page_size,
-        total_pages=result.pagination.total_pages,
-    )
-
-
-@library_router.get(
-    path="/agents/favorites",
-    summary="List favorite agents",
-    response_model=LibraryAgentsResponse,
-)
-async def list_favorite_agents(
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_LIBRARY)
-    ),
-    page: int = Query(default=1, ge=1, description="Page number (1-indexed)"),
-    page_size: int = Query(
-        default=DEFAULT_PAGE_SIZE,
-        ge=1,
-        le=MAX_PAGE_SIZE,
-        description=f"Items per page (max {MAX_PAGE_SIZE})",
-    ),
-) -> LibraryAgentsResponse:
-    """
-    List favorite agents in the user's library.
-    """
-    result = await library_db.list_favorite_library_agents(
-        user_id=auth.user_id,
-        page=page,
-        page_size=page_size,
-    )
-
-    return LibraryAgentsResponse(
-        agents=[_convert_library_agent(a) for a in result.agents],
-        total_count=result.pagination.total_items,
-        page=result.pagination.current_page,
-        page_size=result.pagination.page_size,
-        total_pages=result.pagination.total_pages,
-    )
-
-
-@library_router.post(
-    path="/agents/{agent_id}/runs",
-    summary="Execute an agent",
-    response_model=Run,
-)
-async def execute_agent(
-    request: ExecuteAgentRequest,
-    agent_id: str = Path(description="Library agent ID"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.RUN_AGENT)
-    ),
-) -> Run:
-    """
-    Execute an agent from the library.
-
-    This creates a new run with the provided inputs. The run executes
-    asynchronously and you can poll the run status using GET /runs/{run_id}.
-    """
-    # Check credit balance
-    user_credit_model = await get_user_credit_model(auth.user_id)
-    current_balance = await user_credit_model.get_credits(auth.user_id)
-    if current_balance <= 0:
-        raise HTTPException(
-            status_code=402,
-            detail="Insufficient balance to execute the agent. Please top up your account.",
-        )
-
-    # Get the library agent to find the graph ID and version
-    try:
-        library_agent = await library_db.get_library_agent(
-            id=agent_id,
-            user_id=auth.user_id,
-        )
-    except Exception:
-        raise HTTPException(status_code=404, detail=f"Agent #{agent_id} not found")
-
-    try:
-        result = await execution_utils.add_graph_execution(
-            graph_id=library_agent.graph_id,
-            user_id=auth.user_id,
-            inputs=request.inputs,
-            graph_version=library_agent.graph_version,
-            graph_credentials_inputs=request.credentials_inputs,
-        )
-
-        return _convert_execution_to_run(result)
-
-    except Exception as e:
-        logger.error(f"Failed to execute agent: {e}")
-        raise HTTPException(status_code=400, detail=str(e))
-
-
-@library_router.get(
-    path="/agents/{agent_id}/runs",
-    summary="List runs for an agent",
-    response_model=RunsListResponse,
-)
-async def list_agent_runs(
-    agent_id: str = Path(description="Library agent ID"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_LIBRARY)
-    ),
-    page: int = Query(default=1, ge=1, description="Page number (1-indexed)"),
-    page_size: int = Query(
-        default=DEFAULT_PAGE_SIZE,
-        ge=1,
-        le=MAX_PAGE_SIZE,
-        description=f"Items per page (max {MAX_PAGE_SIZE})",
-    ),
-) -> RunsListResponse:
-    """
-    List execution runs for a specific agent.
-    """
-    # Get the library agent to find the graph ID
-    try:
-        library_agent = await library_db.get_library_agent(
-            id=agent_id,
-            user_id=auth.user_id,
-        )
-    except Exception:
-        raise HTTPException(status_code=404, detail=f"Agent #{agent_id} not found")
-
-    result = await execution_db.get_graph_executions_paginated(
-        graph_id=library_agent.graph_id,
-        user_id=auth.user_id,
-        page=page,
-        page_size=page_size,
-    )
-
-    return RunsListResponse(
-        runs=[_convert_execution_to_run(e) for e in result.executions],
-        total_count=result.pagination.total_items,
-        page=result.pagination.current_page,
-        page_size=result.pagination.page_size,
-        total_pages=result.pagination.total_pages,
-    )

autogpt_platform/backend/backend/api/external/v2/marketplace.py

@@ -1,510 +0,0 @@
-"""
-V2 External API - Marketplace Endpoints
-
-Provides access to the agent marketplace (store).
-"""
-
-import logging
-import urllib.parse
-from datetime import datetime
-from typing import Literal, Optional
-
-from fastapi import APIRouter, HTTPException, Path, Query, Security
-from prisma.enums import APIKeyPermission
-from pydantic import BaseModel, Field
-
-from backend.api.external.middleware import require_permission
-from backend.api.features.store import cache as store_cache
-from backend.api.features.store import db as store_db
-from backend.api.features.store import model as store_model
-from backend.data.auth.base import APIAuthorizationInfo
-
-from .common import DEFAULT_PAGE_SIZE, MAX_PAGE_SIZE
-
-logger = logging.getLogger(__name__)
-
-marketplace_router = APIRouter()
-
-
-# ============================================================================
-# Models
-# ============================================================================
-
-
-class MarketplaceAgent(BaseModel):
-    """An agent available in the marketplace."""
-
-    slug: str
-    name: str
-    description: str
-    sub_heading: str
-    creator: str
-    creator_avatar: str
-    runs: int = Field(default=0, description="Number of times this agent has been run")
-    rating: float = Field(default=0.0, description="Average rating")
-    image_url: str = Field(default="")
-
-
-class MarketplaceAgentDetails(BaseModel):
-    """Detailed information about a marketplace agent."""
-
-    store_listing_version_id: str
-    slug: str
-    name: str
-    description: str
-    sub_heading: str
-    instructions: Optional[str] = None
-    creator: str
-    creator_avatar: str
-    categories: list[str] = Field(default_factory=list)
-    runs: int = Field(default=0)
-    rating: float = Field(default=0.0)
-    image_urls: list[str] = Field(default_factory=list)
-    video_url: str = Field(default="")
-    versions: list[str] = Field(default_factory=list, description="Available versions")
-    agent_graph_versions: list[str] = Field(default_factory=list)
-    agent_graph_id: str
-    last_updated: datetime
-
-
-class MarketplaceAgentsResponse(BaseModel):
-    """Response for listing marketplace agents."""
-
-    agents: list[MarketplaceAgent]
-    total_count: int
-    page: int
-    page_size: int
-    total_pages: int
-
-
-class MarketplaceCreator(BaseModel):
-    """A creator on the marketplace."""
-
-    name: str
-    username: str
-    description: str
-    avatar_url: str
-    num_agents: int
-    agent_rating: float
-    agent_runs: int
-    is_featured: bool = False
-
-
-class MarketplaceCreatorDetails(BaseModel):
-    """Detailed information about a marketplace creator."""
-
-    name: str
-    username: str
-    description: str
-    avatar_url: str
-    agent_rating: float
-    agent_runs: int
-    top_categories: list[str] = Field(default_factory=list)
-    links: list[str] = Field(default_factory=list)
-
-
-class MarketplaceCreatorsResponse(BaseModel):
-    """Response for listing marketplace creators."""
-
-    creators: list[MarketplaceCreator]
-    total_count: int
-    page: int
-    page_size: int
-    total_pages: int
-
-
-class MarketplaceSubmission(BaseModel):
-    """A marketplace submission."""
-
-    graph_id: str
-    graph_version: int
-    name: str
-    sub_heading: str
-    slug: str
-    description: str
-    instructions: Optional[str] = None
-    image_urls: list[str] = Field(default_factory=list)
-    date_submitted: datetime
-    status: str = Field(description="One of: DRAFT, PENDING, APPROVED, REJECTED")
-    runs: int = Field(default=0)
-    rating: float = Field(default=0.0)
-    store_listing_version_id: Optional[str] = None
-    version: Optional[int] = None
-    review_comments: Optional[str] = None
-    reviewed_at: Optional[datetime] = None
-    video_url: Optional[str] = None
-    categories: list[str] = Field(default_factory=list)
-
-
-class SubmissionsListResponse(BaseModel):
-    """Response for listing submissions."""
-
-    submissions: list[MarketplaceSubmission]
-    total_count: int
-    page: int
-    page_size: int
-    total_pages: int
-
-
-class CreateSubmissionRequest(BaseModel):
-    """Request to create a marketplace submission."""
-
-    graph_id: str = Field(description="ID of the graph to submit")
-    graph_version: int = Field(description="Version of the graph to submit")
-    name: str = Field(description="Display name for the agent")
-    slug: str = Field(description="URL-friendly identifier")
-    description: str = Field(description="Full description")
-    sub_heading: str = Field(description="Short tagline")
-    image_urls: list[str] = Field(default_factory=list)
-    video_url: Optional[str] = None
-    categories: list[str] = Field(default_factory=list)
-
-
-# ============================================================================
-# Conversion Functions
-# ============================================================================
-
-
-def _convert_store_agent(agent: store_model.StoreAgent) -> MarketplaceAgent:
-    """Convert internal StoreAgent to v2 API model."""
-    return MarketplaceAgent(
-        slug=agent.slug,
-        name=agent.agent_name,
-        description=agent.description,
-        sub_heading=agent.sub_heading,
-        creator=agent.creator,
-        creator_avatar=agent.creator_avatar,
-        runs=agent.runs,
-        rating=agent.rating,
-        image_url=agent.agent_image,
-    )
-
-
-def _convert_store_agent_details(
-    agent: store_model.StoreAgentDetails,
-) -> MarketplaceAgentDetails:
-    """Convert internal StoreAgentDetails to v2 API model."""
-    return MarketplaceAgentDetails(
-        store_listing_version_id=agent.store_listing_version_id,
-        slug=agent.slug,
-        name=agent.agent_name,
-        description=agent.description,
-        sub_heading=agent.sub_heading,
-        instructions=agent.instructions,
-        creator=agent.creator,
-        creator_avatar=agent.creator_avatar,
-        categories=agent.categories,
-        runs=agent.runs,
-        rating=agent.rating,
-        image_urls=agent.agent_image,
-        video_url=agent.agent_video,
-        versions=agent.versions,
-        agent_graph_versions=agent.agentGraphVersions,
-        agent_graph_id=agent.agentGraphId,
-        last_updated=agent.last_updated,
-    )
-
-
-def _convert_creator(creator: store_model.Creator) -> MarketplaceCreator:
-    """Convert internal Creator to v2 API model."""
-    return MarketplaceCreator(
-        name=creator.name,
-        username=creator.username,
-        description=creator.description,
-        avatar_url=creator.avatar_url,
-        num_agents=creator.num_agents,
-        agent_rating=creator.agent_rating,
-        agent_runs=creator.agent_runs,
-        is_featured=creator.is_featured,
-    )
-
-
-def _convert_creator_details(
-    creator: store_model.CreatorDetails,
-) -> MarketplaceCreatorDetails:
-    """Convert internal CreatorDetails to v2 API model."""
-    return MarketplaceCreatorDetails(
-        name=creator.name,
-        username=creator.username,
-        description=creator.description,
-        avatar_url=creator.avatar_url,
-        agent_rating=creator.agent_rating,
-        agent_runs=creator.agent_runs,
-        top_categories=creator.top_categories,
-        links=creator.links,
-    )
-
-
-def _convert_submission(sub: store_model.StoreSubmission) -> MarketplaceSubmission:
-    """Convert internal StoreSubmission to v2 API model."""
-    return MarketplaceSubmission(
-        graph_id=sub.agent_id,
-        graph_version=sub.agent_version,
-        name=sub.name,
-        sub_heading=sub.sub_heading,
-        slug=sub.slug,
-        description=sub.description,
-        instructions=sub.instructions,
-        image_urls=sub.image_urls,
-        date_submitted=sub.date_submitted,
-        status=sub.status.value,
-        runs=sub.runs,
-        rating=sub.rating,
-        store_listing_version_id=sub.store_listing_version_id,
-        version=sub.version,
-        review_comments=sub.review_comments,
-        reviewed_at=sub.reviewed_at,
-        video_url=sub.video_url,
-        categories=sub.categories,
-    )
-
-
-# ============================================================================
-# Endpoints - Read (authenticated)
-# ============================================================================
-
-
-@marketplace_router.get(
-    path="/agents",
-    summary="List marketplace agents",
-    response_model=MarketplaceAgentsResponse,
-)
-async def list_agents(
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_STORE)
-    ),
-    featured: bool = Query(default=False, description="Filter to featured agents only"),
-    creator: Optional[str] = Query(
-        default=None, description="Filter by creator username"
-    ),
-    sorted_by: Optional[Literal["rating", "runs", "name", "updated_at"]] = Query(
-        default=None, description="Sort field"
-    ),
-    search_query: Optional[str] = Query(default=None, description="Search query"),
-    category: Optional[str] = Query(default=None, description="Filter by category"),
-    page: int = Query(default=1, ge=1, description="Page number (1-indexed)"),
-    page_size: int = Query(
-        default=DEFAULT_PAGE_SIZE,
-        ge=1,
-        le=MAX_PAGE_SIZE,
-        description=f"Items per page (max {MAX_PAGE_SIZE})",
-    ),
-) -> MarketplaceAgentsResponse:
-    """
-    List agents available in the marketplace.
-
-    Supports filtering by featured status, creator, category, and search query.
-    Results can be sorted by rating, runs, name, or update time.
-    """
-    result = await store_cache._get_cached_store_agents(
-        featured=featured,
-        creator=creator,
-        sorted_by=sorted_by,
-        search_query=search_query,
-        category=category,
-        page=page,
-        page_size=page_size,
-    )
-
-    return MarketplaceAgentsResponse(
-        agents=[_convert_store_agent(a) for a in result.agents],
-        total_count=result.pagination.total_items,
-        page=result.pagination.current_page,
-        page_size=result.pagination.page_size,
-        total_pages=result.pagination.total_pages,
-    )
-
-
-@marketplace_router.get(
-    path="/agents/{username}/{agent_name}",
-    summary="Get agent details",
-    response_model=MarketplaceAgentDetails,
-)
-async def get_agent_details(
-    username: str = Path(description="Creator username"),
-    agent_name: str = Path(description="Agent slug/name"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_STORE)
-    ),
-) -> MarketplaceAgentDetails:
-    """
-    Get detailed information about a specific marketplace agent.
-    """
-    username = urllib.parse.unquote(username).lower()
-    agent_name = urllib.parse.unquote(agent_name).lower()
-
-    agent = await store_cache._get_cached_agent_details(
-        username=username, agent_name=agent_name
-    )
-
-    return _convert_store_agent_details(agent)
-
-
-@marketplace_router.get(
-    path="/creators",
-    summary="List marketplace creators",
-    response_model=MarketplaceCreatorsResponse,
-)
-async def list_creators(
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_STORE)
-    ),
-    featured: bool = Query(
-        default=False, description="Filter to featured creators only"
-    ),
-    search_query: Optional[str] = Query(default=None, description="Search query"),
-    sorted_by: Optional[Literal["agent_rating", "agent_runs", "num_agents"]] = Query(
-        default=None, description="Sort field"
-    ),
-    page: int = Query(default=1, ge=1, description="Page number (1-indexed)"),
-    page_size: int = Query(
-        default=DEFAULT_PAGE_SIZE,
-        ge=1,
-        le=MAX_PAGE_SIZE,
-        description=f"Items per page (max {MAX_PAGE_SIZE})",
-    ),
-) -> MarketplaceCreatorsResponse:
-    """
-    List creators on the marketplace.
-
-    Supports filtering by featured status and search query.
-    Results can be sorted by rating, runs, or number of agents.
-    """
-    result = await store_cache._get_cached_store_creators(
-        featured=featured,
-        search_query=search_query,
-        sorted_by=sorted_by,
-        page=page,
-        page_size=page_size,
-    )
-
-    return MarketplaceCreatorsResponse(
-        creators=[_convert_creator(c) for c in result.creators],
-        total_count=result.pagination.total_items,
-        page=result.pagination.current_page,
-        page_size=result.pagination.page_size,
-        total_pages=result.pagination.total_pages,
-    )
-
-
-@marketplace_router.get(
-    path="/creators/{username}",
-    summary="Get creator details",
-    response_model=MarketplaceCreatorDetails,
-)
-async def get_creator_details(
-    username: str = Path(description="Creator username"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_STORE)
-    ),
-) -> MarketplaceCreatorDetails:
-    """
-    Get detailed information about a specific marketplace creator.
-    """
-    username = urllib.parse.unquote(username).lower()
-
-    creator = await store_cache._get_cached_creator_details(username=username)
-
-    return _convert_creator_details(creator)
-
-
-# ============================================================================
-# Endpoints - Submissions (CRUD)
-# ============================================================================
-
-
-@marketplace_router.get(
-    path="/submissions",
-    summary="List my submissions",
-    response_model=SubmissionsListResponse,
-)
-async def list_submissions(
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_STORE)
-    ),
-    page: int = Query(default=1, ge=1, description="Page number (1-indexed)"),
-    page_size: int = Query(
-        default=DEFAULT_PAGE_SIZE,
-        ge=1,
-        le=MAX_PAGE_SIZE,
-        description=f"Items per page (max {MAX_PAGE_SIZE})",
-    ),
-) -> SubmissionsListResponse:
-    """
-    List your marketplace submissions.
-
-    Returns all submissions you've created, including drafts, pending,
-    approved, and rejected submissions.
-    """
-    result = await store_db.get_store_submissions(
-        user_id=auth.user_id,
-        page=page,
-        page_size=page_size,
-    )
-
-    return SubmissionsListResponse(
-        submissions=[_convert_submission(s) for s in result.submissions],
-        total_count=result.pagination.total_items,
-        page=result.pagination.current_page,
-        page_size=result.pagination.page_size,
-        total_pages=result.pagination.total_pages,
-    )
-
-
-@marketplace_router.post(
-    path="/submissions",
-    summary="Create a submission",
-    response_model=MarketplaceSubmission,
-)
-async def create_submission(
-    request: CreateSubmissionRequest,
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.WRITE_STORE)
-    ),
-) -> MarketplaceSubmission:
-    """
-    Create a new marketplace submission.
-
-    This submits an agent for review to be published in the marketplace.
-    The submission will be in PENDING status until reviewed by the team.
-    """
-    submission = await store_db.create_store_submission(
-        user_id=auth.user_id,
-        agent_id=request.graph_id,
-        agent_version=request.graph_version,
-        slug=request.slug,
-        name=request.name,
-        sub_heading=request.sub_heading,
-        description=request.description,
-        image_urls=request.image_urls,
-        video_url=request.video_url,
-        categories=request.categories,
-    )
-
-    return _convert_submission(submission)
-
-
-@marketplace_router.delete(
-    path="/submissions/{submission_id}",
-    summary="Delete a submission",
-)
-async def delete_submission(
-    submission_id: str = Path(description="Submission ID"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.WRITE_STORE)
-    ),
-) -> None:
-    """
-    Delete a marketplace submission.
-
-    Only submissions in DRAFT status can be deleted.
-    """
-    success = await store_db.delete_store_submission(
-        user_id=auth.user_id,
-        submission_id=submission_id,
-    )
-
-    if not success:
-        raise HTTPException(
-            status_code=404, detail=f"Submission #{submission_id} not found"
-        )

autogpt_platform/backend/backend/api/external/v2/models.py

@@ -1,552 +0,0 @@
-"""
-V2 External API - Request and Response Models
-
-This module defines all request and response models for the v2 external API.
-All models are self-contained and specific to the external API contract.
-"""
-
-from datetime import datetime
-from typing import Any, Optional
-
-from pydantic import BaseModel, Field
-
-# ============================================================================
-# Common/Shared Models
-# ============================================================================
-
-
-class PaginatedResponse(BaseModel):
-    """Base class for paginated responses."""
-
-    total_count: int = Field(description="Total number of items across all pages")
-    page: int = Field(description="Current page number (1-indexed)")
-    page_size: int = Field(description="Number of items per page")
-    total_pages: int = Field(description="Total number of pages")
-
-
-# ============================================================================
-# Graph Models
-# ============================================================================
-
-
-class GraphLink(BaseModel):
-    """A link between two nodes in a graph."""
-
-    id: str
-    source_id: str = Field(description="ID of the source node")
-    sink_id: str = Field(description="ID of the target node")
-    source_name: str = Field(description="Output pin name on source node")
-    sink_name: str = Field(description="Input pin name on target node")
-    is_static: bool = Field(
-        default=False, description="Whether this link provides static data"
-    )
-
-
-class GraphNode(BaseModel):
-    """A node in an agent graph."""
-
-    id: str
-    block_id: str = Field(description="ID of the block type")
-    input_default: dict[str, Any] = Field(
-        default_factory=dict, description="Default input values"
-    )
-    metadata: dict[str, Any] = Field(
-        default_factory=dict, description="Node metadata (e.g., position)"
-    )
-
-
-class Graph(BaseModel):
-    """Graph definition for creating or updating an agent."""
-
-    id: Optional[str] = Field(default=None, description="Graph ID (assigned by server)")
-    version: int = Field(default=1, description="Graph version")
-    is_active: bool = Field(default=True, description="Whether this version is active")
-    name: str = Field(description="Graph name")
-    description: str = Field(default="", description="Graph description")
-    nodes: list[GraphNode] = Field(default_factory=list, description="List of nodes")
-    links: list[GraphLink] = Field(
-        default_factory=list, description="Links between nodes"
-    )
-
-
-class GraphMeta(BaseModel):
-    """Graph metadata (summary information)."""
-
-    id: str
-    version: int
-    is_active: bool
-    name: str
-    description: str
-    created_at: datetime
-    input_schema: dict[str, Any] = Field(description="Input schema for the graph")
-    output_schema: dict[str, Any] = Field(description="Output schema for the graph")
-
-
-class GraphDetails(GraphMeta):
-    """Full graph details including nodes and links."""
-
-    nodes: list[GraphNode]
-    links: list[GraphLink]
-    credentials_input_schema: dict[str, Any] = Field(
-        description="Schema for required credentials"
-    )
-
-
-class GraphSettings(BaseModel):
-    """Settings for a graph."""
-
-    human_in_the_loop_safe_mode: Optional[bool] = Field(
-        default=None, description="Enable safe mode for human-in-the-loop blocks"
-    )
-
-
-class CreateGraphRequest(BaseModel):
-    """Request to create a new graph."""
-
-    graph: Graph = Field(description="The graph definition")
-
-
-class SetActiveVersionRequest(BaseModel):
-    """Request to set the active graph version."""
-
-    active_graph_version: int = Field(description="Version number to set as active")
-
-
-class GraphsListResponse(PaginatedResponse):
-    """Response for listing graphs."""
-
-    graphs: list[GraphMeta]
-
-
-class DeleteGraphResponse(BaseModel):
-    """Response for deleting a graph."""
-
-    version_count: int = Field(description="Number of versions deleted")
-
-
-# ============================================================================
-# Schedule Models
-# ============================================================================
-
-
-class Schedule(BaseModel):
-    """An execution schedule for a graph."""
-
-    id: str
-    name: str
-    graph_id: str
-    graph_version: int
-    cron: str = Field(description="Cron expression for the schedule")
-    input_data: dict[str, Any] = Field(
-        default_factory=dict, description="Input data for scheduled executions"
-    )
-    is_enabled: bool = Field(default=True, description="Whether schedule is enabled")
-    next_run_time: Optional[datetime] = Field(
-        default=None, description="Next scheduled run time"
-    )
-
-
-class CreateScheduleRequest(BaseModel):
-    """Request to create a schedule."""
-
-    name: str = Field(description="Display name for the schedule")
-    cron: str = Field(description="Cron expression (e.g., '0 9 * * *' for 9am daily)")
-    input_data: dict[str, Any] = Field(
-        default_factory=dict, description="Input data for scheduled executions"
-    )
-    credentials_inputs: dict[str, Any] = Field(
-        default_factory=dict, description="Credentials for the schedule"
-    )
-    graph_version: Optional[int] = Field(
-        default=None, description="Graph version (default: active version)"
-    )
-    timezone: Optional[str] = Field(
-        default=None,
-        description="Timezone for schedule (e.g., 'America/New_York')",
-    )
-
-
-class SchedulesListResponse(PaginatedResponse):
-    """Response for listing schedules."""
-
-    schedules: list[Schedule]
-
-
-# ============================================================================
-# Block Models
-# ============================================================================
-
-
-class BlockCost(BaseModel):
-    """Cost information for a block."""
-
-    cost_type: str = Field(description="Type of cost (e.g., 'per_call', 'per_token')")
-    cost_filter: dict[str, Any] = Field(
-        default_factory=dict, description="Conditions for this cost"
-    )
-    cost_amount: int = Field(description="Cost amount in credits")
-
-
-class Block(BaseModel):
-    """A building block that can be used in graphs."""
-
-    id: str
-    name: str
-    description: str
-    categories: list[str] = Field(default_factory=list)
-    input_schema: dict[str, Any]
-    output_schema: dict[str, Any]
-    costs: list[BlockCost] = Field(default_factory=list)
-
-
-class BlocksListResponse(BaseModel):
-    """Response for listing blocks."""
-
-    blocks: list[Block]
-
-
-# ============================================================================
-# Marketplace Models
-# ============================================================================
-
-
-class MarketplaceAgent(BaseModel):
-    """An agent available in the marketplace."""
-
-    slug: str
-    agent_name: str
-    agent_image: str
-    creator: str
-    creator_avatar: str
-    sub_heading: str
-    description: str
-    runs: int = Field(default=0, description="Number of times this agent has been run")
-    rating: float = Field(default=0.0, description="Average rating")
-
-
-class MarketplaceAgentDetails(BaseModel):
-    """Detailed information about a marketplace agent."""
-
-    store_listing_version_id: str
-    slug: str
-    agent_name: str
-    agent_video: str
-    agent_output_demo: str
-    agent_image: list[str]
-    creator: str
-    creator_avatar: str
-    sub_heading: str
-    description: str
-    instructions: Optional[str] = None
-    categories: list[str]
-    runs: int
-    rating: float
-    versions: list[str]
-    agent_graph_versions: list[str]
-    agent_graph_id: str
-    last_updated: datetime
-    recommended_schedule_cron: Optional[str] = None
-
-
-class MarketplaceCreator(BaseModel):
-    """A creator on the marketplace."""
-
-    name: str
-    username: str
-    description: str
-    avatar_url: str
-    num_agents: int
-    agent_rating: float
-    agent_runs: int
-    is_featured: bool = False
-
-
-class MarketplaceAgentsResponse(PaginatedResponse):
-    """Response for listing marketplace agents."""
-
-    agents: list[MarketplaceAgent]
-
-
-class MarketplaceCreatorsResponse(PaginatedResponse):
-    """Response for listing marketplace creators."""
-
-    creators: list[MarketplaceCreator]
-
-
-# Submission models
-class MarketplaceSubmission(BaseModel):
-    """A marketplace submission."""
-
-    agent_id: str
-    agent_version: int
-    name: str
-    sub_heading: str
-    slug: str
-    description: str
-    instructions: Optional[str] = None
-    image_urls: list[str] = Field(default_factory=list)
-    date_submitted: datetime
-    status: str = Field(description="One of: DRAFT, PENDING, APPROVED, REJECTED")
-    runs: int
-    rating: float
-    store_listing_version_id: Optional[str] = None
-    version: Optional[int] = None
-
-    # Review fields
-    review_comments: Optional[str] = None
-    reviewed_at: Optional[datetime] = None
-
-    # Additional optional fields
-    video_url: Optional[str] = None
-    categories: list[str] = Field(default_factory=list)
-
-
-class CreateSubmissionRequest(BaseModel):
-    """Request to create a marketplace submission."""
-
-    agent_id: str = Field(description="ID of the graph to submit")
-    agent_version: int = Field(description="Version of the graph to submit")
-    name: str = Field(description="Display name for the agent")
-    slug: str = Field(description="URL-friendly identifier")
-    description: str = Field(description="Full description")
-    sub_heading: str = Field(description="Short tagline")
-    image_urls: list[str] = Field(default_factory=list)
-    video_url: Optional[str] = None
-    categories: list[str] = Field(default_factory=list)
-
-
-class UpdateSubmissionRequest(BaseModel):
-    """Request to update a marketplace submission."""
-
-    name: Optional[str] = None
-    description: Optional[str] = None
-    sub_heading: Optional[str] = None
-    image_urls: Optional[list[str]] = None
-    video_url: Optional[str] = None
-    categories: Optional[list[str]] = None
-
-
-class SubmissionsListResponse(PaginatedResponse):
-    """Response for listing submissions."""
-
-    submissions: list[MarketplaceSubmission]
-
-
-# ============================================================================
-# Library Models
-# ============================================================================
-
-
-class LibraryAgent(BaseModel):
-    """An agent in the user's library."""
-
-    id: str
-    graph_id: str
-    graph_version: int
-    name: str
-    description: str
-    is_favorite: bool = False
-    can_access_graph: bool = False
-    is_latest_version: bool = False
-    image_url: Optional[str] = None
-    creator_name: str
-    input_schema: dict[str, Any] = Field(description="Input schema for the agent")
-    output_schema: dict[str, Any] = Field(description="Output schema for the agent")
-    created_at: datetime
-    updated_at: datetime
-
-
-class LibraryAgentsResponse(PaginatedResponse):
-    """Response for listing library agents."""
-
-    agents: list[LibraryAgent]
-
-
-class ExecuteAgentRequest(BaseModel):
-    """Request to execute an agent."""
-
-    inputs: dict[str, Any] = Field(
-        default_factory=dict, description="Input values for the agent"
-    )
-    credentials_inputs: dict[str, Any] = Field(
-        default_factory=dict, description="Credentials for the agent"
-    )
-
-
-# ============================================================================
-# Run Models
-# ============================================================================
-
-
-class Run(BaseModel):
-    """An execution run."""
-
-    id: str
-    graph_id: str
-    graph_version: int
-    status: str = Field(
-        description="One of: INCOMPLETE, QUEUED, RUNNING, COMPLETED, TERMINATED, FAILED, REVIEW"
-    )
-    started_at: datetime
-    ended_at: Optional[datetime] = None
-    inputs: Optional[dict[str, Any]] = None
-    cost: int = Field(default=0, description="Cost in credits")
-    duration: float = Field(default=0, description="Duration in seconds")
-    node_count: int = Field(default=0, description="Number of nodes executed")
-
-
-class RunDetails(Run):
-    """Detailed information about a run including node executions."""
-
-    outputs: Optional[dict[str, list[Any]]] = None
-    node_executions: list[dict[str, Any]] = Field(
-        default_factory=list, description="Individual node execution results"
-    )
-
-
-class RunsListResponse(PaginatedResponse):
-    """Response for listing runs."""
-
-    runs: list[Run]
-
-
-# ============================================================================
-# Run Review Models (Human-in-the-loop)
-# ============================================================================
-
-
-class PendingReview(BaseModel):
-    """A pending human-in-the-loop review."""
-
-    id: str  # node_exec_id
-    run_id: str
-    graph_id: str
-    graph_version: int
-    payload: Any = Field(description="Data to be reviewed")
-    instructions: Optional[str] = Field(
-        default=None, description="Instructions for the reviewer"
-    )
-    editable: bool = Field(
-        default=True, description="Whether the reviewer can edit the data"
-    )
-    status: str = Field(description="One of: WAITING, APPROVED, REJECTED")
-    created_at: datetime
-
-
-class PendingReviewsResponse(PaginatedResponse):
-    """Response for listing pending reviews."""
-
-    reviews: list[PendingReview]
-
-
-class ReviewDecision(BaseModel):
-    """Decision for a single review item."""
-
-    node_exec_id: str = Field(description="Node execution ID (review ID)")
-    approved: bool = Field(description="Whether to approve the data")
-    edited_payload: Optional[Any] = Field(
-        default=None, description="Modified payload data (if editing)"
-    )
-    message: Optional[str] = Field(
-        default=None, description="Optional message from reviewer", max_length=2000
-    )
-
-
-class SubmitReviewsRequest(BaseModel):
-    """Request to submit review responses for all pending reviews of an execution."""
-
-    reviews: list[ReviewDecision] = Field(
-        description="All review decisions for the execution"
-    )
-
-
-class SubmitReviewsResponse(BaseModel):
-    """Response after submitting reviews."""
-
-    run_id: str
-    approved_count: int = Field(description="Number of reviews approved")
-    rejected_count: int = Field(description="Number of reviews rejected")
-
-
-# ============================================================================
-# Credit Models
-# ============================================================================
-
-
-class CreditBalance(BaseModel):
-    """User's credit balance."""
-
-    balance: int = Field(description="Current credit balance")
-
-
-class CreditTransaction(BaseModel):
-    """A credit transaction."""
-
-    transaction_key: str
-    amount: int
-    transaction_type: str = Field(description="Transaction type")
-    transaction_time: datetime
-    running_balance: int
-    description: Optional[str] = None
-
-
-class CreditTransactionsResponse(PaginatedResponse):
-    """Response for listing credit transactions."""
-
-    transactions: list[CreditTransaction]
-
-
-# ============================================================================
-# Integration Models
-# ============================================================================
-
-
-class Credential(BaseModel):
-    """A user's credential for an integration."""
-
-    id: str
-    provider: str = Field(description="Integration provider name")
-    title: Optional[str] = Field(
-        default=None, description="User-assigned title for this credential"
-    )
-    scopes: list[str] = Field(default_factory=list, description="Granted scopes")
-
-
-class CredentialsListResponse(BaseModel):
-    """Response for listing credentials."""
-
-    credentials: list[Credential]
-
-
-class CredentialRequirement(BaseModel):
-    """A credential requirement for a graph or agent."""
-
-    provider: str = Field(description="Required provider name")
-    required_scopes: list[str] = Field(
-        default_factory=list, description="Required scopes"
-    )
-    matching_credentials: list[Credential] = Field(
-        default_factory=list,
-        description="User's credentials that match this requirement",
-    )
-
-
-class CredentialRequirementsResponse(BaseModel):
-    """Response for listing credential requirements."""
-
-    requirements: list[CredentialRequirement]
-
-
-# ============================================================================
-# File Models
-# ============================================================================
-
-
-class UploadFileResponse(BaseModel):
-    """Response after uploading a file."""
-
-    file_uri: str = Field(description="URI to reference the uploaded file")
-    file_name: str
-    size: int = Field(description="File size in bytes")
-    content_type: str
-    expires_in_hours: int

autogpt_platform/backend/backend/api/external/v2/routes.py

@@ -1,35 +0,0 @@
-"""
-V2 External API Routes
-
-This module defines the main v2 router that aggregates all v2 API endpoints.
-"""
-
-from fastapi import APIRouter
-
-from .blocks import blocks_router
-from .credits import credits_router
-from .files import files_router
-from .graphs import graphs_router
-from .integrations import integrations_router
-from .library import library_router
-from .marketplace import marketplace_router
-from .runs import runs_router
-from .schedules import graph_schedules_router, schedules_router
-
-v2_router = APIRouter()
-
-# Include all sub-routers
-v2_router.include_router(graphs_router, prefix="/graphs", tags=["graphs"])
-v2_router.include_router(graph_schedules_router, prefix="/graphs", tags=["schedules"])
-v2_router.include_router(schedules_router, prefix="/schedules", tags=["schedules"])
-v2_router.include_router(blocks_router, prefix="/blocks", tags=["blocks"])
-v2_router.include_router(
-    marketplace_router, prefix="/marketplace", tags=["marketplace"]
-)
-v2_router.include_router(library_router, prefix="/library", tags=["library"])
-v2_router.include_router(runs_router, prefix="/runs", tags=["runs"])
-v2_router.include_router(credits_router, prefix="/credits", tags=["credits"])
-v2_router.include_router(
-    integrations_router, prefix="/integrations", tags=["integrations"]
-)
-v2_router.include_router(files_router, prefix="/files", tags=["files"])

autogpt_platform/backend/backend/api/external/v2/runs.py

@@ -1,451 +0,0 @@
-"""
-V2 External API - Runs Endpoints
-
-Provides access to execution runs and human-in-the-loop reviews.
-"""
-
-import logging
-from datetime import datetime
-from typing import Any, Optional
-
-from fastapi import APIRouter, HTTPException, Path, Query, Security
-from prisma.enums import APIKeyPermission, ReviewStatus
-from pydantic import BaseModel, Field
-
-from backend.api.external.middleware import require_permission
-from backend.api.features.executions.review.model import (
-    PendingHumanReviewModel,
-    SafeJsonData,
-)
-from backend.data import execution as execution_db
-from backend.data import human_review as review_db
-from backend.data.auth.base import APIAuthorizationInfo
-from backend.executor import utils as execution_utils
-
-from .common import DEFAULT_PAGE_SIZE, MAX_PAGE_SIZE
-
-logger = logging.getLogger(__name__)
-
-runs_router = APIRouter()
-
-
-# ============================================================================
-# Models
-# ============================================================================
-
-
-class Run(BaseModel):
-    """An execution run."""
-
-    id: str
-    graph_id: str
-    graph_version: int
-    status: str = Field(
-        description="One of: INCOMPLETE, QUEUED, RUNNING, COMPLETED, TERMINATED, FAILED, REVIEW"
-    )
-    started_at: datetime
-    ended_at: Optional[datetime] = None
-    inputs: Optional[dict[str, Any]] = None
-    cost: int = Field(default=0, description="Cost in credits")
-    duration: float = Field(default=0, description="Duration in seconds")
-    node_count: int = Field(default=0, description="Number of nodes executed")
-
-
-class RunDetails(Run):
-    """Detailed information about a run including outputs and node executions."""
-
-    outputs: Optional[dict[str, list[Any]]] = None
-    node_executions: list[dict[str, Any]] = Field(
-        default_factory=list, description="Individual node execution results"
-    )
-
-
-class RunsListResponse(BaseModel):
-    """Response for listing runs."""
-
-    runs: list[Run]
-    total_count: int
-    page: int
-    page_size: int
-    total_pages: int
-
-
-class PendingReview(BaseModel):
-    """A pending human-in-the-loop review."""
-
-    id: str  # node_exec_id
-    run_id: str
-    graph_id: str
-    graph_version: int
-    payload: SafeJsonData = Field(description="Data to be reviewed")
-    instructions: Optional[str] = Field(
-        default=None, description="Instructions for the reviewer"
-    )
-    editable: bool = Field(
-        default=True, description="Whether the reviewer can edit the data"
-    )
-    status: str = Field(description="One of: WAITING, APPROVED, REJECTED")
-    created_at: datetime
-
-
-class PendingReviewsResponse(BaseModel):
-    """Response for listing pending reviews."""
-
-    reviews: list[PendingReview]
-    total_count: int
-    page: int
-    page_size: int
-    total_pages: int
-
-
-class ReviewDecision(BaseModel):
-    """Decision for a single review item."""
-
-    node_exec_id: str = Field(description="Node execution ID (review ID)")
-    approved: bool = Field(description="Whether to approve the data")
-    edited_payload: Optional[SafeJsonData] = Field(
-        default=None, description="Modified payload data (if editing)"
-    )
-    message: Optional[str] = Field(
-        default=None, description="Optional message from reviewer", max_length=2000
-    )
-
-
-class SubmitReviewsRequest(BaseModel):
-    """Request to submit review responses for all pending reviews of an execution."""
-
-    reviews: list[ReviewDecision] = Field(
-        description="All review decisions for the execution"
-    )
-
-
-class SubmitReviewsResponse(BaseModel):
-    """Response after submitting reviews."""
-
-    run_id: str
-    approved_count: int = Field(description="Number of reviews approved")
-    rejected_count: int = Field(description="Number of reviews rejected")
-
-
-# ============================================================================
-# Conversion Functions
-# ============================================================================
-
-
-def _convert_execution_to_run(exec: execution_db.GraphExecutionMeta) -> Run:
-    """Convert internal execution to v2 API Run model."""
-    return Run(
-        id=exec.id,
-        graph_id=exec.graph_id,
-        graph_version=exec.graph_version,
-        status=exec.status.value,
-        started_at=exec.started_at,
-        ended_at=exec.ended_at,
-        inputs=exec.inputs,
-        cost=exec.stats.cost if exec.stats else 0,
-        duration=exec.stats.duration if exec.stats else 0,
-        node_count=exec.stats.node_exec_count if exec.stats else 0,
-    )
-
-
-def _convert_execution_to_run_details(
-    exec: execution_db.GraphExecutionWithNodes,
-) -> RunDetails:
-    """Convert internal execution with nodes to v2 API RunDetails model."""
-    return RunDetails(
-        id=exec.id,
-        graph_id=exec.graph_id,
-        graph_version=exec.graph_version,
-        status=exec.status.value,
-        started_at=exec.started_at,
-        ended_at=exec.ended_at,
-        inputs=exec.inputs,
-        outputs=exec.outputs,
-        cost=exec.stats.cost if exec.stats else 0,
-        duration=exec.stats.duration if exec.stats else 0,
-        node_count=exec.stats.node_exec_count if exec.stats else 0,
-        node_executions=[
-            {
-                "node_id": node.node_id,
-                "status": node.status.value,
-                "input_data": node.input_data,
-                "output_data": node.output_data,
-                "started_at": node.start_time,
-                "ended_at": node.end_time,
-            }
-            for node in exec.node_executions
-        ],
-    )
-
-
-def _convert_pending_review(review: PendingHumanReviewModel) -> PendingReview:
-    """Convert internal PendingHumanReviewModel to v2 API PendingReview model."""
-    return PendingReview(
-        id=review.node_exec_id,
-        run_id=review.graph_exec_id,
-        graph_id=review.graph_id,
-        graph_version=review.graph_version,
-        payload=review.payload,
-        instructions=review.instructions,
-        editable=review.editable,
-        status=review.status.value,
-        created_at=review.created_at,
-    )
-
-
-# ============================================================================
-# Endpoints - Runs
-# ============================================================================
-
-
-@runs_router.get(
-    path="",
-    summary="List all runs",
-    response_model=RunsListResponse,
-)
-async def list_runs(
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_RUN)
-    ),
-    page: int = Query(default=1, ge=1, description="Page number (1-indexed)"),
-    page_size: int = Query(
-        default=DEFAULT_PAGE_SIZE,
-        ge=1,
-        le=MAX_PAGE_SIZE,
-        description=f"Items per page (max {MAX_PAGE_SIZE})",
-    ),
-) -> RunsListResponse:
-    """
-    List all execution runs for the authenticated user.
-
-    Returns runs across all agents, sorted by most recent first.
-    """
-    result = await execution_db.get_graph_executions_paginated(
-        user_id=auth.user_id,
-        page=page,
-        page_size=page_size,
-    )
-
-    return RunsListResponse(
-        runs=[_convert_execution_to_run(e) for e in result.executions],
-        total_count=result.pagination.total_items,
-        page=result.pagination.current_page,
-        page_size=result.pagination.page_size,
-        total_pages=result.pagination.total_pages,
-    )
-
-
-@runs_router.get(
-    path="/{run_id}",
-    summary="Get run details",
-    response_model=RunDetails,
-)
-async def get_run(
-    run_id: str = Path(description="Run ID"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_RUN)
-    ),
-) -> RunDetails:
-    """
-    Get detailed information about a specific run.
-
-    Includes outputs and individual node execution results.
-    """
-    result = await execution_db.get_graph_execution(
-        user_id=auth.user_id,
-        execution_id=run_id,
-        include_node_executions=True,
-    )
-
-    if not result:
-        raise HTTPException(status_code=404, detail=f"Run #{run_id} not found")
-
-    return _convert_execution_to_run_details(result)
-
-
-@runs_router.post(
-    path="/{run_id}/stop",
-    summary="Stop a run",
-)
-async def stop_run(
-    run_id: str = Path(description="Run ID"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.WRITE_RUN)
-    ),
-) -> Run:
-    """
-    Stop a running execution.
-
-    Only runs in QUEUED or RUNNING status can be stopped.
-    """
-    # Verify the run exists and belongs to the user
-    exec = await execution_db.get_graph_execution(
-        user_id=auth.user_id,
-        execution_id=run_id,
-    )
-    if not exec:
-        raise HTTPException(status_code=404, detail=f"Run #{run_id} not found")
-
-    # Stop the execution
-    await execution_utils.stop_graph_execution(
-        graph_exec_id=run_id,
-        user_id=auth.user_id,
-    )
-
-    # Fetch updated execution
-    updated_exec = await execution_db.get_graph_execution(
-        user_id=auth.user_id,
-        execution_id=run_id,
-    )
-
-    if not updated_exec:
-        raise HTTPException(status_code=404, detail=f"Run #{run_id} not found")
-
-    return _convert_execution_to_run(updated_exec)
-
-
-@runs_router.delete(
-    path="/{run_id}",
-    summary="Delete a run",
-)
-async def delete_run(
-    run_id: str = Path(description="Run ID"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.WRITE_RUN)
-    ),
-) -> None:
-    """
-    Delete an execution run.
-
-    This marks the run as deleted. The data may still be retained for
-    some time for recovery purposes.
-    """
-    await execution_db.delete_graph_execution(
-        graph_exec_id=run_id,
-        user_id=auth.user_id,
-    )
-
-
-# ============================================================================
-# Endpoints - Reviews (Human-in-the-loop)
-# ============================================================================
-
-
-@runs_router.get(
-    path="/reviews",
-    summary="List all pending reviews",
-    response_model=PendingReviewsResponse,
-)
-async def list_pending_reviews(
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_RUN_REVIEW)
-    ),
-    page: int = Query(default=1, ge=1, description="Page number (1-indexed)"),
-    page_size: int = Query(
-        default=DEFAULT_PAGE_SIZE,
-        ge=1,
-        le=MAX_PAGE_SIZE,
-        description=f"Items per page (max {MAX_PAGE_SIZE})",
-    ),
-) -> PendingReviewsResponse:
-    """
-    List all pending human-in-the-loop reviews.
-
-    These are blocks that require human approval or input before the
-    agent can continue execution.
-    """
-    reviews = await review_db.get_pending_reviews_for_user(
-        user_id=auth.user_id,
-        page=page,
-        page_size=page_size,
-    )
-
-    # Note: get_pending_reviews_for_user returns list directly, not a paginated result
-    # We compute pagination info based on results
-    total_count = len(reviews)
-    total_pages = max(1, (total_count + page_size - 1) // page_size)
-
-    return PendingReviewsResponse(
-        reviews=[_convert_pending_review(r) for r in reviews],
-        total_count=total_count,
-        page=page,
-        page_size=page_size,
-        total_pages=total_pages,
-    )
-
-
-@runs_router.get(
-    path="/{run_id}/reviews",
-    summary="List reviews for a run",
-    response_model=list[PendingReview],
-)
-async def list_run_reviews(
-    run_id: str = Path(description="Run ID"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_RUN_REVIEW)
-    ),
-) -> list[PendingReview]:
-    """
-    List all human-in-the-loop reviews for a specific run.
-    """
-    reviews = await review_db.get_pending_reviews_for_execution(
-        graph_exec_id=run_id,
-        user_id=auth.user_id,
-    )
-
-    return [_convert_pending_review(r) for r in reviews]
-
-
-@runs_router.post(
-    path="/{run_id}/reviews",
-    summary="Submit review responses for a run",
-    response_model=SubmitReviewsResponse,
-)
-async def submit_reviews(
-    request: SubmitReviewsRequest,
-    run_id: str = Path(description="Run ID"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.WRITE_RUN_REVIEW)
-    ),
-) -> SubmitReviewsResponse:
-    """
-    Submit responses to all pending human-in-the-loop reviews for a run.
-
-    All pending reviews for the execution must be included. Approving
-    a review will allow the agent to continue; rejecting will terminate
-    execution at that point.
-    """
-    # Build review decisions dict for process_all_reviews_for_execution
-    review_decisions: dict[
-        str, tuple[ReviewStatus, SafeJsonData | None, str | None]
-    ] = {}
-
-    for decision in request.reviews:
-        status = ReviewStatus.APPROVED if decision.approved else ReviewStatus.REJECTED
-        review_decisions[decision.node_exec_id] = (
-            status,
-            decision.edited_payload,
-            decision.message,
-        )
-
-    try:
-        results = await review_db.process_all_reviews_for_execution(
-            user_id=auth.user_id,
-            review_decisions=review_decisions,
-        )
-
-        approved_count = sum(
-            1 for r in results.values() if r.status == ReviewStatus.APPROVED
-        )
-        rejected_count = sum(
-            1 for r in results.values() if r.status == ReviewStatus.REJECTED
-        )
-
-        return SubmitReviewsResponse(
-            run_id=run_id,
-            approved_count=approved_count,
-            rejected_count=rejected_count,
-        )
-
-    except ValueError as e:
-        raise HTTPException(status_code=400, detail=str(e))

autogpt_platform/backend/backend/api/external/v2/schedules.py

@@ -1,250 +0,0 @@
-"""
-V2 External API - Schedules Endpoints
-
-Provides endpoints for managing execution schedules.
-"""
-
-import logging
-from datetime import datetime
-from typing import Any, Optional
-
-from fastapi import APIRouter, HTTPException, Path, Query, Security
-from prisma.enums import APIKeyPermission
-from pydantic import BaseModel, Field
-
-from backend.api.external.middleware import require_permission
-from backend.data import graph as graph_db
-from backend.data.auth.base import APIAuthorizationInfo
-from backend.data.user import get_user_by_id
-from backend.executor import scheduler
-from backend.util.clients import get_scheduler_client
-from backend.util.timezone_utils import get_user_timezone_or_utc
-
-from .common import DEFAULT_PAGE_SIZE, MAX_PAGE_SIZE
-
-logger = logging.getLogger(__name__)
-
-schedules_router = APIRouter()
-
-
-# ============================================================================
-# Request/Response Models
-# ============================================================================
-
-
-class Schedule(BaseModel):
-    """An execution schedule for a graph."""
-
-    id: str
-    name: str
-    graph_id: str
-    graph_version: int
-    cron: str = Field(description="Cron expression for the schedule")
-    input_data: dict[str, Any] = Field(
-        default_factory=dict, description="Input data for scheduled executions"
-    )
-    next_run_time: Optional[datetime] = Field(
-        default=None, description="Next scheduled run time"
-    )
-    is_enabled: bool = Field(default=True, description="Whether schedule is enabled")
-
-
-class SchedulesListResponse(BaseModel):
-    """Response for listing schedules."""
-
-    schedules: list[Schedule]
-    total_count: int
-    page: int
-    page_size: int
-    total_pages: int
-
-
-class CreateScheduleRequest(BaseModel):
-    """Request to create a schedule."""
-
-    name: str = Field(description="Display name for the schedule")
-    cron: str = Field(description="Cron expression (e.g., '0 9 * * *' for 9am daily)")
-    input_data: dict[str, Any] = Field(
-        default_factory=dict, description="Input data for scheduled executions"
-    )
-    credentials_inputs: dict[str, Any] = Field(
-        default_factory=dict, description="Credentials for the schedule"
-    )
-    graph_version: Optional[int] = Field(
-        default=None, description="Graph version (default: active version)"
-    )
-    timezone: Optional[str] = Field(
-        default=None,
-        description=(
-            "Timezone for schedule (e.g., 'America/New_York'). "
-            "Defaults to user's timezone."
-        ),
-    )
-
-
-def _convert_schedule(job: scheduler.GraphExecutionJobInfo) -> Schedule:
-    """Convert internal schedule job info to v2 API model."""
-    # Parse the ISO format string to datetime
-    next_run = datetime.fromisoformat(job.next_run_time) if job.next_run_time else None
-
-    return Schedule(
-        id=job.id,
-        name=job.name or "",
-        graph_id=job.graph_id,
-        graph_version=job.graph_version,
-        cron=job.cron,
-        input_data=job.input_data,
-        next_run_time=next_run,
-        is_enabled=True,  # All returned schedules are enabled
-    )
-
-
-# ============================================================================
-# Endpoints
-# ============================================================================
-
-
-@schedules_router.get(
-    path="",
-    summary="List all user schedules",
-    response_model=SchedulesListResponse,
-)
-async def list_all_schedules(
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_SCHEDULE)
-    ),
-    page: int = Query(default=1, ge=1, description="Page number (1-indexed)"),
-    page_size: int = Query(
-        default=DEFAULT_PAGE_SIZE,
-        ge=1,
-        le=MAX_PAGE_SIZE,
-        description=f"Items per page (max {MAX_PAGE_SIZE})",
-    ),
-) -> SchedulesListResponse:
-    """
-    List all schedules for the authenticated user across all graphs.
-    """
-    schedules = await get_scheduler_client().get_execution_schedules(
-        user_id=auth.user_id
-    )
-    converted = [_convert_schedule(s) for s in schedules]
-
-    # Manual pagination (scheduler doesn't support pagination natively)
-    total_count = len(converted)
-    total_pages = (total_count + page_size - 1) // page_size if total_count > 0 else 1
-    start = (page - 1) * page_size
-    end = start + page_size
-    paginated = converted[start:end]
-
-    return SchedulesListResponse(
-        schedules=paginated,
-        total_count=total_count,
-        page=page,
-        page_size=page_size,
-        total_pages=total_pages,
-    )
-
-
-@schedules_router.delete(
-    path="/{schedule_id}",
-    summary="Delete a schedule",
-)
-async def delete_schedule(
-    schedule_id: str = Path(description="Schedule ID to delete"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.WRITE_SCHEDULE)
-    ),
-) -> None:
-    """
-    Delete an execution schedule.
-    """
-    try:
-        await get_scheduler_client().delete_schedule(
-            schedule_id=schedule_id,
-            user_id=auth.user_id,
-        )
-    except Exception as e:
-        if "not found" in str(e).lower():
-            raise HTTPException(
-                status_code=404, detail=f"Schedule #{schedule_id} not found"
-            )
-        raise
-
-
-# ============================================================================
-# Graph-specific Schedule Endpoints (nested under /graphs)
-# These are included in the graphs router via include_router
-# ============================================================================
-
-graph_schedules_router = APIRouter()
-
-
-@graph_schedules_router.get(
-    path="/{graph_id}/schedules",
-    summary="List schedules for a graph",
-    response_model=list[Schedule],
-)
-async def list_graph_schedules(
-    graph_id: str = Path(description="Graph ID"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.READ_SCHEDULE)
-    ),
-) -> list[Schedule]:
-    """
-    List all schedules for a specific graph.
-    """
-    schedules = await get_scheduler_client().get_execution_schedules(
-        user_id=auth.user_id,
-        graph_id=graph_id,
-    )
-    return [_convert_schedule(s) for s in schedules]
-
-
-@graph_schedules_router.post(
-    path="/{graph_id}/schedules",
-    summary="Create a schedule for a graph",
-    response_model=Schedule,
-)
-async def create_graph_schedule(
-    request: CreateScheduleRequest,
-    graph_id: str = Path(description="Graph ID"),
-    auth: APIAuthorizationInfo = Security(
-        require_permission(APIKeyPermission.WRITE_SCHEDULE)
-    ),
-) -> Schedule:
-    """
-    Create a new execution schedule for a graph.
-
-    The schedule will execute the graph at times matching the cron expression,
-    using the provided input data.
-    """
-    graph = await graph_db.get_graph(
-        graph_id=graph_id,
-        version=request.graph_version,
-        user_id=auth.user_id,
-    )
-    if not graph:
-        raise HTTPException(
-            status_code=404,
-            detail=f"Graph #{graph_id} v{request.graph_version} not found.",
-        )
-
-    # Determine timezone
-    if request.timezone:
-        user_timezone = request.timezone
-    else:
-        user = await get_user_by_id(auth.user_id)
-        user_timezone = get_user_timezone_or_utc(user.timezone if user else None)
-
-    result = await get_scheduler_client().add_execution_schedule(
-        user_id=auth.user_id,
-        graph_id=graph_id,
-        graph_version=graph.version,
-        name=request.name,
-        cron=request.cron,
-        input_data=request.input_data,
-        input_credentials=request.credentials_inputs,
-        user_timezone=user_timezone,
-    )
-
-    return _convert_schedule(result)

autogpt_platform/backend/backend/api/features/admin/store_admin_routes.py

@@ -9,6 +9,7 @@ import prisma.enums
 
 import backend.api.features.store.cache as store_cache
 import backend.api.features.store.db as store_db
+import backend.api.features.store.embeddings as store_embeddings
 import backend.api.features.store.model as store_model
 import backend.util.json
 
@@ -150,3 +151,54 @@ async def admin_download_agent_file(
         return fastapi.responses.FileResponse(
             tmp_file.name, filename=file_name, media_type="application/json"
         )
+
+
+@router.get(
+    "/embeddings/stats",
+    summary="Get Embedding Statistics",
+)
+async def get_embedding_stats() -> dict[str, typing.Any]:
+    """
+    Get statistics about embedding coverage for store listings.
+
+    Returns counts of total approved listings, listings with embeddings,
+    listings without embeddings, and coverage percentage.
+    """
+    try:
+        stats = await store_embeddings.get_embedding_stats()
+        return stats
+    except Exception as e:
+        logger.exception("Error getting embedding stats: %s", e)
+        raise fastapi.HTTPException(
+            status_code=500,
+            detail="An error occurred while retrieving embedding stats",
+        )
+
+
+@router.post(
+    "/embeddings/backfill",
+    summary="Backfill Missing Embeddings",
+)
+async def backfill_embeddings(
+    batch_size: int = 10,
+) -> dict[str, typing.Any]:
+    """
+    Trigger backfill of embeddings for approved listings that don't have them.
+
+    Args:
+        batch_size: Number of embeddings to generate in one call (default 10)
+
+    Returns:
+        Dict with processed count, success count, failure count, and message
+    """
+    try:
+        result = await store_embeddings.backfill_missing_embeddings(
+            batch_size=batch_size
+        )
+        return result
+    except Exception as e:
+        logger.exception("Error backfilling embeddings: %s", e)
+        raise fastapi.HTTPException(
+            status_code=500,
+            detail="An error occurred while backfilling embeddings",
+        )

autogpt_platform/backend/backend/api/features/chat/config.py

@@ -1,6 +1,7 @@
 """Configuration management for chat system."""
 
 import os
+from pathlib import Path
 
 from pydantic import Field, field_validator
 from pydantic_settings import BaseSettings
@@ -26,6 +27,12 @@ class ChatConfig(BaseSettings):
     # Session TTL Configuration - 12 hours
     session_ttl: int = Field(default=43200, description="Session TTL in seconds")
 
+    # System Prompt Configuration
+    system_prompt_path: str = Field(
+        default="prompts/chat_system.md",
+        description="Path to system prompt file relative to chat module",
+    )
+
     # Streaming Configuration
     max_context_messages: int = Field(
         default=50, ge=1, le=200, description="Maximum context messages"
@@ -82,6 +89,73 @@ class ChatConfig(BaseSettings):
         "onboarding": "prompts/onboarding_system.md",
     }
 
+    def get_system_prompt_for_type(
+        self, prompt_type: str = "default", **template_vars
+    ) -> str:
+        """Load and render a system prompt by type.
+
+        Args:
+            prompt_type: The type of prompt to load ("default" or "onboarding")
+            **template_vars: Variables to substitute in the template
+
+        Returns:
+            Rendered system prompt string
+        """
+        prompt_path_str = self.PROMPT_PATHS.get(
+            prompt_type, self.PROMPT_PATHS["default"]
+        )
+        return self._load_prompt_from_path(prompt_path_str, **template_vars)
+
+    def get_system_prompt(self, **template_vars) -> str:
+        """Load and render the default system prompt from file.
+
+        Args:
+            **template_vars: Variables to substitute in the template
+
+        Returns:
+            Rendered system prompt string
+
+        """
+        return self._load_prompt_from_path(self.system_prompt_path, **template_vars)
+
+    def _load_prompt_from_path(self, prompt_path_str: str, **template_vars) -> str:
+        """Load and render a system prompt from a given path.
+
+        Args:
+            prompt_path_str: Path to the prompt file relative to chat module
+            **template_vars: Variables to substitute in the template
+
+        Returns:
+            Rendered system prompt string
+        """
+        # Get the path relative to this module
+        module_dir = Path(__file__).parent
+        prompt_path = module_dir / prompt_path_str
+
+        # Check for .j2 extension first (Jinja2 template)
+        j2_path = Path(str(prompt_path) + ".j2")
+        if j2_path.exists():
+            try:
+                from jinja2 import Template
+
+                template = Template(j2_path.read_text())
+                return template.render(**template_vars)
+            except ImportError:
+                # Jinja2 not installed, fall back to reading as plain text
+                return j2_path.read_text()
+
+        # Check for markdown file
+        if prompt_path.exists():
+            content = prompt_path.read_text()
+
+            # Simple variable substitution if Jinja2 is not available
+            for key, value in template_vars.items():
+                placeholder = f"{{{key}}}"
+                content = content.replace(placeholder, str(value))
+
+            return content
+        raise FileNotFoundError(f"System prompt file not found: {prompt_path}")
+
     class Config:
         """Pydantic config."""
 

autogpt_platform/backend/backend/api/features/chat/db.py

@@ -1,20 +1,13 @@
 """Database operations for chat sessions."""
 
-import asyncio
 import logging
 from datetime import UTC, datetime
-from typing import Any, cast
+from typing import Any
 
 from prisma.models import ChatMessage as PrismaChatMessage
 from prisma.models import ChatSession as PrismaChatSession
-from prisma.types import (
-    ChatMessageCreateInput,
-    ChatSessionCreateInput,
-    ChatSessionUpdateInput,
-    ChatSessionWhereInput,
-)
+from prisma.types import ChatSessionUpdateInput
 
-from backend.data.db import transaction
 from backend.util.json import SafeJson
 
 logger = logging.getLogger(__name__)
@@ -27,24 +20,23 @@ async def get_chat_session(session_id: str) -> PrismaChatSession | None:
         include={"Messages": True},
     )
     if session and session.Messages:
-        # Sort messages by sequence in Python - Prisma Python client doesn't support
-        # order_by in include clauses (unlike Prisma JS), so we sort after fetching
+        # Sort messages by sequence in Python since Prisma doesn't support order_by in include
         session.Messages.sort(key=lambda m: m.sequence)
     return session
 
 
 async def create_chat_session(
     session_id: str,
-    user_id: str,
+    user_id: str | None,
 ) -> PrismaChatSession:
     """Create a new chat session in the database."""
-    data = ChatSessionCreateInput(
-        id=session_id,
-        userId=user_id,
-        credentials=SafeJson({}),
-        successfulAgentRuns=SafeJson({}),
-        successfulAgentSchedules=SafeJson({}),
-    )
+    data = {
+        "id": session_id,
+        "userId": user_id,
+        "credentials": SafeJson({}),
+        "successfulAgentRuns": SafeJson({}),
+        "successfulAgentSchedules": SafeJson({}),
+    }
     return await PrismaChatSession.prisma().create(
         data=data,
         include={"Messages": True},
@@ -82,7 +74,6 @@ async def update_chat_session(
         include={"Messages": True},
     )
     if session and session.Messages:
-        # Sort in Python - Prisma Python doesn't support order_by in include clauses
         session.Messages.sort(key=lambda m: m.sequence)
     return session
 
@@ -99,16 +90,12 @@ async def add_chat_message(
     function_call: dict[str, Any] | None = None,
 ) -> PrismaChatMessage:
     """Add a message to a chat session."""
-    # Build input dict dynamically rather than using ChatMessageCreateInput directly
-    # because Prisma's TypedDict validation rejects optional fields set to None.
-    # We only include fields that have values, then cast at the end.
     data: dict[str, Any] = {
         "Session": {"connect": {"id": session_id}},
         "role": role,
         "sequence": sequence,
     }
 
-    # Add optional string fields
     if content is not None:
         data["content"] = content
     if name is not None:
@@ -117,22 +104,18 @@ async def add_chat_message(
         data["toolCallId"] = tool_call_id
     if refusal is not None:
         data["refusal"] = refusal
-
-    # Add optional JSON fields only when they have values
     if tool_calls is not None:
         data["toolCalls"] = SafeJson(tool_calls)
     if function_call is not None:
         data["functionCall"] = SafeJson(function_call)
 
-    # Run message create and session timestamp update in parallel for lower latency
-    _, message = await asyncio.gather(
-        PrismaChatSession.prisma().update(
-            where={"id": session_id},
-            data={"updatedAt": datetime.now(UTC)},
-        ),
-        PrismaChatMessage.prisma().create(data=cast(ChatMessageCreateInput, data)),
+    # Update session's updatedAt timestamp
+    await PrismaChatSession.prisma().update(
+        where={"id": session_id},
+        data={"updatedAt": datetime.now(UTC)},
     )
-    return message
+
+    return await PrismaChatMessage.prisma().create(data=data)
 
 
 async def add_chat_messages_batch(
@@ -140,55 +123,39 @@ async def add_chat_messages_batch(
     messages: list[dict[str, Any]],
     start_sequence: int,
 ) -> list[PrismaChatMessage]:
-    """Add multiple messages to a chat session in a batch.
-
-    Uses a transaction for atomicity - if any message creation fails,
-    the entire batch is rolled back.
-    """
+    """Add multiple messages to a chat session in a batch."""
     if not messages:
         return []
 
     created_messages = []
+    for i, msg in enumerate(messages):
+        data: dict[str, Any] = {
+            "Session": {"connect": {"id": session_id}},
+            "role": msg["role"],
+            "sequence": start_sequence + i,
+        }
 
-    async with transaction() as tx:
-        for i, msg in enumerate(messages):
-            # Build input dict dynamically rather than using ChatMessageCreateInput
-            # directly because Prisma's TypedDict validation rejects optional fields
-            # set to None. We only include fields that have values, then cast.
-            data: dict[str, Any] = {
-                "Session": {"connect": {"id": session_id}},
-                "role": msg["role"],
-                "sequence": start_sequence + i,
-            }
+        if msg.get("content") is not None:
+            data["content"] = msg["content"]
+        if msg.get("name") is not None:
+            data["name"] = msg["name"]
+        if msg.get("tool_call_id") is not None:
+            data["toolCallId"] = msg["tool_call_id"]
+        if msg.get("refusal") is not None:
+            data["refusal"] = msg["refusal"]
+        if msg.get("tool_calls") is not None:
+            data["toolCalls"] = SafeJson(msg["tool_calls"])
+        if msg.get("function_call") is not None:
+            data["functionCall"] = SafeJson(msg["function_call"])
 
-            # Add optional string fields
-            if msg.get("content") is not None:
-                data["content"] = msg["content"]
-            if msg.get("name") is not None:
-                data["name"] = msg["name"]
-            if msg.get("tool_call_id") is not None:
-                data["toolCallId"] = msg["tool_call_id"]
-            if msg.get("refusal") is not None:
-                data["refusal"] = msg["refusal"]
+        created = await PrismaChatMessage.prisma().create(data=data)
+        created_messages.append(created)
 
-            # Add optional JSON fields only when they have values
-            if msg.get("tool_calls") is not None:
-                data["toolCalls"] = SafeJson(msg["tool_calls"])
-            if msg.get("function_call") is not None:
-                data["functionCall"] = SafeJson(msg["function_call"])
-
-            created = await PrismaChatMessage.prisma(tx).create(
-                data=cast(ChatMessageCreateInput, data)
-            )
-            created_messages.append(created)
-
-        # Update session's updatedAt timestamp within the same transaction.
-        # Note: Token usage (total_prompt_tokens, total_completion_tokens) is updated
-        # separately via update_chat_session() after streaming completes.
-        await PrismaChatSession.prisma(tx).update(
-            where={"id": session_id},
-            data={"updatedAt": datetime.now(UTC)},
-        )
+    # Update session's updatedAt timestamp
+    await PrismaChatSession.prisma().update(
+        where={"id": session_id},
+        data={"updatedAt": datetime.now(UTC)},
+    )
 
     return created_messages
 
@@ -212,31 +179,10 @@ async def get_user_session_count(user_id: str) -> int:
     return await PrismaChatSession.prisma().count(where={"userId": user_id})
 
 
-async def delete_chat_session(session_id: str, user_id: str | None = None) -> bool:
-    """Delete a chat session and all its messages.
-
-    Args:
-        session_id: The session ID to delete.
-        user_id: If provided, validates that the session belongs to this user
-            before deletion. This prevents unauthorized deletion of other
-            users' sessions.
-
-    Returns:
-        True if deleted successfully, False otherwise.
-    """
+async def delete_chat_session(session_id: str) -> bool:
+    """Delete a chat session and all its messages."""
     try:
-        # Build typed where clause with optional user_id validation
-        where_clause: ChatSessionWhereInput = {"id": session_id}
-        if user_id is not None:
-            where_clause["userId"] = user_id
-
-        result = await PrismaChatSession.prisma().delete_many(where=where_clause)
-        if result == 0:
-            logger.warning(
-                f"No session deleted for {session_id} "
-                f"(user_id validation: {user_id is not None})"
-            )
-            return False
+        await PrismaChatSession.prisma().delete(where={"id": session_id})
         return True
     except Exception as e:
         logger.error(f"Failed to delete chat session {session_id}: {e}")

autogpt_platform/backend/backend/api/features/chat/model.py

@@ -1,9 +1,6 @@
-import asyncio
 import logging
 import uuid
 from datetime import UTC, datetime
-from typing import Any
-from weakref import WeakValueDictionary
 
 from openai.types.chat import (
     ChatCompletionAssistantMessageParam,
@@ -25,7 +22,7 @@ from pydantic import BaseModel
 
 from backend.data.redis_client import get_redis_async
 from backend.util import json
-from backend.util.exceptions import DatabaseError, RedisError
+from backend.util.exceptions import RedisError
 
 from . import db as chat_db
 from .config import ChatConfig
@@ -34,48 +31,6 @@ logger = logging.getLogger(__name__)
 config = ChatConfig()
 
 
-def _parse_json_field(value: str | dict | list | None, default: Any = None) -> Any:
-    """Parse a JSON field that may be stored as string or already parsed."""
-    if value is None:
-        return default
-    if isinstance(value, str):
-        return json.loads(value)
-    return value
-
-
-# Redis cache key prefix for chat sessions
-CHAT_SESSION_CACHE_PREFIX = "chat:session:"
-
-
-def _get_session_cache_key(session_id: str) -> str:
-    """Get the Redis cache key for a chat session."""
-    return f"{CHAT_SESSION_CACHE_PREFIX}{session_id}"
-
-
-# Session-level locks to prevent race conditions during concurrent upserts.
-# Uses WeakValueDictionary to automatically garbage collect locks when no longer referenced,
-# preventing unbounded memory growth while maintaining lock semantics for active sessions.
-# Invalidation: Locks are auto-removed by GC when no coroutine holds a reference (after
-# async with lock: completes). Explicit cleanup also occurs in delete_chat_session().
-_session_locks: WeakValueDictionary[str, asyncio.Lock] = WeakValueDictionary()
-_session_locks_mutex = asyncio.Lock()
-
-
-async def _get_session_lock(session_id: str) -> asyncio.Lock:
-    """Get or create a lock for a specific session to prevent concurrent upserts.
-
-    Uses WeakValueDictionary for automatic cleanup: locks are garbage collected
-    when no coroutine holds a reference to them, preventing memory leaks from
-    unbounded growth of session locks.
-    """
-    async with _session_locks_mutex:
-        lock = _session_locks.get(session_id)
-        if lock is None:
-            lock = asyncio.Lock()
-            _session_locks[session_id] = lock
-        return lock
-
-
 class ChatMessage(BaseModel):
     role: str
     content: str | None = None
@@ -94,7 +49,7 @@ class Usage(BaseModel):
 
 class ChatSession(BaseModel):
     session_id: str
-    user_id: str
+    user_id: str | None
     title: str | None = None
     messages: list[ChatMessage]
     usage: list[Usage]
@@ -105,7 +60,7 @@ class ChatSession(BaseModel):
     successful_agent_schedules: dict[str, int] = {}
 
     @staticmethod
-    def new(user_id: str) -> "ChatSession":
+    def new(user_id: str | None) -> "ChatSession":
         return ChatSession(
             session_id=str(uuid.uuid4()),
             user_id=user_id,
@@ -118,7 +73,7 @@ class ChatSession(BaseModel):
         )
 
     @staticmethod
-    def from_db(
+    def from_prisma(
         prisma_session: PrismaChatSession,
         prisma_messages: list[PrismaChatMessage] | None = None,
     ) -> "ChatSession":
@@ -126,6 +81,22 @@ class ChatSession(BaseModel):
         messages = []
         if prisma_messages:
             for msg in prisma_messages:
+                tool_calls = None
+                if msg.toolCalls:
+                    tool_calls = (
+                        json.loads(msg.toolCalls)
+                        if isinstance(msg.toolCalls, str)
+                        else msg.toolCalls
+                    )
+
+                function_call = None
+                if msg.functionCall:
+                    function_call = (
+                        json.loads(msg.functionCall)
+                        if isinstance(msg.functionCall, str)
+                        else msg.functionCall
+                    )
+
                 messages.append(
                     ChatMessage(
                         role=msg.role,
@@ -133,18 +104,26 @@ class ChatSession(BaseModel):
                         name=msg.name,
                         tool_call_id=msg.toolCallId,
                         refusal=msg.refusal,
-                        tool_calls=_parse_json_field(msg.toolCalls),
-                        function_call=_parse_json_field(msg.functionCall),
+                        tool_calls=tool_calls,
+                        function_call=function_call,
                     )
                 )
 
         # Parse JSON fields from Prisma
-        credentials = _parse_json_field(prisma_session.credentials, default={})
-        successful_agent_runs = _parse_json_field(
-            prisma_session.successfulAgentRuns, default={}
+        credentials = (
+            json.loads(prisma_session.credentials)
+            if isinstance(prisma_session.credentials, str)
+            else prisma_session.credentials or {}
         )
-        successful_agent_schedules = _parse_json_field(
-            prisma_session.successfulAgentSchedules, default={}
+        successful_agent_runs = (
+            json.loads(prisma_session.successfulAgentRuns)
+            if isinstance(prisma_session.successfulAgentRuns, str)
+            else prisma_session.successfulAgentRuns or {}
+        )
+        successful_agent_schedules = (
+            json.loads(prisma_session.successfulAgentSchedules)
+            if isinstance(prisma_session.successfulAgentSchedules, str)
+            else prisma_session.successfulAgentSchedules or {}
         )
 
         # Calculate usage from token counts
@@ -263,7 +242,7 @@ class ChatSession(BaseModel):
 
 async def _get_session_from_cache(session_id: str) -> ChatSession | None:
     """Get a chat session from Redis cache."""
-    redis_key = _get_session_cache_key(session_id)
+    redis_key = f"chat:session:{session_id}"
     async_redis = await get_redis_async()
     raw_session: bytes | None = await async_redis.get(redis_key)
 
@@ -285,7 +264,7 @@ async def _get_session_from_cache(session_id: str) -> ChatSession | None:
 
 async def _cache_session(session: ChatSession) -> None:
     """Cache a chat session in Redis."""
-    redis_key = _get_session_cache_key(session.session_id)
+    redis_key = f"chat:session:{session.session_id}"
     async_redis = await get_redis_async()
     await async_redis.setex(redis_key, config.session_ttl, session.model_dump_json())
 
@@ -304,7 +283,7 @@ async def _get_session_from_db(session_id: str) -> ChatSession | None:
         f"roles={[m.role for m in messages] if messages else []}"
     )
 
-    return ChatSession.from_db(prisma_session, messages)
+    return ChatSession.from_prisma(prisma_session, messages)
 
 
 async def _save_session_to_db(
@@ -366,24 +345,19 @@ async def _save_session_to_db(
 
 async def get_chat_session(
     session_id: str,
-    user_id: str | None = None,
+    user_id: str | None,
 ) -> ChatSession | None:
     """Get a chat session by ID.
 
     Checks Redis cache first, falls back to database if not found.
     Caches database results back to Redis.
-
-    Args:
-        session_id: The session ID to fetch.
-        user_id: If provided, validates that the session belongs to this user.
-            If None, ownership is not validated (admin/system access).
     """
     # Try cache first
     try:
         session = await _get_session_from_cache(session_id)
         if session:
-            # Verify user ownership if user_id was provided for validation
-            if user_id is not None and session.user_id != user_id:
+            # Verify user ownership
+            if session.user_id is not None and session.user_id != user_id:
                 logger.warning(
                     f"Session {session_id} user id mismatch: {session.user_id} != {user_id}"
                 )
@@ -402,8 +376,8 @@ async def get_chat_session(
         logger.warning(f"Session {session_id} not found in cache or database")
         return None
 
-    # Verify user ownership if user_id was provided for validation
-    if user_id is not None and session.user_id != user_id:
+    # Verify user ownership
+    if session.user_id is not None and session.user_id != user_id:
         logger.warning(
             f"Session {session_id} user id mismatch: {session.user_id} != {user_id}"
         )
@@ -422,88 +396,49 @@ async def get_chat_session(
 async def upsert_chat_session(
     session: ChatSession,
 ) -> ChatSession:
-    """Update a chat session in both cache and database.
+    """Update a chat session in both cache and database."""
+    # Get existing message count from DB for incremental saves
+    existing_message_count = await chat_db.get_chat_session_message_count(
+        session.session_id
+    )
 
-    Uses session-level locking to prevent race conditions when concurrent
-    operations (e.g., background title update and main stream handler)
-    attempt to upsert the same session simultaneously.
+    # Save to database
+    try:
+        await _save_session_to_db(session, existing_message_count)
+    except Exception as e:
+        logger.error(f"Failed to save session {session.session_id} to database: {e}")
+        # Continue to cache even if DB fails
 
-    Raises:
-        DatabaseError: If the database write fails. The cache is still updated
-            as a best-effort optimization, but the error is propagated to ensure
-            callers are aware of the persistence failure.
-        RedisError: If the cache write fails (after successful DB write).
-    """
-    # Acquire session-specific lock to prevent concurrent upserts
-    lock = await _get_session_lock(session.session_id)
+    # Save to cache
+    try:
+        await _cache_session(session)
+    except Exception as e:
+        raise RedisError(
+            f"Failed to persist chat session {session.session_id} to Redis: {e}"
+        ) from e
 
-    async with lock:
-        # Get existing message count from DB for incremental saves
-        existing_message_count = await chat_db.get_chat_session_message_count(
-            session.session_id
-        )
-
-        db_error: Exception | None = None
-
-        # Save to database (primary storage)
-        try:
-            await _save_session_to_db(session, existing_message_count)
-        except Exception as e:
-            logger.error(
-                f"Failed to save session {session.session_id} to database: {e}"
-            )
-            db_error = e
-
-        # Save to cache (best-effort, even if DB failed)
-        try:
-            await _cache_session(session)
-        except Exception as e:
-            # If DB succeeded but cache failed, raise cache error
-            if db_error is None:
-                raise RedisError(
-                    f"Failed to persist chat session {session.session_id} to Redis: {e}"
-                ) from e
-            # If both failed, log cache error but raise DB error (more critical)
-            logger.warning(
-                f"Cache write also failed for session {session.session_id}: {e}"
-            )
-
-        # Propagate DB error after attempting cache (prevents data loss)
-        if db_error is not None:
-            raise DatabaseError(
-                f"Failed to persist chat session {session.session_id} to database"
-            ) from db_error
-
-        return session
+    return session
 
 
-async def create_chat_session(user_id: str) -> ChatSession:
-    """Create a new chat session and persist it.
-
-    Raises:
-        DatabaseError: If the database write fails. We fail fast to ensure
-            callers never receive a non-persisted session that only exists
-            in cache (which would be lost when the cache expires).
-    """
+async def create_chat_session(user_id: str | None) -> ChatSession:
+    """Create a new chat session and persist it."""
     session = ChatSession.new(user_id)
 
-    # Create in database first - fail fast if this fails
+    # Create in database first
     try:
         await chat_db.create_chat_session(
             session_id=session.session_id,
             user_id=user_id,
         )
     except Exception as e:
-        logger.error(f"Failed to create session {session.session_id} in database: {e}")
-        raise DatabaseError(
-            f"Failed to create chat session {session.session_id} in database"
-        ) from e
+        logger.error(f"Failed to create session in database: {e}")
+        # Continue even if DB fails - cache will still work
 
-    # Cache the session (best-effort optimization, DB is source of truth)
+    # Cache the session
     try:
         await _cache_session(session)
     except Exception as e:
-        logger.warning(f"Failed to cache new session {session.session_id}: {e}")
+        logger.warning(f"Failed to cache new session: {e}")
 
     return session
 
@@ -512,86 +447,27 @@ async def get_user_sessions(
     user_id: str,
     limit: int = 50,
     offset: int = 0,
-) -> tuple[list[ChatSession], int]:
-    """Get chat sessions for a user from the database with total count.
-
-    Returns:
-        A tuple of (sessions, total_count) where total_count is the overall
-        number of sessions for the user (not just the current page).
-    """
+) -> list[ChatSession]:
+    """Get all chat sessions for a user from the database."""
     prisma_sessions = await chat_db.get_user_chat_sessions(user_id, limit, offset)
-    total_count = await chat_db.get_user_session_count(user_id)
 
     sessions = []
     for prisma_session in prisma_sessions:
         # Convert without messages for listing (lighter weight)
-        sessions.append(ChatSession.from_db(prisma_session, None))
+        sessions.append(ChatSession.from_prisma(prisma_session, None))
 
-    return sessions, total_count
+    return sessions
 
 
-async def delete_chat_session(session_id: str, user_id: str | None = None) -> bool:
-    """Delete a chat session from both cache and database.
-
-    Args:
-        session_id: The session ID to delete.
-        user_id: If provided, validates that the session belongs to this user
-            before deletion. This prevents unauthorized deletion.
-
-    Returns:
-        True if deleted successfully, False otherwise.
-    """
-    # Delete from database first (with optional user_id validation)
-    # This confirms ownership before invalidating cache
-    deleted = await chat_db.delete_chat_session(session_id, user_id)
-
-    if not deleted:
-        return False
-
-    # Only invalidate cache and clean up lock after DB confirms deletion
+async def delete_chat_session(session_id: str) -> bool:
+    """Delete a chat session from both cache and database."""
+    # Delete from cache
     try:
-        redis_key = _get_session_cache_key(session_id)
+        redis_key = f"chat:session:{session_id}"
         async_redis = await get_redis_async()
         await async_redis.delete(redis_key)
     except Exception as e:
         logger.warning(f"Failed to delete session {session_id} from cache: {e}")
 
-    # Clean up session lock (belt-and-suspenders with WeakValueDictionary)
-    async with _session_locks_mutex:
-        _session_locks.pop(session_id, None)
-
-    return True
-
-
-async def update_session_title(session_id: str, title: str) -> bool:
-    """Update only the title of a chat session.
-
-    This is a lightweight operation that doesn't touch messages, avoiding
-    race conditions with concurrent message updates. Use this for background
-    title generation instead of upsert_chat_session.
-
-    Args:
-        session_id: The session ID to update.
-        title: The new title to set.
-
-    Returns:
-        True if updated successfully, False otherwise.
-    """
-    try:
-        result = await chat_db.update_chat_session(session_id=session_id, title=title)
-        if result is None:
-            logger.warning(f"Session {session_id} not found for title update")
-            return False
-
-        # Invalidate cache so next fetch gets updated title
-        try:
-            redis_key = _get_session_cache_key(session_id)
-            async_redis = await get_redis_async()
-            await async_redis.delete(redis_key)
-        except Exception as e:
-            logger.warning(f"Failed to invalidate cache for session {session_id}: {e}")
-
-        return True
-    except Exception as e:
-        logger.error(f"Failed to update title for session {session_id}: {e}")
-        return False
+    # Delete from database
+    return await chat_db.delete_chat_session(session_id)

autogpt_platform/backend/backend/api/features/chat/model_test.py

@@ -43,9 +43,9 @@ async def test_chatsession_serialization_deserialization():
 
 
 @pytest.mark.asyncio(loop_scope="session")
-async def test_chatsession_redis_storage(setup_test_user, test_user_id):
+async def test_chatsession_redis_storage():
 
-    s = ChatSession.new(user_id=test_user_id)
+    s = ChatSession.new(user_id=None)
     s.messages = messages
 
     s = await upsert_chat_session(s)
@@ -59,28 +59,26 @@ async def test_chatsession_redis_storage(setup_test_user, test_user_id):
 
 
 @pytest.mark.asyncio(loop_scope="session")
-async def test_chatsession_redis_storage_user_id_mismatch(
-    setup_test_user, test_user_id
-):
+async def test_chatsession_redis_storage_user_id_mismatch():
 
-    s = ChatSession.new(user_id=test_user_id)
+    s = ChatSession.new(user_id="abc123")
     s.messages = messages
     s = await upsert_chat_session(s)
 
-    s2 = await get_chat_session(s.session_id, "different_user_id")
+    s2 = await get_chat_session(s.session_id, None)
 
     assert s2 is None
 
 
 @pytest.mark.asyncio(loop_scope="session")
-async def test_chatsession_db_storage(setup_test_user, test_user_id):
+async def test_chatsession_db_storage():
     """Test that messages are correctly saved to and loaded from DB (not cache)."""
     from backend.data.redis_client import get_redis_async
 
     # Create session with messages including assistant message
-    s = ChatSession.new(user_id=test_user_id)
+    s = ChatSession.new(user_id=None)
     s.messages = messages  # Contains user, assistant, and tool messages
-    assert s.session_id is not None, "Session id is not set"
+
     # Upsert to save to both cache and DB
     s = await upsert_chat_session(s)
 

autogpt_platform/backend/backend/api/features/chat/prompts/chat_system.md

@@ -0,0 +1,192 @@
+You are Otto, an AI Co-Pilot and Forward Deployed Engineer for AutoGPT, an AI Business Automation tool. Your mission is to help users quickly find, create, and set up AutoGPT agents to solve their business problems.
+
+Here are the functions available to you:
+
+<functions>
+**Understanding & Discovery:**
+1. **add_understanding** - Save information about the user's business context (use this as you learn about them)
+2. **find_agent** - Search the marketplace for pre-built agents that solve the user's problem
+3. **find_library_agent** - Search the user's personal library of saved agents
+4. **find_block** - Search for individual blocks (building components for agents)
+5. **search_platform_docs** - Search AutoGPT documentation for help
+
+**Agent Creation & Editing:**
+6. **create_agent** - Create a new custom agent from scratch based on user requirements
+7. **edit_agent** - Modify an existing agent (add/remove blocks, change configuration)
+
+**Execution & Output:**
+8. **run_agent** - Run or schedule an agent (automatically handles setup)
+9. **run_block** - Run a single block directly without creating an agent
+10. **agent_output** - Get the output/results from a running or completed agent execution
+</functions>
+
+## ALWAYS GET THE USER'S NAME
+
+**This is critical:** If you don't know the user's name, ask for it in your first response. Use a friendly, natural approach:
+- "Hi! I'm Otto. What's your name?"
+- "Hey there! Before we dive in, what should I call you?"
+
+Once you have their name, immediately save it with `add_understanding(user_name="...")` and use it throughout the conversation.
+
+## BUILDING USER UNDERSTANDING
+
+**If no User Business Context is provided below**, gather information naturally during conversation - don't interrogate them.
+
+**Key information to gather (in priority order):**
+1. Their name (ALWAYS first if unknown)
+2. Their job title and role
+3. Their business/company and industry
+4. Pain points and what they want to automate
+5. Tools they currently use
+
+**How to gather this information:**
+- Ask naturally as part of helping them (e.g., "What's your role?" or "What industry are you in?")
+- When they share information, immediately save it using `add_understanding`
+- Don't ask all questions at once - spread them across the conversation
+- Prioritize understanding their immediate problem first
+
+**Example:**
+```
+User: "I need help automating my social media"
+Otto: I can help with that! I'm Otto - what's your name?
+User: "I'm Sarah"
+Otto: [calls add_understanding with user_name="Sarah"]
+Nice to meet you, Sarah! What's your role - are you a social media manager or business owner?
+User: "I'm the marketing director at a fintech startup"
+Otto: [calls add_understanding with job_title="Marketing Director", industry="fintech", business_size="startup"]
+Great! Let me find social media automation agents for you.
+[calls find_agent with query="social media automation marketing"]
+```
+
+## WHEN TO USE WHICH TOOL
+
+**Finding existing agents:**
+- `find_agent` - Search the marketplace for pre-built agents others have created
+- `find_library_agent` - Search agents the user has already saved to their library
+
+**Creating/editing agents:**
+- `create_agent` - When user wants a custom agent that doesn't exist, or has specific requirements
+- `edit_agent` - When user wants to modify an existing agent (change inputs, add blocks, etc.)
+
+**Running agents:**
+- `run_agent` - To execute an agent (handles credentials and inputs automatically)
+- `agent_output` - To check the results of a running or completed agent execution
+
+**Direct execution:**
+- `run_block` - Run a single block directly without needing a full agent
+
+## HOW run_agent WORKS
+
+The `run_agent` tool automatically handles the entire setup flow:
+
+1. **First call** (no inputs) → Returns available inputs so user can decide what values to use
+2. **Credentials check** → If missing, UI automatically prompts user to add them (you don't need to mention this)
+3. **Execution** → Runs when you provide `inputs` OR set `use_defaults=true`
+
+Parameters:
+- `username_agent_slug` (required): Agent identifier like "creator/agent-name"
+- `inputs`: Object with input values for the agent
+- `use_defaults`: Set to `true` to run with default values (only after user confirms)
+- `schedule_name` + `cron`: For scheduled execution
+
+## HOW create_agent WORKS
+
+Use `create_agent` when the user wants to build a custom automation:
+- Describe what the agent should do
+- The tool will create the agent structure with appropriate blocks
+- Returns the agent ID for further editing or running
+
+## HOW agent_output WORKS
+
+Use `agent_output` to get results from agent executions:
+- Pass the execution_id from a run_agent response
+- Returns the current status and any outputs produced
+- Useful for checking if an agent has completed and what it produced
+
+## WORKFLOW
+
+1. **Get their name** - If unknown, ask for it first
+2. **Understand context** - Ask 1-2 questions about their problem while helping
+3. **Find or create** - Use find_agent for existing solutions, create_agent for custom needs
+4. **Set up and run** - Use run_agent to execute, agent_output to get results
+
+## YOUR APPROACH
+
+**Step 1: Greet and Identify**
+- If you don't know their name, ask for it
+- Be friendly and conversational
+
+**Step 2: Understand the Problem**
+- Ask maximum 1-2 targeted questions
+- Focus on: What business problem are they solving?
+- If they want to create/edit an agent, understand what it should do
+
+**Step 3: Find or Create**
+- For existing solutions: Use `find_agent` with relevant keywords
+- For custom needs: Use `create_agent` with their requirements
+- For modifications: Use `edit_agent` on an existing agent
+
+**Step 4: Execute**
+- Call `run_agent` without inputs first to see what's available
+- Ask user what values they want or if defaults are okay
+- Call `run_agent` again with inputs or `use_defaults=true`
+- Use `agent_output` to check results when needed
+
+## USING add_understanding
+
+Call `add_understanding` whenever you learn something about the user:
+
+**User info:** `user_name`, `job_title`
+**Business:** `business_name`, `industry`, `business_size` (1-10, 11-50, 51-200, 201-1000, 1000+), `user_role` (decision maker, implementer, end user)
+**Processes:** `key_workflows` (array), `daily_activities` (array)
+**Pain points:** `pain_points` (array), `bottlenecks` (array), `manual_tasks` (array), `automation_goals` (array)
+**Tools:** `current_software` (array), `existing_automation` (array)
+**Other:** `additional_notes`
+
+Example: `add_understanding(user_name="Sarah", job_title="Marketing Director", industry="fintech")`
+
+## KEY RULES
+
+**What You DON'T Do:**
+- Don't help with login (frontend handles this)
+- Don't mention or explain credentials to the user (frontend handles this automatically)
+- Don't run agents without first showing available inputs to the user
+- Don't use `use_defaults=true` without user explicitly confirming
+- Don't write responses longer than 3 sentences
+- Don't interrogate users with many questions - gather info naturally
+
+**What You DO:**
+- ALWAYS ask for user's name if you don't have it
+- Save user information with `add_understanding` as you learn it
+- Use their name when addressing them
+- Always call run_agent first without inputs to see what's available
+- Ask user what values they want OR if they want to use defaults
+- Keep all responses to maximum 3 sentences
+- Include the agent link in your response after successful execution
+
+**Error Handling:**
+- Authentication needed → "Please sign in via the interface"
+- Credentials missing → The UI handles this automatically. Focus on asking the user about input values instead.
+
+## RESPONSE STRUCTURE
+
+Before responding, wrap your analysis in <thinking> tags to systematically plan your approach:
+- Check if you know the user's name - if not, ask for it
+- Check if you have user context - if not, plan to gather some naturally
+- Extract the key business problem or request from the user's message
+- Determine what function call (if any) you need to make next
+- Plan your response to stay under the 3-sentence maximum
+
+Example interaction:
+```
+User: "Hi, I want to build an agent that monitors my competitors"
+Otto: <thinking>I don't know this user's name. I should ask for it while acknowledging their request.</thinking>
+Hi! I'm Otto and I'd love to help you build a competitor monitoring agent. What's your name?
+User: "I'm Mike"
+Otto: [calls add_understanding with user_name="Mike"]
+<thinking>Now I know Mike wants competitor monitoring. I should search for existing agents first.</thinking>
+Great to meet you, Mike! Let me search for competitor monitoring agents.
+[calls find_agent with query="competitor monitoring analysis"]
+```
+
+KEEP ANSWERS TO 3 SENTENCES

autogpt_platform/backend/backend/api/features/chat/prompts/onboarding_system.md

@@ -0,0 +1,155 @@
+You are Otto, an AI Co-Pilot helping new users get started with AutoGPT, an AI Business Automation platform. Your mission is to welcome them, learn about their needs, and help them run their first successful agent.
+
+Here are the functions available to you:
+
+<functions>
+**Understanding & Discovery:**
+1. **add_understanding** - Save information about the user's business context (use this as you learn about them)
+2. **find_agent** - Search the marketplace for pre-built agents that solve the user's problem
+3. **find_library_agent** - Search the user's personal library of saved agents
+4. **find_block** - Search for individual blocks (building components for agents)
+5. **search_platform_docs** - Search AutoGPT documentation for help
+
+**Agent Creation & Editing:**
+6. **create_agent** - Create a new custom agent from scratch based on user requirements
+7. **edit_agent** - Modify an existing agent (add/remove blocks, change configuration)
+
+**Execution & Output:**
+8. **run_agent** - Run or schedule an agent (automatically handles setup)
+9. **run_block** - Run a single block directly without creating an agent
+10. **agent_output** - Get the output/results from a running or completed agent execution
+</functions>
+
+## YOUR ONBOARDING MISSION
+
+You are guiding a new user through their first experience with AutoGPT. Your goal is to:
+1. Welcome them warmly and get their name
+2. Learn about them and their business
+3. Find or create an agent that solves a real problem for them
+4. Get that agent running successfully
+5. Celebrate their success and point them to next steps
+
+## PHASE 1: WELCOME & INTRODUCTION
+
+**Start every conversation by:**
+- Giving a warm, friendly greeting
+- Introducing yourself as Otto, their AI assistant
+- Asking for their name immediately
+
+**Example opening:**
+```
+Hi! I'm Otto, your AI assistant. Welcome to AutoGPT! I'm here to help you set up your first automation. What's your name?
+```
+
+Once you have their name, save it immediately with `add_understanding(user_name="...")` and use it throughout.
+
+## PHASE 2: DISCOVERY
+
+**After getting their name, learn about them:**
+- What's their role/job title?
+- What industry/business are they in?
+- What's one thing they'd love to automate?
+
+**Keep it conversational - don't interrogate. Example:**
+```
+Nice to meet you, Sarah! What do you do for work, and what's one task you wish you could automate?
+```
+
+Save everything you learn with `add_understanding`.
+
+## PHASE 3: FIND OR CREATE AN AGENT
+
+**Once you understand their need:**
+- Search for existing agents with `find_agent`
+- Present the best match and explain how it helps them
+- If nothing fits, offer to create a custom agent with `create_agent`
+
+**Be enthusiastic about the solution:**
+```
+I found a great agent for you! The "Social Media Scheduler" can automatically post to your accounts on a schedule. Want to try it?
+```
+
+## PHASE 4: SETUP & RUN
+
+**Guide them through running the agent:**
+1. Call `run_agent` without inputs first to see what's needed
+2. Explain each input in simple terms
+3. Ask what values they want to use
+4. Run the agent with their inputs or defaults
+
+**Don't mention credentials** - the UI handles that automatically.
+
+## PHASE 5: CELEBRATE & HANDOFF
+
+**After successful execution:**
+- Congratulate them on their first automation!
+- Tell them where to find this agent (their Library)
+- Mention they can explore more agents in the Marketplace
+- Offer to help with anything else
+
+**Example:**
+```
+You did it! Your first agent is running. You can find it anytime in your Library. Ready to explore more automations?
+```
+
+## KEY RULES
+
+**What You DON'T Do:**
+- Don't help with login (frontend handles this)
+- Don't mention credentials (UI handles automatically)
+- Don't run agents without showing inputs first
+- Don't use `use_defaults=true` without explicit confirmation
+- Don't write responses longer than 3 sentences
+- Don't overwhelm with too many questions at once
+
+**What You DO:**
+- ALWAYS get the user's name first
+- Be warm, encouraging, and celebratory
+- Save info with `add_understanding` as you learn it
+- Use their name when addressing them
+- Keep responses to maximum 3 sentences
+- Make them feel successful at each step
+
+## USING add_understanding
+
+Save information as you learn it:
+
+**User info:** `user_name`, `job_title`
+**Business:** `business_name`, `industry`, `business_size`, `user_role`
+**Pain points:** `pain_points`, `manual_tasks`, `automation_goals`
+**Tools:** `current_software`
+
+Example: `add_understanding(user_name="Sarah", job_title="Marketing Manager", automation_goals=["social media scheduling"])`
+
+## HOW run_agent WORKS
+
+1. **First call** (no inputs) → Shows available inputs
+2. **Credentials** → UI handles automatically (don't mention)
+3. **Execution** → Run with `inputs={...}` or `use_defaults=true`
+
+## RESPONSE STRUCTURE
+
+Before responding, plan your approach in <thinking> tags:
+- What phase am I in? (Welcome/Discovery/Find/Setup/Celebrate)
+- Do I know their name? If not, ask for it
+- What's the next step to move them forward?
+- Keep response under 3 sentences
+
+**Example flow:**
+```
+User: "Hi"
+Otto: <thinking>Phase 1 - I need to welcome them and get their name.</thinking>
+Hi! I'm Otto, welcome to AutoGPT! I'm here to help you set up your first automation - what's your name?
+
+User: "I'm Alex"
+Otto: [calls add_understanding with user_name="Alex"]
+<thinking>Got their name. Phase 2 - learn about them.</thinking>
+Great to meet you, Alex! What do you do for work, and what's one task you'd love to automate?
+
+User: "I run an e-commerce store and spend hours on customer support emails"
+Otto: [calls add_understanding with industry="e-commerce", pain_points=["customer support emails"]]
+<thinking>Phase 3 - search for agents.</thinking>
+[calls find_agent with query="customer support email automation"]
+```
+
+KEEP ANSWERS TO 3 SENTENCES - Be warm, helpful, and focused on their success!

autogpt_platform/backend/backend/api/features/chat/response_model.py

@@ -1,10 +1,3 @@
-"""
-Response models for Vercel AI SDK UI Stream Protocol.
-
-This module implements the AI SDK UI Stream Protocol (v1) for streaming chat responses.
-See: https://ai-sdk.dev/docs/ai-sdk-ui/stream-protocol
-"""
-
 from enum import Enum
 from typing import Any
 
@@ -12,133 +5,97 @@ from pydantic import BaseModel, Field
 
 
 class ResponseType(str, Enum):
-    """Types of streaming responses following AI SDK protocol."""
+    """Types of streaming responses."""
 
-    # Message lifecycle
-    START = "start"
-    FINISH = "finish"
-
-    # Text streaming
-    TEXT_START = "text-start"
-    TEXT_DELTA = "text-delta"
-    TEXT_END = "text-end"
-
-    # Tool interaction
-    TOOL_INPUT_START = "tool-input-start"
-    TOOL_INPUT_AVAILABLE = "tool-input-available"
-    TOOL_OUTPUT_AVAILABLE = "tool-output-available"
-
-    # Other
+    TEXT_CHUNK = "text_chunk"
+    TEXT_ENDED = "text_ended"
+    TOOL_CALL = "tool_call"
+    TOOL_CALL_START = "tool_call_start"
+    TOOL_RESPONSE = "tool_response"
     ERROR = "error"
     USAGE = "usage"
+    STREAM_END = "stream_end"
 
 
 class StreamBaseResponse(BaseModel):
     """Base response model for all streaming responses."""
 
     type: ResponseType
+    timestamp: str | None = None
 
     def to_sse(self) -> str:
         """Convert to SSE format."""
         return f"data: {self.model_dump_json()}\n\n"
 
 
-# ========== Message Lifecycle ==========
+class StreamTextChunk(StreamBaseResponse):
+    """Streaming text content from the assistant."""
+
+    type: ResponseType = ResponseType.TEXT_CHUNK
+    content: str = Field(..., description="Text content chunk")
 
 
-class StreamStart(StreamBaseResponse):
-    """Start of a new message."""
-
-    type: ResponseType = ResponseType.START
-    messageId: str = Field(..., description="Unique message ID")
-
-
-class StreamFinish(StreamBaseResponse):
-    """End of message/stream."""
-
-    type: ResponseType = ResponseType.FINISH
-
-
-# ========== Text Streaming ==========
-
-
-class StreamTextStart(StreamBaseResponse):
-    """Start of a text block."""
-
-    type: ResponseType = ResponseType.TEXT_START
-    id: str = Field(..., description="Text block ID")
-
-
-class StreamTextDelta(StreamBaseResponse):
-    """Streaming text content delta."""
-
-    type: ResponseType = ResponseType.TEXT_DELTA
-    id: str = Field(..., description="Text block ID")
-    delta: str = Field(..., description="Text content delta")
-
-
-class StreamTextEnd(StreamBaseResponse):
-    """End of a text block."""
-
-    type: ResponseType = ResponseType.TEXT_END
-    id: str = Field(..., description="Text block ID")
-
-
-# ========== Tool Interaction ==========
-
-
-class StreamToolInputStart(StreamBaseResponse):
+class StreamToolCallStart(StreamBaseResponse):
     """Tool call started notification."""
 
-    type: ResponseType = ResponseType.TOOL_INPUT_START
-    toolCallId: str = Field(..., description="Unique tool call ID")
-    toolName: str = Field(..., description="Name of the tool being called")
+    type: ResponseType = ResponseType.TOOL_CALL_START
+    tool_name: str = Field(..., description="Name of the tool that was executed")
+    tool_id: str = Field(..., description="Unique tool call ID")
 
 
-class StreamToolInputAvailable(StreamBaseResponse):
-    """Tool input is ready for execution."""
+class StreamToolCall(StreamBaseResponse):
+    """Tool invocation notification."""
 
-    type: ResponseType = ResponseType.TOOL_INPUT_AVAILABLE
-    toolCallId: str = Field(..., description="Unique tool call ID")
-    toolName: str = Field(..., description="Name of the tool being called")
-    input: dict[str, Any] = Field(
-        default_factory=dict, description="Tool input arguments"
+    type: ResponseType = ResponseType.TOOL_CALL
+    tool_id: str = Field(..., description="Unique tool call ID")
+    tool_name: str = Field(..., description="Name of the tool being called")
+    arguments: dict[str, Any] = Field(
+        default_factory=dict, description="Tool arguments"
     )
 
 
-class StreamToolOutputAvailable(StreamBaseResponse):
+class StreamToolExecutionResult(StreamBaseResponse):
     """Tool execution result."""
 
-    type: ResponseType = ResponseType.TOOL_OUTPUT_AVAILABLE
-    toolCallId: str = Field(..., description="Tool call ID this responds to")
-    output: str | dict[str, Any] = Field(..., description="Tool execution output")
-    # Additional fields for internal use (not part of AI SDK spec but useful)
-    toolName: str | None = Field(
-        default=None, description="Name of the tool that was executed"
-    )
+    type: ResponseType = ResponseType.TOOL_RESPONSE
+    tool_id: str = Field(..., description="Tool call ID this responds to")
+    tool_name: str = Field(..., description="Name of the tool that was executed")
+    result: str | dict[str, Any] = Field(..., description="Tool execution result")
     success: bool = Field(
         default=True, description="Whether the tool execution succeeded"
     )
 
 
-# ========== Other ==========
-
-
 class StreamUsage(StreamBaseResponse):
     """Token usage statistics."""
 
     type: ResponseType = ResponseType.USAGE
-    promptTokens: int = Field(..., description="Number of prompt tokens")
-    completionTokens: int = Field(..., description="Number of completion tokens")
-    totalTokens: int = Field(..., description="Total number of tokens")
+    prompt_tokens: int
+    completion_tokens: int
+    total_tokens: int
 
 
 class StreamError(StreamBaseResponse):
     """Error response."""
 
     type: ResponseType = ResponseType.ERROR
-    errorText: str = Field(..., description="Error message text")
+    message: str = Field(..., description="Error message")
     code: str | None = Field(default=None, description="Error code")
     details: dict[str, Any] | None = Field(
         default=None, description="Additional error details"
     )
+
+
+class StreamTextEnded(StreamBaseResponse):
+    """Text streaming completed marker."""
+
+    type: ResponseType = ResponseType.TEXT_ENDED
+
+
+class StreamEnd(StreamBaseResponse):
+    """End of stream marker."""
+
+    type: ResponseType = ResponseType.STREAM_END
+    summary: dict[str, Any] | None = Field(
+        default=None, description="Stream summary statistics"
+    )

autogpt_platform/backend/backend/api/features/chat/routes.py

@@ -13,25 +13,12 @@ from backend.util.exceptions import NotFoundError
 
 from . import service as chat_service
 from .config import ChatConfig
-from .model import ChatSession, create_chat_session, get_chat_session, get_user_sessions
 
 config = ChatConfig()
 
 
 logger = logging.getLogger(__name__)
 
-
-async def _validate_and_get_session(
-    session_id: str,
-    user_id: str | None,
-) -> ChatSession:
-    """Validate session exists and belongs to user."""
-    session = await get_chat_session(session_id, user_id)
-    if not session:
-        raise NotFoundError(f"Session {session_id} not found.")
-    return session
-
-
 router = APIRouter(
     tags=["chat"],
 )
@@ -107,7 +94,7 @@ async def list_sessions(
     Returns:
         ListSessionsResponse: List of session summaries and total count.
     """
-    sessions, total_count = await get_user_sessions(user_id, limit, offset)
+    sessions = await chat_service.get_user_sessions(user_id, limit, offset)
 
     return ListSessionsResponse(
         sessions=[
@@ -115,11 +102,11 @@ async def list_sessions(
                 id=session.session_id,
                 created_at=session.started_at.isoformat(),
                 updated_at=session.updated_at.isoformat(),
-                title=session.title,
+                title=None,  # TODO: Add title support
             )
             for session in sessions
         ],
-        total=total_count,
+        total=len(sessions),
     )
 
 
@@ -127,15 +114,15 @@ async def list_sessions(
     "/sessions",
 )
 async def create_session(
-    user_id: Annotated[str, Depends(auth.get_user_id)],
+    user_id: Annotated[str | None, Depends(auth.get_user_id)],
 ) -> CreateSessionResponse:
     """
     Create a new chat session.
 
-    Initiates a new chat session for the authenticated user.
+    Initiates a new chat session for either an authenticated or anonymous user.
 
     Args:
-        user_id: The authenticated user ID parsed from the JWT (required).
+        user_id: The optional authenticated user ID parsed from the JWT. If missing, creates an anonymous session.
 
     Returns:
         CreateSessionResponse: Details of the created session.
@@ -143,15 +130,15 @@ async def create_session(
     """
     logger.info(
         f"Creating session with user_id: "
-        f"...{user_id[-8:] if len(user_id) > 8 else '<redacted>'}"
+        f"...{user_id[-8:] if user_id and len(user_id) > 8 else '<redacted>'}"
     )
 
-    session = await create_chat_session(user_id)
+    session = await chat_service.create_chat_session(user_id)
 
     return CreateSessionResponse(
         id=session.session_id,
         created_at=session.started_at.isoformat(),
-        user_id=session.user_id,
+        user_id=session.user_id or None,
     )
 
 
@@ -175,7 +162,7 @@ async def get_session(
         SessionDetailResponse: Details for the requested session; raises NotFoundError if not found.
 
     """
-    session = await get_chat_session(session_id, user_id)
+    session = await chat_service.get_session(session_id, user_id)
     if not session:
         raise NotFoundError(f"Session {session_id} not found")
 
@@ -219,7 +206,14 @@ async def stream_chat_post(
         StreamingResponse: SSE-formatted response chunks.
 
     """
-    session = await _validate_and_get_session(session_id, user_id)
+    # Validate session exists before starting the stream
+    # This prevents errors after the response has already started
+    session = await chat_service.get_session(session_id, user_id)
+
+    if not session:
+        raise NotFoundError(f"Session {session_id} not found. ")
+    if session.user_id is None and user_id is not None:
+        session = await chat_service.assign_user_to_session(session_id, user_id)
 
     async def event_generator() -> AsyncGenerator[str, None]:
         async for chunk in chat_service.stream_chat_completion(
@@ -231,8 +225,6 @@ async def stream_chat_post(
             context=request.context,
         ):
             yield chunk.to_sse()
-        # AI SDK protocol termination
-        yield "data: [DONE]\n\n"
 
     return StreamingResponse(
         event_generator(),
@@ -241,7 +233,6 @@ async def stream_chat_post(
             "Cache-Control": "no-cache",
             "Connection": "keep-alive",
             "X-Accel-Buffering": "no",  # Disable nginx buffering
-            "x-vercel-ai-ui-message-stream": "v1",  # AI SDK protocol header
         },
     )
 
@@ -272,7 +263,14 @@ async def stream_chat_get(
         StreamingResponse: SSE-formatted response chunks.
 
     """
-    session = await _validate_and_get_session(session_id, user_id)
+    # Validate session exists before starting the stream
+    # This prevents errors after the response has already started
+    session = await chat_service.get_session(session_id, user_id)
+
+    if not session:
+        raise NotFoundError(f"Session {session_id} not found. ")
+    if session.user_id is None and user_id is not None:
+        session = await chat_service.assign_user_to_session(session_id, user_id)
 
     async def event_generator() -> AsyncGenerator[str, None]:
         async for chunk in chat_service.stream_chat_completion(
@@ -283,8 +281,6 @@ async def stream_chat_get(
             session=session,  # Pass pre-fetched session to avoid double-fetch
         ):
             yield chunk.to_sse()
-        # AI SDK protocol termination
-        yield "data: [DONE]\n\n"
 
     return StreamingResponse(
         event_generator(),
@@ -293,7 +289,6 @@ async def stream_chat_get(
             "Cache-Control": "no-cache",
             "Connection": "keep-alive",
             "X-Accel-Buffering": "no",  # Disable nginx buffering
-            "x-vercel-ai-ui-message-stream": "v1",  # AI SDK protocol header
         },
     )
 
@@ -324,6 +319,133 @@ async def session_assign_user(
     return {"status": "ok"}
 
 
+# ========== Onboarding Routes ==========
+# These routes use a specialized onboarding system prompt
+
+
+@router.post(
+    "/onboarding/sessions",
+)
+async def create_onboarding_session(
+    user_id: Annotated[str | None, Depends(auth.get_user_id)],
+) -> CreateSessionResponse:
+    """
+    Create a new onboarding chat session.
+
+    Initiates a new chat session specifically for user onboarding,
+    using a specialized prompt that guides users through their first
+    experience with AutoGPT.
+
+    Args:
+        user_id: The optional authenticated user ID parsed from the JWT.
+
+    Returns:
+        CreateSessionResponse: Details of the created onboarding session.
+    """
+    logger.info(
+        f"Creating onboarding session with user_id: "
+        f"...{user_id[-8:] if user_id and len(user_id) > 8 else '<redacted>'}"
+    )
+
+    session = await chat_service.create_chat_session(user_id)
+
+    return CreateSessionResponse(
+        id=session.session_id,
+        created_at=session.started_at.isoformat(),
+        user_id=session.user_id or None,
+    )
+
+
+@router.get(
+    "/onboarding/sessions/{session_id}",
+)
+async def get_onboarding_session(
+    session_id: str,
+    user_id: Annotated[str | None, Depends(auth.get_user_id)],
+) -> SessionDetailResponse:
+    """
+    Retrieve the details of an onboarding chat session.
+
+    Args:
+        session_id: The unique identifier for the onboarding session.
+        user_id: The optional authenticated user ID.
+
+    Returns:
+        SessionDetailResponse: Details for the requested session.
+    """
+    session = await chat_service.get_session(session_id, user_id)
+    if not session:
+        raise NotFoundError(f"Session {session_id} not found")
+
+    messages = [message.model_dump() for message in session.messages]
+    logger.info(
+        f"Returning onboarding session {session_id}: "
+        f"message_count={len(messages)}, "
+        f"roles={[m.get('role') for m in messages]}"
+    )
+
+    return SessionDetailResponse(
+        id=session.session_id,
+        created_at=session.started_at.isoformat(),
+        updated_at=session.updated_at.isoformat(),
+        user_id=session.user_id or None,
+        messages=messages,
+    )
+
+
+@router.post(
+    "/onboarding/sessions/{session_id}/stream",
+)
+async def stream_onboarding_chat(
+    session_id: str,
+    request: StreamChatRequest,
+    user_id: str | None = Depends(auth.get_user_id),
+):
+    """
+    Stream onboarding chat responses for a session.
+
+    Uses the specialized onboarding system prompt to guide new users
+    through their first experience with AutoGPT. Streams AI responses
+    in real time over Server-Sent Events (SSE).
+
+    Args:
+        session_id: The onboarding session identifier.
+        request: Request body containing message and optional context.
+        user_id: Optional authenticated user ID.
+
+    Returns:
+        StreamingResponse: SSE-formatted response chunks.
+    """
+    session = await chat_service.get_session(session_id, user_id)
+
+    if not session:
+        raise NotFoundError(f"Session {session_id} not found.")
+    if session.user_id is None and user_id is not None:
+        session = await chat_service.assign_user_to_session(session_id, user_id)
+
+    async def event_generator() -> AsyncGenerator[str, None]:
+        async for chunk in chat_service.stream_chat_completion(
+            session_id,
+            request.message,
+            is_user_message=request.is_user_message,
+            user_id=user_id,
+            session=session,
+            context=request.context,
+            prompt_type="onboarding",  # Use onboarding system prompt
+        ):
+            yield chunk.to_sse()
+
+    return StreamingResponse(
+        event_generator(),
+        media_type="text/event-stream",
+        headers={
+            "Cache-Control": "no-cache",
+            "Connection": "keep-alive",
+            "X-Accel-Buffering": "no",
+        },
+    )
+
+
 # ========== Health Check ==========
 
 
@@ -332,28 +454,16 @@ async def health_check() -> dict:
     """
     Health check endpoint for the chat service.
 
-    Performs a full cycle test of session creation and retrieval. Should always return healthy
+    Performs a full cycle test of session creation, assignment, and retrieval. Should always return healthy
     if the service and data layer are operational.
 
     Returns:
         dict: A status dictionary indicating health, service name, and API version.
 
     """
-    from backend.data.user import get_or_create_user
-
-    # Ensure health check user exists (required for FK constraint)
-    health_check_user_id = "health-check-user"
-    await get_or_create_user(
-        {
-            "sub": health_check_user_id,
-            "email": "health-check@system.local",
-            "user_metadata": {"name": "Health Check User"},
-        }
-    )
-
-    # Create and retrieve session to verify full data layer
-    session = await create_chat_session(health_check_user_id)
-    await get_chat_session(session.session_id, health_check_user_id)
+    session = await chat_service.create_chat_session(None)
+    await chat_service.assign_user_to_session(session.session_id, "test_user")
+    await chat_service.get_session(session.session_id, "test_user")
 
     return {
         "status": "healthy",

autogpt_platform/backend/backend/api/features/chat/service_test.py

@@ -4,19 +4,18 @@ from os import getenv
 import pytest
 
 from . import service as chat_service
-from .model import create_chat_session, get_chat_session, upsert_chat_session
 from .response_model import (
+    StreamEnd,
     StreamError,
-    StreamFinish,
-    StreamTextDelta,
-    StreamToolOutputAvailable,
+    StreamTextChunk,
+    StreamToolExecutionResult,
 )
 
 logger = logging.getLogger(__name__)
 
 
 @pytest.mark.asyncio(loop_scope="session")
-async def test_stream_chat_completion(setup_test_user, test_user_id):
+async def test_stream_chat_completion():
     """
     Test the stream_chat_completion function.
     """
@@ -24,7 +23,7 @@ async def test_stream_chat_completion(setup_test_user, test_user_id):
     if not api_key:
         return pytest.skip("OPEN_ROUTER_API_KEY is not set, skipping test")
 
-    session = await create_chat_session(test_user_id)
+    session = await chat_service.create_chat_session()
 
     has_errors = False
     has_ended = False
@@ -35,9 +34,9 @@ async def test_stream_chat_completion(setup_test_user, test_user_id):
         logger.info(chunk)
         if isinstance(chunk, StreamError):
             has_errors = True
-        if isinstance(chunk, StreamTextDelta):
-            assistant_message += chunk.delta
-        if isinstance(chunk, StreamFinish):
+        if isinstance(chunk, StreamTextChunk):
+            assistant_message += chunk.content
+        if isinstance(chunk, StreamEnd):
             has_ended = True
 
     assert has_ended, "Chat completion did not end"
@@ -46,7 +45,7 @@ async def test_stream_chat_completion(setup_test_user, test_user_id):
 
 
 @pytest.mark.asyncio(loop_scope="session")
-async def test_stream_chat_completion_with_tool_calls(setup_test_user, test_user_id):
+async def test_stream_chat_completion_with_tool_calls():
     """
     Test the stream_chat_completion function.
     """
@@ -54,8 +53,8 @@ async def test_stream_chat_completion_with_tool_calls(setup_test_user, test_user
     if not api_key:
         return pytest.skip("OPEN_ROUTER_API_KEY is not set, skipping test")
 
-    session = await create_chat_session(test_user_id)
-    session = await upsert_chat_session(session)
+    session = await chat_service.create_chat_session()
+    session = await chat_service.upsert_chat_session(session)
 
     has_errors = False
     has_ended = False
@@ -69,14 +68,14 @@ async def test_stream_chat_completion_with_tool_calls(setup_test_user, test_user
         if isinstance(chunk, StreamError):
             has_errors = True
 
-        if isinstance(chunk, StreamFinish):
+        if isinstance(chunk, StreamEnd):
             has_ended = True
-        if isinstance(chunk, StreamToolOutputAvailable):
+        if isinstance(chunk, StreamToolExecutionResult):
             had_tool_calls = True
 
     assert has_ended, "Chat completion did not end"
     assert not has_errors, "Error occurred while streaming chat completion"
     assert had_tool_calls, "Tool calls did not occur"
-    session = await get_chat_session(session.session_id)
+    session = await chat_service.get_session(session.session_id)
     assert session, "Session not found"
     assert session.usage, "Usage is empty"

autogpt_platform/backend/backend/api/features/chat/tools/__init__.py

@@ -7,29 +7,42 @@ from backend.api.features.chat.model import ChatSession
 from .add_understanding import AddUnderstandingTool
 from .agent_output import AgentOutputTool
 from .base import BaseTool
+from .create_agent import CreateAgentTool
+from .edit_agent import EditAgentTool
 from .find_agent import FindAgentTool
+from .find_block import FindBlockTool
 from .find_library_agent import FindLibraryAgentTool
 from .run_agent import RunAgentTool
+from .run_block import RunBlockTool
+from .search_docs import SearchDocsTool
 
 if TYPE_CHECKING:
-    from backend.api.features.chat.response_model import StreamToolOutputAvailable
+    from backend.api.features.chat.response_model import StreamToolExecutionResult
 
-# Single source of truth for all tools
-TOOL_REGISTRY: dict[str, BaseTool] = {
-    "add_understanding": AddUnderstandingTool(),
-    "find_agent": FindAgentTool(),
-    "find_library_agent": FindLibraryAgentTool(),
-    "run_agent": RunAgentTool(),
-    "agent_output": AgentOutputTool(),
-}
+# Initialize tool instances
+add_understanding_tool = AddUnderstandingTool()
+create_agent_tool = CreateAgentTool()
+edit_agent_tool = EditAgentTool()
+find_agent_tool = FindAgentTool()
+find_block_tool = FindBlockTool()
+find_library_agent_tool = FindLibraryAgentTool()
+run_agent_tool = RunAgentTool()
+run_block_tool = RunBlockTool()
+search_docs_tool = SearchDocsTool()
+agent_output_tool = AgentOutputTool()
 
-# Export individual tool instances for backwards compatibility
-find_agent_tool = TOOL_REGISTRY["find_agent"]
-run_agent_tool = TOOL_REGISTRY["run_agent"]
-
-# Generated from registry for OpenAI API
+# Export tools as OpenAI format
 tools: list[ChatCompletionToolParam] = [
-    tool.as_openai_tool() for tool in TOOL_REGISTRY.values()
+    add_understanding_tool.as_openai_tool(),
+    create_agent_tool.as_openai_tool(),
+    edit_agent_tool.as_openai_tool(),
+    find_agent_tool.as_openai_tool(),
+    find_block_tool.as_openai_tool(),
+    find_library_agent_tool.as_openai_tool(),
+    run_agent_tool.as_openai_tool(),
+    run_block_tool.as_openai_tool(),
+    search_docs_tool.as_openai_tool(),
+    agent_output_tool.as_openai_tool(),
 ]
 
 
@@ -39,9 +52,22 @@ async def execute_tool(
     user_id: str | None,
     session: ChatSession,
     tool_call_id: str,
-) -> "StreamToolOutputAvailable":
-    """Execute a tool by name."""
-    tool = TOOL_REGISTRY.get(tool_name)
-    if not tool:
+) -> "StreamToolExecutionResult":
+
+    tool_map: dict[str, BaseTool] = {
+        "add_understanding": add_understanding_tool,
+        "create_agent": create_agent_tool,
+        "edit_agent": edit_agent_tool,
+        "find_agent": find_agent_tool,
+        "find_block": find_block_tool,
+        "find_library_agent": find_library_agent_tool,
+        "run_agent": run_agent_tool,
+        "run_block": run_block_tool,
+        "search_platform_docs": search_docs_tool,
+        "agent_output": agent_output_tool,
+    }
+    if tool_name not in tool_map:
         raise ValueError(f"Tool {tool_name} not found")
-    return await tool.execute(user_id, session, tool_call_id, **parameters)
+    return await tool_map[tool_name].execute(
+        user_id, session, tool_call_id, **parameters
+    )

autogpt_platform/backend/backend/api/features/chat/tools/_test_data.py

@@ -3,7 +3,6 @@ from datetime import UTC, datetime
 from os import getenv
 
 import pytest
-from prisma.types import ProfileCreateInput
 from pydantic import SecretStr
 
 from backend.api.features.chat.model import ChatSession
@@ -18,7 +17,7 @@ from backend.data.user import get_or_create_user
 from backend.integrations.credentials_store import IntegrationCredentialsStore
 
 
-def make_session(user_id: str):
+def make_session(user_id: str | None = None):
     return ChatSession(
         session_id=str(uuid.uuid4()),
         user_id=user_id,
@@ -50,13 +49,13 @@ async def setup_test_data():
     # 1b. Create a profile with username for the user (required for store agent lookup)
     username = user.email.split("@")[0]
     await prisma.profile.create(
-        data=ProfileCreateInput(
-            userId=user.id,
-            username=username,
-            name=f"Test User {username}",
-            description="Test user profile",
-            links=[],  # Required field - empty array for test profiles
-        )
+        data={
+            "userId": user.id,
+            "username": username,
+            "name": f"Test User {username}",
+            "description": "Test user profile",
+            "links": [],  # Required field - empty array for test profiles
+        }
     )
 
     # 2. Create a test graph with agent input -> agent output
@@ -173,13 +172,13 @@ async def setup_llm_test_data():
     # 1b. Create a profile with username for the user (required for store agent lookup)
     username = user.email.split("@")[0]
     await prisma.profile.create(
-        data=ProfileCreateInput(
-            userId=user.id,
-            username=username,
-            name=f"Test User {username}",
-            description="Test user profile for LLM tests",
-            links=[],  # Required field - empty array for test profiles
-        )
+        data={
+            "userId": user.id,
+            "username": username,
+            "name": f"Test User {username}",
+            "description": "Test user profile for LLM tests",
+            "links": [],  # Required field - empty array for test profiles
+        }
     )
 
     # 2. Create test OpenAI credentials for the user
@@ -333,13 +332,13 @@ async def setup_firecrawl_test_data():
     # 1b. Create a profile with username for the user (required for store agent lookup)
     username = user.email.split("@")[0]
     await prisma.profile.create(
-        data=ProfileCreateInput(
-            userId=user.id,
-            username=username,
-            name=f"Test User {username}",
-            description="Test user profile for Firecrawl tests",
-            links=[],  # Required field - empty array for test profiles
-        )
+        data={
+            "userId": user.id,
+            "username": username,
+            "name": f"Test User {username}",
+            "description": "Test user profile for Firecrawl tests",
+            "links": [],  # Required field - empty array for test profiles
+        }
     )
 
     # NOTE: We deliberately do NOT create Firecrawl credentials for this user

autogpt_platform/backend/backend/api/features/chat/tools/add_understanding.py

@@ -10,7 +10,11 @@ from backend.data.understanding import (
 )
 
 from .base import BaseTool
-from .models import ErrorResponse, ToolResponseBase, UnderstandingUpdatedResponse
+from .models import (
+    ErrorResponse,
+    ToolResponseBase,
+    UnderstandingUpdatedResponse,
+)
 
 logger = logging.getLogger(__name__)
 
@@ -34,25 +38,80 @@ and automations for the user's specific needs."""
 
     @property
     def parameters(self) -> dict[str, Any]:
-        # Auto-generate from Pydantic model schema
-        schema = BusinessUnderstandingInput.model_json_schema()
-        properties = {}
-        for field_name, field_schema in schema.get("properties", {}).items():
-            prop: dict[str, Any] = {"description": field_schema.get("description", "")}
-            # Handle anyOf for Optional types
-            if "anyOf" in field_schema:
-                for option in field_schema["anyOf"]:
-                    if option.get("type") != "null":
-                        prop["type"] = option.get("type", "string")
-                        if "items" in option:
-                            prop["items"] = option["items"]
-                        break
-            else:
-                prop["type"] = field_schema.get("type", "string")
-                if "items" in field_schema:
-                    prop["items"] = field_schema["items"]
-            properties[field_name] = prop
-        return {"type": "object", "properties": properties, "required": []}
+        return {
+            "type": "object",
+            "properties": {
+                "user_name": {
+                    "type": "string",
+                    "description": "The user's name",
+                },
+                "job_title": {
+                    "type": "string",
+                    "description": "The user's job title (e.g., 'Marketing Manager', 'CEO', 'Software Engineer')",
+                },
+                "business_name": {
+                    "type": "string",
+                    "description": "Name of the user's business or organization",
+                },
+                "industry": {
+                    "type": "string",
+                    "description": "Industry or sector (e.g., 'e-commerce', 'healthcare', 'finance')",
+                },
+                "business_size": {
+                    "type": "string",
+                    "description": "Company size: '1-10', '11-50', '51-200', '201-1000', or '1000+'",
+                },
+                "user_role": {
+                    "type": "string",
+                    "description": "User's role in organization context (e.g., 'decision maker', 'implementer', 'end user')",
+                },
+                "key_workflows": {
+                    "type": "array",
+                    "items": {"type": "string"},
+                    "description": "Key business workflows (e.g., 'lead qualification', 'content publishing')",
+                },
+                "daily_activities": {
+                    "type": "array",
+                    "items": {"type": "string"},
+                    "description": "Regular daily activities the user performs",
+                },
+                "pain_points": {
+                    "type": "array",
+                    "items": {"type": "string"},
+                    "description": "Current pain points or challenges",
+                },
+                "bottlenecks": {
+                    "type": "array",
+                    "items": {"type": "string"},
+                    "description": "Process bottlenecks slowing things down",
+                },
+                "manual_tasks": {
+                    "type": "array",
+                    "items": {"type": "string"},
+                    "description": "Manual or repetitive tasks that could be automated",
+                },
+                "automation_goals": {
+                    "type": "array",
+                    "items": {"type": "string"},
+                    "description": "Desired automation outcomes or goals",
+                },
+                "current_software": {
+                    "type": "array",
+                    "items": {"type": "string"},
+                    "description": "Software and tools currently in use",
+                },
+                "existing_automation": {
+                    "type": "array",
+                    "items": {"type": "string"},
+                    "description": "Any existing automations or integrations",
+                },
+                "additional_notes": {
+                    "type": "string",
+                    "description": "Any other relevant context or notes",
+                },
+            },
+            "required": [],
+        }
 
     @property
     def requires_auth(self) -> bool:
@@ -87,26 +146,54 @@ and automations for the user's specific needs."""
                 session_id=session_id,
             )
 
-        # Build input model from kwargs (only include fields defined in the model)
-        valid_fields = set(BusinessUnderstandingInput.model_fields.keys())
+        # Build input model
         input_data = BusinessUnderstandingInput(
-            **{k: v for k, v in kwargs.items() if k in valid_fields}
+            user_name=kwargs.get("user_name"),
+            job_title=kwargs.get("job_title"),
+            business_name=kwargs.get("business_name"),
+            industry=kwargs.get("industry"),
+            business_size=kwargs.get("business_size"),
+            user_role=kwargs.get("user_role"),
+            key_workflows=kwargs.get("key_workflows"),
+            daily_activities=kwargs.get("daily_activities"),
+            pain_points=kwargs.get("pain_points"),
+            bottlenecks=kwargs.get("bottlenecks"),
+            manual_tasks=kwargs.get("manual_tasks"),
+            automation_goals=kwargs.get("automation_goals"),
+            current_software=kwargs.get("current_software"),
+            existing_automation=kwargs.get("existing_automation"),
+            additional_notes=kwargs.get("additional_notes"),
         )
 
         # Track which fields were updated
-        updated_fields = [
-            k for k, v in kwargs.items() if k in valid_fields and v is not None
-        ]
+        updated_fields = [k for k, v in kwargs.items() if v is not None]
 
         # Upsert with merge
         understanding = await upsert_business_understanding(user_id, input_data)
 
-        # Build current understanding summary (filter out empty values)
+        # Build current understanding summary for the response
+        current_understanding = {
+            "user_name": understanding.user_name,
+            "job_title": understanding.job_title,
+            "business_name": understanding.business_name,
+            "industry": understanding.industry,
+            "business_size": understanding.business_size,
+            "user_role": understanding.user_role,
+            "key_workflows": understanding.key_workflows,
+            "daily_activities": understanding.daily_activities,
+            "pain_points": understanding.pain_points,
+            "bottlenecks": understanding.bottlenecks,
+            "manual_tasks": understanding.manual_tasks,
+            "automation_goals": understanding.automation_goals,
+            "current_software": understanding.current_software,
+            "existing_automation": understanding.existing_automation,
+            "additional_notes": understanding.additional_notes,
+        }
+
+        # Filter out empty values for cleaner response
         current_understanding = {
             k: v
-            for k, v in understanding.model_dump(
-                exclude={"id", "user_id", "created_at", "updated_at"}
-            ).items()
+            for k, v in current_understanding.items()
             if v is not None and v != [] and v != ""
         }
 

autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/__init__.py

@@ -0,0 +1,29 @@
+"""Agent generator package - Creates agents from natural language."""
+
+from .core import (
+    apply_agent_patch,
+    decompose_goal,
+    generate_agent,
+    generate_agent_patch,
+    get_agent_as_json,
+    save_agent_to_library,
+)
+from .fixer import apply_all_fixes
+from .utils import get_blocks_info
+from .validator import validate_agent
+
+__all__ = [
+    # Core functions
+    "decompose_goal",
+    "generate_agent",
+    "generate_agent_patch",
+    "apply_agent_patch",
+    "save_agent_to_library",
+    "get_agent_as_json",
+    # Fixer
+    "apply_all_fixes",
+    # Validator
+    "validate_agent",
+    # Utils
+    "get_blocks_info",
+]

autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/client.py

@@ -0,0 +1,25 @@
+"""OpenRouter client configuration for agent generation."""
+
+import os
+
+from openai import AsyncOpenAI
+
+# Configuration - use OPEN_ROUTER_API_KEY for consistency with chat/config.py
+OPENROUTER_API_KEY = os.getenv("OPEN_ROUTER_API_KEY") or os.getenv("OPENROUTER_API_KEY")
+AGENT_GENERATOR_MODEL = os.getenv("AGENT_GENERATOR_MODEL", "anthropic/claude-opus-4.5")
+
+# OpenRouter client (OpenAI-compatible API)
+_client: AsyncOpenAI | None = None
+
+
+def get_client() -> AsyncOpenAI:
+    """Get or create the OpenRouter client."""
+    global _client
+    if _client is None:
+        if not OPENROUTER_API_KEY:
+            raise ValueError("OPENROUTER_API_KEY environment variable is required")
+        _client = AsyncOpenAI(
+            base_url="https://openrouter.ai/api/v1",
+            api_key=OPENROUTER_API_KEY,
+        )
+    return _client

autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/core.py

@@ -0,0 +1,390 @@
+"""Core agent generation functions."""
+
+import copy
+import json
+import logging
+import uuid
+from typing import Any
+
+from backend.api.features.library import db as library_db
+from backend.data.graph import Graph, Link, Node, create_graph
+
+from .client import AGENT_GENERATOR_MODEL, get_client
+from .prompts import DECOMPOSITION_PROMPT, GENERATION_PROMPT, PATCH_PROMPT
+from .utils import get_block_summaries, parse_json_from_llm
+
+logger = logging.getLogger(__name__)
+
+
+async def decompose_goal(description: str, context: str = "") -> dict[str, Any] | None:
+    """Break down a goal into steps or return clarifying questions.
+
+    Args:
+        description: Natural language goal description
+        context: Additional context (e.g., answers to previous questions)
+
+    Returns:
+        Dict with either:
+        - {"type": "clarifying_questions", "questions": [...]}
+        - {"type": "instructions", "steps": [...]}
+        Or None on error
+    """
+    client = get_client()
+    prompt = DECOMPOSITION_PROMPT.format(block_summaries=get_block_summaries())
+
+    full_description = description
+    if context:
+        full_description = f"{description}\n\nAdditional context:\n{context}"
+
+    try:
+        response = await client.chat.completions.create(
+            model=AGENT_GENERATOR_MODEL,
+            messages=[
+                {"role": "system", "content": prompt},
+                {"role": "user", "content": full_description},
+            ],
+            temperature=0,
+        )
+
+        content = response.choices[0].message.content
+        if content is None:
+            logger.error("LLM returned empty content for decomposition")
+            return None
+
+        result = parse_json_from_llm(content)
+
+        if result is None:
+            logger.error(f"Failed to parse decomposition response: {content[:200]}")
+            return None
+
+        return result
+
+    except Exception as e:
+        logger.error(f"Error decomposing goal: {e}")
+        return None
+
+
+async def generate_agent(instructions: dict[str, Any]) -> dict[str, Any] | None:
+    """Generate agent JSON from instructions.
+
+    Args:
+        instructions: Structured instructions from decompose_goal
+
+    Returns:
+        Agent JSON dict or None on error
+    """
+    client = get_client()
+    prompt = GENERATION_PROMPT.format(block_summaries=get_block_summaries())
+
+    try:
+        response = await client.chat.completions.create(
+            model=AGENT_GENERATOR_MODEL,
+            messages=[
+                {"role": "system", "content": prompt},
+                {"role": "user", "content": json.dumps(instructions, indent=2)},
+            ],
+            temperature=0,
+        )
+
+        content = response.choices[0].message.content
+        if content is None:
+            logger.error("LLM returned empty content for agent generation")
+            return None
+
+        result = parse_json_from_llm(content)
+
+        if result is None:
+            logger.error(f"Failed to parse agent JSON: {content[:200]}")
+            return None
+
+        # Ensure required fields
+        if "id" not in result:
+            result["id"] = str(uuid.uuid4())
+        if "version" not in result:
+            result["version"] = 1
+        if "is_active" not in result:
+            result["is_active"] = True
+
+        return result
+
+    except Exception as e:
+        logger.error(f"Error generating agent: {e}")
+        return None
+
+
+def json_to_graph(agent_json: dict[str, Any]) -> Graph:
+    """Convert agent JSON dict to Graph model.
+
+    Args:
+        agent_json: Agent JSON with nodes and links
+
+    Returns:
+        Graph ready for saving
+    """
+    nodes = []
+    for n in agent_json.get("nodes", []):
+        node = Node(
+            id=n.get("id", str(uuid.uuid4())),
+            block_id=n["block_id"],
+            input_default=n.get("input_default", {}),
+            metadata=n.get("metadata", {}),
+        )
+        nodes.append(node)
+
+    links = []
+    for link_data in agent_json.get("links", []):
+        link = Link(
+            id=link_data.get("id", str(uuid.uuid4())),
+            source_id=link_data["source_id"],
+            sink_id=link_data["sink_id"],
+            source_name=link_data["source_name"],
+            sink_name=link_data["sink_name"],
+            is_static=link_data.get("is_static", False),
+        )
+        links.append(link)
+
+    return Graph(
+        id=agent_json.get("id", str(uuid.uuid4())),
+        version=agent_json.get("version", 1),
+        is_active=agent_json.get("is_active", True),
+        name=agent_json.get("name", "Generated Agent"),
+        description=agent_json.get("description", ""),
+        nodes=nodes,
+        links=links,
+    )
+
+
+def _reassign_node_ids(graph: Graph) -> None:
+    """Reassign all node and link IDs to new UUIDs.
+
+    This is needed when creating a new version to avoid unique constraint violations.
+    """
+    # Create mapping from old node IDs to new UUIDs
+    id_map = {node.id: str(uuid.uuid4()) for node in graph.nodes}
+
+    # Reassign node IDs
+    for node in graph.nodes:
+        node.id = id_map[node.id]
+
+    # Update link references to use new node IDs
+    for link in graph.links:
+        link.id = str(uuid.uuid4())  # Also give links new IDs
+        if link.source_id in id_map:
+            link.source_id = id_map[link.source_id]
+        if link.sink_id in id_map:
+            link.sink_id = id_map[link.sink_id]
+
+
+async def save_agent_to_library(
+    agent_json: dict[str, Any], user_id: str, is_update: bool = False
+) -> tuple[Graph, Any]:
+    """Save agent to database and user's library.
+
+    Args:
+        agent_json: Agent JSON dict
+        user_id: User ID
+        is_update: Whether this is an update to an existing agent
+
+    Returns:
+        Tuple of (created Graph, LibraryAgent)
+    """
+    from backend.data.graph import get_graph_all_versions
+
+    graph = json_to_graph(agent_json)
+
+    if is_update:
+        # For updates, keep the same graph ID but increment version
+        # and reassign node/link IDs to avoid conflicts
+        if graph.id:
+            existing_versions = await get_graph_all_versions(graph.id, user_id)
+            if existing_versions:
+                latest_version = max(v.version for v in existing_versions)
+                graph.version = latest_version + 1
+                # Reassign node IDs (but keep graph ID the same)
+                _reassign_node_ids(graph)
+                logger.info(f"Updating agent {graph.id} to version {graph.version}")
+    else:
+        # For new agents, always generate a fresh UUID to avoid collisions
+        graph.id = str(uuid.uuid4())
+        graph.version = 1
+        # Reassign all node IDs as well
+        _reassign_node_ids(graph)
+        logger.info(f"Creating new agent with ID {graph.id}")
+
+    # Save to database
+    created_graph = await create_graph(graph, user_id)
+
+    # Add to user's library (or update existing library agent)
+    library_agents = await library_db.create_library_agent(
+        graph=created_graph,
+        user_id=user_id,
+        create_library_agents_for_sub_graphs=False,
+    )
+
+    return created_graph, library_agents[0]
+
+
+async def get_agent_as_json(
+    graph_id: str, user_id: str | None
+) -> dict[str, Any] | None:
+    """Fetch an agent and convert to JSON format for editing.
+
+    Args:
+        graph_id: Graph ID or library agent ID
+        user_id: User ID
+
+    Returns:
+        Agent as JSON dict or None if not found
+    """
+    from backend.data.graph import get_graph
+
+    # Try to get the graph (version=None gets the active version)
+    graph = await get_graph(graph_id, version=None, user_id=user_id)
+    if not graph:
+        return None
+
+    # Convert to JSON format
+    nodes = []
+    for node in graph.nodes:
+        nodes.append(
+            {
+                "id": node.id,
+                "block_id": node.block_id,
+                "input_default": node.input_default,
+                "metadata": node.metadata,
+            }
+        )
+
+    links = []
+    for node in graph.nodes:
+        for link in node.output_links:
+            links.append(
+                {
+                    "id": link.id,
+                    "source_id": link.source_id,
+                    "sink_id": link.sink_id,
+                    "source_name": link.source_name,
+                    "sink_name": link.sink_name,
+                    "is_static": link.is_static,
+                }
+            )
+
+    return {
+        "id": graph.id,
+        "name": graph.name,
+        "description": graph.description,
+        "version": graph.version,
+        "is_active": graph.is_active,
+        "nodes": nodes,
+        "links": links,
+    }
+
+
+async def generate_agent_patch(
+    update_request: str, current_agent: dict[str, Any]
+) -> dict[str, Any] | None:
+    """Generate a patch to update an existing agent.
+
+    Args:
+        update_request: Natural language description of changes
+        current_agent: Current agent JSON
+
+    Returns:
+        Patch dict or clarifying questions, or None on error
+    """
+    client = get_client()
+    prompt = PATCH_PROMPT.format(
+        current_agent=json.dumps(current_agent, indent=2),
+        block_summaries=get_block_summaries(),
+    )
+
+    try:
+        response = await client.chat.completions.create(
+            model=AGENT_GENERATOR_MODEL,
+            messages=[
+                {"role": "system", "content": prompt},
+                {"role": "user", "content": update_request},
+            ],
+            temperature=0,
+        )
+
+        content = response.choices[0].message.content
+        if content is None:
+            logger.error("LLM returned empty content for patch generation")
+            return None
+
+        return parse_json_from_llm(content)
+
+    except Exception as e:
+        logger.error(f"Error generating patch: {e}")
+        return None
+
+
+def apply_agent_patch(
+    current_agent: dict[str, Any], patch: dict[str, Any]
+) -> dict[str, Any]:
+    """Apply a patch to an existing agent.
+
+    Args:
+        current_agent: Current agent JSON
+        patch: Patch dict with operations
+
+    Returns:
+        Updated agent JSON
+    """
+    agent = copy.deepcopy(current_agent)
+    patches = patch.get("patches", [])
+
+    for p in patches:
+        patch_type = p.get("type")
+
+        if patch_type == "modify":
+            node_id = p.get("node_id")
+            changes = p.get("changes", {})
+
+            for node in agent.get("nodes", []):
+                if node["id"] == node_id:
+                    _deep_update(node, changes)
+                    logger.debug(f"Modified node {node_id}")
+                    break
+
+        elif patch_type == "add":
+            new_nodes = p.get("new_nodes", [])
+            new_links = p.get("new_links", [])
+
+            agent["nodes"] = agent.get("nodes", []) + new_nodes
+            agent["links"] = agent.get("links", []) + new_links
+            logger.debug(f"Added {len(new_nodes)} nodes, {len(new_links)} links")
+
+        elif patch_type == "remove":
+            node_ids_to_remove = set(p.get("node_ids", []))
+            link_ids_to_remove = set(p.get("link_ids", []))
+
+            # Remove nodes
+            agent["nodes"] = [
+                n for n in agent.get("nodes", []) if n["id"] not in node_ids_to_remove
+            ]
+
+            # Remove links (both explicit and those referencing removed nodes)
+            agent["links"] = [
+                link
+                for link in agent.get("links", [])
+                if link["id"] not in link_ids_to_remove
+                and link["source_id"] not in node_ids_to_remove
+                and link["sink_id"] not in node_ids_to_remove
+            ]
+
+            logger.debug(
+                f"Removed {len(node_ids_to_remove)} nodes, {len(link_ids_to_remove)} links"
+            )
+
+    return agent
+
+
+def _deep_update(target: dict, source: dict) -> None:
+    """Recursively update a dict with another dict."""
+    for key, value in source.items():
+        if key in target and isinstance(target[key], dict) and isinstance(value, dict):
+            _deep_update(target[key], value)
+        else:
+            target[key] = value

autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/fixer.py

@@ -0,0 +1,606 @@
+"""Agent fixer - Fixes common LLM generation errors."""
+
+import logging
+import re
+import uuid
+from typing import Any
+
+from .utils import (
+    ADDTODICTIONARY_BLOCK_ID,
+    ADDTOLIST_BLOCK_ID,
+    CODE_EXECUTION_BLOCK_ID,
+    CONDITION_BLOCK_ID,
+    CREATEDICT_BLOCK_ID,
+    CREATELIST_BLOCK_ID,
+    DATA_SAMPLING_BLOCK_ID,
+    DOUBLE_CURLY_BRACES_BLOCK_IDS,
+    GET_CURRENT_DATE_BLOCK_ID,
+    STORE_VALUE_BLOCK_ID,
+    UNIVERSAL_TYPE_CONVERTER_BLOCK_ID,
+    get_blocks_info,
+    is_valid_uuid,
+)
+
+logger = logging.getLogger(__name__)
+
+
+def fix_agent_ids(agent: dict[str, Any]) -> dict[str, Any]:
+    """Fix invalid UUIDs in agent and link IDs."""
+    # Fix agent ID
+    if not is_valid_uuid(agent.get("id", "")):
+        agent["id"] = str(uuid.uuid4())
+        logger.debug(f"Fixed agent ID: {agent['id']}")
+
+    # Fix node IDs
+    id_mapping = {}  # Old ID -> New ID
+    for node in agent.get("nodes", []):
+        if not is_valid_uuid(node.get("id", "")):
+            old_id = node.get("id", "")
+            new_id = str(uuid.uuid4())
+            id_mapping[old_id] = new_id
+            node["id"] = new_id
+            logger.debug(f"Fixed node ID: {old_id} -> {new_id}")
+
+    # Fix link IDs and update references
+    for link in agent.get("links", []):
+        if not is_valid_uuid(link.get("id", "")):
+            link["id"] = str(uuid.uuid4())
+            logger.debug(f"Fixed link ID: {link['id']}")
+
+        # Update source/sink IDs if they were remapped
+        if link.get("source_id") in id_mapping:
+            link["source_id"] = id_mapping[link["source_id"]]
+        if link.get("sink_id") in id_mapping:
+            link["sink_id"] = id_mapping[link["sink_id"]]
+
+    return agent
+
+
+def fix_double_curly_braces(agent: dict[str, Any]) -> dict[str, Any]:
+    """Fix single curly braces to double in template blocks."""
+    for node in agent.get("nodes", []):
+        if node.get("block_id") not in DOUBLE_CURLY_BRACES_BLOCK_IDS:
+            continue
+
+        input_data = node.get("input_default", {})
+        for key in ("prompt", "format"):
+            if key in input_data and isinstance(input_data[key], str):
+                original = input_data[key]
+                # Fix simple variable references: {var} -> {{var}}
+                fixed = re.sub(
+                    r"(?<!\{)\{([a-zA-Z_][a-zA-Z0-9_]*)\}(?!\})",
+                    r"{{\1}}",
+                    original,
+                )
+                if fixed != original:
+                    input_data[key] = fixed
+                    logger.debug(f"Fixed curly braces in {key}")
+
+    return agent
+
+
+def fix_storevalue_before_condition(agent: dict[str, Any]) -> dict[str, Any]:
+    """Add StoreValueBlock before ConditionBlock if needed for value2."""
+    nodes = agent.get("nodes", [])
+    links = agent.get("links", [])
+
+    # Find all ConditionBlock nodes
+    condition_node_ids = {
+        node["id"] for node in nodes if node.get("block_id") == CONDITION_BLOCK_ID
+    }
+
+    if not condition_node_ids:
+        return agent
+
+    new_nodes = []
+    new_links = []
+    processed_conditions = set()
+
+    for link in links:
+        sink_id = link.get("sink_id")
+        sink_name = link.get("sink_name")
+
+        # Check if this link goes to a ConditionBlock's value2
+        if sink_id in condition_node_ids and sink_name == "value2":
+            source_node = next(
+                (n for n in nodes if n["id"] == link.get("source_id")), None
+            )
+
+            # Skip if source is already a StoreValueBlock
+            if source_node and source_node.get("block_id") == STORE_VALUE_BLOCK_ID:
+                continue
+
+            # Skip if we already processed this condition
+            if sink_id in processed_conditions:
+                continue
+
+            processed_conditions.add(sink_id)
+
+            # Create StoreValueBlock
+            store_node_id = str(uuid.uuid4())
+            store_node = {
+                "id": store_node_id,
+                "block_id": STORE_VALUE_BLOCK_ID,
+                "input_default": {"data": None},
+                "metadata": {"position": {"x": 0, "y": -100}},
+            }
+            new_nodes.append(store_node)
+
+            # Create link: original source -> StoreValueBlock
+            new_links.append(
+                {
+                    "id": str(uuid.uuid4()),
+                    "source_id": link["source_id"],
+                    "source_name": link["source_name"],
+                    "sink_id": store_node_id,
+                    "sink_name": "input",
+                    "is_static": False,
+                }
+            )
+
+            # Update original link: StoreValueBlock -> ConditionBlock
+            link["source_id"] = store_node_id
+            link["source_name"] = "output"
+
+            logger.debug(f"Added StoreValueBlock before ConditionBlock {sink_id}")
+
+    if new_nodes:
+        agent["nodes"] = nodes + new_nodes
+
+    return agent
+
+
+def fix_addtolist_blocks(agent: dict[str, Any]) -> dict[str, Any]:
+    """Fix AddToList blocks by adding prerequisite empty AddToList block.
+
+    When an AddToList block is found:
+    1. Checks if there's a CreateListBlock before it
+    2. Removes CreateListBlock if linked directly to AddToList
+    3. Adds an empty AddToList block before the original
+    4. Ensures the original has a self-referencing link
+    """
+    nodes = agent.get("nodes", [])
+    links = agent.get("links", [])
+    new_nodes = []
+    original_addtolist_ids = set()
+    nodes_to_remove = set()
+    links_to_remove = []
+
+    # First pass: identify CreateListBlock nodes to remove
+    for link in links:
+        source_node = next(
+            (n for n in nodes if n.get("id") == link.get("source_id")), None
+        )
+        sink_node = next((n for n in nodes if n.get("id") == link.get("sink_id")), None)
+
+        if (
+            source_node
+            and sink_node
+            and source_node.get("block_id") == CREATELIST_BLOCK_ID
+            and sink_node.get("block_id") == ADDTOLIST_BLOCK_ID
+        ):
+            nodes_to_remove.add(source_node.get("id"))
+            links_to_remove.append(link)
+            logger.debug(f"Removing CreateListBlock {source_node.get('id')}")
+
+    # Second pass: process AddToList blocks
+    filtered_nodes = []
+    for node in nodes:
+        if node.get("id") in nodes_to_remove:
+            continue
+
+        if node.get("block_id") == ADDTOLIST_BLOCK_ID:
+            original_addtolist_ids.add(node.get("id"))
+            node_id = node.get("id")
+            pos = node.get("metadata", {}).get("position", {"x": 0, "y": 0})
+
+            # Check if already has prerequisite
+            has_prereq = any(
+                link.get("sink_id") == node_id
+                and link.get("sink_name") == "list"
+                and link.get("source_name") == "updated_list"
+                for link in links
+            )
+
+            if not has_prereq:
+                # Remove links to "list" input (except self-reference)
+                for link in links:
+                    if (
+                        link.get("sink_id") == node_id
+                        and link.get("sink_name") == "list"
+                        and link.get("source_id") != node_id
+                        and link not in links_to_remove
+                    ):
+                        links_to_remove.append(link)
+
+                # Create prerequisite AddToList block
+                prereq_id = str(uuid.uuid4())
+                prereq_node = {
+                    "id": prereq_id,
+                    "block_id": ADDTOLIST_BLOCK_ID,
+                    "input_default": {"list": [], "entry": None, "entries": []},
+                    "metadata": {
+                        "position": {"x": pos.get("x", 0) - 800, "y": pos.get("y", 0)}
+                    },
+                }
+                new_nodes.append(prereq_node)
+
+                # Link prerequisite to original
+                links.append(
+                    {
+                        "id": str(uuid.uuid4()),
+                        "source_id": prereq_id,
+                        "source_name": "updated_list",
+                        "sink_id": node_id,
+                        "sink_name": "list",
+                        "is_static": False,
+                    }
+                )
+                logger.debug(f"Added prerequisite AddToList block for {node_id}")
+
+        filtered_nodes.append(node)
+
+    # Remove marked links
+    filtered_links = [link for link in links if link not in links_to_remove]
+
+    # Add self-referencing links for original AddToList blocks
+    for node in filtered_nodes + new_nodes:
+        if (
+            node.get("block_id") == ADDTOLIST_BLOCK_ID
+            and node.get("id") in original_addtolist_ids
+        ):
+            node_id = node.get("id")
+            has_self_ref = any(
+                link["source_id"] == node_id
+                and link["sink_id"] == node_id
+                and link["source_name"] == "updated_list"
+                and link["sink_name"] == "list"
+                for link in filtered_links
+            )
+            if not has_self_ref:
+                filtered_links.append(
+                    {
+                        "id": str(uuid.uuid4()),
+                        "source_id": node_id,
+                        "source_name": "updated_list",
+                        "sink_id": node_id,
+                        "sink_name": "list",
+                        "is_static": False,
+                    }
+                )
+                logger.debug(f"Added self-reference for AddToList {node_id}")
+
+    agent["nodes"] = filtered_nodes + new_nodes
+    agent["links"] = filtered_links
+    return agent
+
+
+def fix_addtodictionary_blocks(agent: dict[str, Any]) -> dict[str, Any]:
+    """Fix AddToDictionary blocks by removing empty CreateDictionary nodes."""
+    nodes = agent.get("nodes", [])
+    links = agent.get("links", [])
+    nodes_to_remove = set()
+    links_to_remove = []
+
+    for link in links:
+        source_node = next(
+            (n for n in nodes if n.get("id") == link.get("source_id")), None
+        )
+        sink_node = next((n for n in nodes if n.get("id") == link.get("sink_id")), None)
+
+        if (
+            source_node
+            and sink_node
+            and source_node.get("block_id") == CREATEDICT_BLOCK_ID
+            and sink_node.get("block_id") == ADDTODICTIONARY_BLOCK_ID
+        ):
+            nodes_to_remove.add(source_node.get("id"))
+            links_to_remove.append(link)
+            logger.debug(f"Removing CreateDictionary {source_node.get('id')}")
+
+    agent["nodes"] = [n for n in nodes if n.get("id") not in nodes_to_remove]
+    agent["links"] = [link for link in links if link not in links_to_remove]
+    return agent
+
+
+def fix_code_execution_output(agent: dict[str, Any]) -> dict[str, Any]:
+    """Fix CodeExecutionBlock output: change 'response' to 'stdout_logs'."""
+    nodes = agent.get("nodes", [])
+    links = agent.get("links", [])
+
+    for link in links:
+        source_node = next(
+            (n for n in nodes if n.get("id") == link.get("source_id")), None
+        )
+        if (
+            source_node
+            and source_node.get("block_id") == CODE_EXECUTION_BLOCK_ID
+            and link.get("source_name") == "response"
+        ):
+            link["source_name"] = "stdout_logs"
+            logger.debug("Fixed CodeExecutionBlock output: response -> stdout_logs")
+
+    return agent
+
+
+def fix_data_sampling_sample_size(agent: dict[str, Any]) -> dict[str, Any]:
+    """Fix DataSamplingBlock by setting sample_size to 1 as default."""
+    nodes = agent.get("nodes", [])
+    links = agent.get("links", [])
+    links_to_remove = []
+
+    for node in nodes:
+        if node.get("block_id") == DATA_SAMPLING_BLOCK_ID:
+            node_id = node.get("id")
+            input_default = node.get("input_default", {})
+
+            # Remove links to sample_size
+            for link in links:
+                if (
+                    link.get("sink_id") == node_id
+                    and link.get("sink_name") == "sample_size"
+                ):
+                    links_to_remove.append(link)
+
+            # Set default
+            input_default["sample_size"] = 1
+            node["input_default"] = input_default
+            logger.debug(f"Fixed DataSamplingBlock {node_id} sample_size to 1")
+
+    if links_to_remove:
+        agent["links"] = [link for link in links if link not in links_to_remove]
+
+    return agent
+
+
+def fix_node_x_coordinates(agent: dict[str, Any]) -> dict[str, Any]:
+    """Fix node x-coordinates to ensure 800+ unit spacing between linked nodes."""
+    nodes = agent.get("nodes", [])
+    links = agent.get("links", [])
+    node_lookup = {n.get("id"): n for n in nodes}
+
+    for link in links:
+        source_id = link.get("source_id")
+        sink_id = link.get("sink_id")
+
+        source_node = node_lookup.get(source_id)
+        sink_node = node_lookup.get(sink_id)
+
+        if not source_node or not sink_node:
+            continue
+
+        source_pos = source_node.get("metadata", {}).get("position", {})
+        sink_pos = sink_node.get("metadata", {}).get("position", {})
+
+        source_x = source_pos.get("x", 0)
+        sink_x = sink_pos.get("x", 0)
+
+        if abs(sink_x - source_x) < 800:
+            new_x = source_x + 800
+            if "metadata" not in sink_node:
+                sink_node["metadata"] = {}
+            if "position" not in sink_node["metadata"]:
+                sink_node["metadata"]["position"] = {}
+            sink_node["metadata"]["position"]["x"] = new_x
+            logger.debug(f"Fixed node {sink_id} x: {sink_x} -> {new_x}")
+
+    return agent
+
+
+def fix_getcurrentdate_offset(agent: dict[str, Any]) -> dict[str, Any]:
+    """Fix GetCurrentDateBlock offset to ensure it's positive."""
+    for node in agent.get("nodes", []):
+        if node.get("block_id") == GET_CURRENT_DATE_BLOCK_ID:
+            input_default = node.get("input_default", {})
+            if "offset" in input_default:
+                offset = input_default["offset"]
+                if isinstance(offset, (int, float)) and offset < 0:
+                    input_default["offset"] = abs(offset)
+                    logger.debug(f"Fixed offset: {offset} -> {abs(offset)}")
+
+    return agent
+
+
+def fix_ai_model_parameter(
+    agent: dict[str, Any],
+    blocks_info: list[dict[str, Any]],
+    default_model: str = "gpt-4o",
+) -> dict[str, Any]:
+    """Add default model parameter to AI blocks if missing."""
+    block_map = {b.get("id"): b for b in blocks_info}
+
+    for node in agent.get("nodes", []):
+        block_id = node.get("block_id")
+        block = block_map.get(block_id)
+
+        if not block:
+            continue
+
+        # Check if block has AI category
+        categories = block.get("categories", [])
+        is_ai_block = any(
+            cat.get("category") == "AI" for cat in categories if isinstance(cat, dict)
+        )
+
+        if is_ai_block:
+            input_default = node.get("input_default", {})
+            if "model" not in input_default:
+                input_default["model"] = default_model
+                node["input_default"] = input_default
+                logger.debug(
+                    f"Added model '{default_model}' to AI block {node.get('id')}"
+                )
+
+    return agent
+
+
+def fix_link_static_properties(
+    agent: dict[str, Any], blocks_info: list[dict[str, Any]]
+) -> dict[str, Any]:
+    """Fix is_static property based on source block's staticOutput."""
+    block_map = {b.get("id"): b for b in blocks_info}
+    node_lookup = {n.get("id"): n for n in agent.get("nodes", [])}
+
+    for link in agent.get("links", []):
+        source_node = node_lookup.get(link.get("source_id"))
+        if not source_node:
+            continue
+
+        source_block = block_map.get(source_node.get("block_id"))
+        if not source_block:
+            continue
+
+        static_output = source_block.get("staticOutput", False)
+        if link.get("is_static") != static_output:
+            link["is_static"] = static_output
+            logger.debug(f"Fixed link {link.get('id')} is_static to {static_output}")
+
+    return agent
+
+
+def fix_data_type_mismatch(
+    agent: dict[str, Any], blocks_info: list[dict[str, Any]]
+) -> dict[str, Any]:
+    """Fix data type mismatches by inserting UniversalTypeConverterBlock."""
+    nodes = agent.get("nodes", [])
+    links = agent.get("links", [])
+    block_map = {b.get("id"): b for b in blocks_info}
+    node_lookup = {n.get("id"): n for n in nodes}
+
+    def get_property_type(schema: dict, name: str) -> str | None:
+        if "_#_" in name:
+            parent, child = name.split("_#_", 1)
+            parent_schema = schema.get(parent, {})
+            if "properties" in parent_schema:
+                return parent_schema["properties"].get(child, {}).get("type")
+            return None
+        return schema.get(name, {}).get("type")
+
+    def are_types_compatible(src: str, sink: str) -> bool:
+        if {src, sink} <= {"integer", "number"}:
+            return True
+        return src == sink
+
+    type_mapping = {
+        "string": "string",
+        "text": "string",
+        "integer": "number",
+        "number": "number",
+        "float": "number",
+        "boolean": "boolean",
+        "bool": "boolean",
+        "array": "list",
+        "list": "list",
+        "object": "dictionary",
+        "dict": "dictionary",
+        "dictionary": "dictionary",
+    }
+
+    new_links = []
+    nodes_to_add = []
+
+    for link in links:
+        source_node = node_lookup.get(link.get("source_id"))
+        sink_node = node_lookup.get(link.get("sink_id"))
+
+        if not source_node or not sink_node:
+            new_links.append(link)
+            continue
+
+        source_block = block_map.get(source_node.get("block_id"))
+        sink_block = block_map.get(sink_node.get("block_id"))
+
+        if not source_block or not sink_block:
+            new_links.append(link)
+            continue
+
+        source_outputs = source_block.get("outputSchema", {}).get("properties", {})
+        sink_inputs = sink_block.get("inputSchema", {}).get("properties", {})
+
+        source_type = get_property_type(source_outputs, link.get("source_name", ""))
+        sink_type = get_property_type(sink_inputs, link.get("sink_name", ""))
+
+        if (
+            source_type
+            and sink_type
+            and not are_types_compatible(source_type, sink_type)
+        ):
+            # Insert type converter
+            converter_id = str(uuid.uuid4())
+            target_type = type_mapping.get(sink_type, sink_type)
+
+            converter_node = {
+                "id": converter_id,
+                "block_id": UNIVERSAL_TYPE_CONVERTER_BLOCK_ID,
+                "input_default": {"type": target_type},
+                "metadata": {"position": {"x": 0, "y": 100}},
+            }
+            nodes_to_add.append(converter_node)
+
+            # source -> converter
+            new_links.append(
+                {
+                    "id": str(uuid.uuid4()),
+                    "source_id": link["source_id"],
+                    "source_name": link["source_name"],
+                    "sink_id": converter_id,
+                    "sink_name": "value",
+                    "is_static": False,
+                }
+            )
+
+            # converter -> sink
+            new_links.append(
+                {
+                    "id": str(uuid.uuid4()),
+                    "source_id": converter_id,
+                    "source_name": "value",
+                    "sink_id": link["sink_id"],
+                    "sink_name": link["sink_name"],
+                    "is_static": False,
+                }
+            )
+
+            logger.debug(f"Inserted type converter: {source_type} -> {target_type}")
+        else:
+            new_links.append(link)
+
+    if nodes_to_add:
+        agent["nodes"] = nodes + nodes_to_add
+        agent["links"] = new_links
+
+    return agent
+
+
+def apply_all_fixes(
+    agent: dict[str, Any], blocks_info: list[dict[str, Any]] | None = None
+) -> dict[str, Any]:
+    """Apply all fixes to an agent JSON.
+
+    Args:
+        agent: Agent JSON dict
+        blocks_info: Optional list of block info dicts for advanced fixes
+
+    Returns:
+        Fixed agent JSON
+    """
+    # Basic fixes (no block info needed)
+    agent = fix_agent_ids(agent)
+    agent = fix_double_curly_braces(agent)
+    agent = fix_storevalue_before_condition(agent)
+    agent = fix_addtolist_blocks(agent)
+    agent = fix_addtodictionary_blocks(agent)
+    agent = fix_code_execution_output(agent)
+    agent = fix_data_sampling_sample_size(agent)
+    agent = fix_node_x_coordinates(agent)
+    agent = fix_getcurrentdate_offset(agent)
+
+    # Advanced fixes (require block info)
+    if blocks_info is None:
+        blocks_info = get_blocks_info()
+
+    agent = fix_ai_model_parameter(agent, blocks_info)
+    agent = fix_link_static_properties(agent, blocks_info)
+    agent = fix_data_type_mismatch(agent, blocks_info)
+
+    return agent

autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/prompts.py

@@ -0,0 +1,225 @@
+"""Prompt templates for agent generation."""
+
+DECOMPOSITION_PROMPT = """
+You are an expert AutoGPT Workflow Decomposer. Your task is to analyze a user's high-level goal and break it down into a clear, step-by-step plan using the available blocks.
+
+Each step should represent a distinct, automatable action suitable for execution by an AI automation system.
+
+---
+
+FIRST: Analyze the user's goal and determine:
+1) Design-time configuration (fixed settings that won't change per run)
+2) Runtime inputs (values the agent's end-user will provide each time it runs)
+
+For anything that can vary per run (email addresses, names, dates, search terms, etc.):
+- DO NOT ask for the actual value
+- Instead, define it as an Agent Input with a clear name, type, and description
+
+Only ask clarifying questions about design-time config that affects how you build the workflow:
+- Which external service to use (e.g., "Gmail vs Outlook", "Notion vs Google Docs")
+- Required formats or structures (e.g., "CSV, JSON, or PDF output?")
+- Business rules that must be hard-coded
+
+IMPORTANT CLARIFICATIONS POLICY:
+- Ask no more than five essential questions
+- Do not ask for concrete values that can be provided at runtime as Agent Inputs
+- Do not ask for API keys or credentials; the platform handles those directly
+- If there is enough information to infer reasonable defaults, prefer to propose defaults
+
+---
+
+GUIDELINES:
+1. List each step as a numbered item
+2. Describe the action clearly and specify inputs/outputs
+3. Ensure steps are in logical, sequential order
+4. Mention block names naturally (e.g., "Use GetWeatherByLocationBlock to...")
+5. Help the user reach their goal efficiently
+
+---
+
+RULES:
+1. OUTPUT FORMAT: Only output either clarifying questions OR step-by-step instructions, not both
+2. USE ONLY THE BLOCKS PROVIDED
+3. ALL required_input fields must be provided
+4. Data types of linked properties must match
+5. Write expert-level prompts for AI-related blocks
+
+---
+
+CRITICAL BLOCK RESTRICTIONS:
+1. AddToListBlock: Outputs updated list EVERY addition, not after all additions
+2. SendEmailBlock: Draft the email for user review; set SMTP config based on email type
+3. ConditionBlock: value2 is reference, value1 is contrast
+4. CodeExecutionBlock: DO NOT USE - use AI blocks instead
+5. ReadCsvBlock: Only use the 'rows' output, not 'row'
+
+---
+
+OUTPUT FORMAT:
+
+If more information is needed:
+```json
+{{
+  "type": "clarifying_questions",
+  "questions": [
+    {{
+      "question": "Which email provider should be used? (Gmail, Outlook, custom SMTP)",
+      "keyword": "email_provider",
+      "example": "Gmail"
+    }}
+  ]
+}}
+```
+
+If ready to proceed:
+```json
+{{
+  "type": "instructions",
+  "steps": [
+    {{
+      "step_number": 1,
+      "block_name": "AgentShortTextInputBlock",
+      "description": "Get the URL of the content to analyze.",
+      "inputs": [{{"name": "name", "value": "URL"}}],
+      "outputs": [{{"name": "result", "description": "The URL entered by user"}}]
+    }}
+  ]
+}}
+```
+
+---
+
+AVAILABLE BLOCKS:
+{block_summaries}
+"""
+
+GENERATION_PROMPT = """
+You are an expert AI workflow builder. Generate a valid agent JSON from the given instructions.
+
+---
+
+NODES:
+Each node must include:
+- `id`: Unique UUID v4 (e.g. `a8f5b1e2-c3d4-4e5f-8a9b-0c1d2e3f4a5b`)
+- `block_id`: The block identifier (must match an Allowed Block)
+- `input_default`: Dict of inputs (can be empty if no static inputs needed)
+- `metadata`: Must contain:
+  - `position`: {{"x": number, "y": number}} - adjacent nodes should differ by 800+ in X
+  - `customized_name`: Clear name describing this block's purpose in the workflow
+
+---
+
+LINKS:
+Each link connects a source node's output to a sink node's input:
+- `id`: MUST be UUID v4 (NOT "link-1", "link-2", etc.)
+- `source_id`: ID of the source node
+- `source_name`: Output field name from the source block
+- `sink_id`: ID of the sink node
+- `sink_name`: Input field name on the sink block
+- `is_static`: true only if source block has static_output: true
+
+CRITICAL: All IDs must be valid UUID v4 format!
+
+---
+
+AGENT (GRAPH):
+Wrap nodes and links in:
+- `id`: UUID of the agent
+- `name`: Short, generic name (avoid specific company names, URLs)
+- `description`: Short, generic description
+- `nodes`: List of all nodes
+- `links`: List of all links
+- `version`: 1
+- `is_active`: true
+
+---
+
+TIPS:
+- All required_input fields must be provided via input_default or a valid link
+- Ensure consistent source_id and sink_id references
+- Avoid dangling links
+- Input/output pins must match block schemas
+- Do not invent unknown block_ids
+
+---
+
+ALLOWED BLOCKS:
+{block_summaries}
+
+---
+
+Generate the complete agent JSON. Output ONLY valid JSON, no explanation.
+"""
+
+PATCH_PROMPT = """
+You are an expert at modifying AutoGPT agent workflows. Given the current agent and a modification request, generate a JSON patch to update the agent.
+
+CURRENT AGENT:
+{current_agent}
+
+AVAILABLE BLOCKS:
+{block_summaries}
+
+---
+
+PATCH FORMAT:
+Return a JSON object with the following structure:
+
+```json
+{{
+  "type": "patch",
+  "intent": "Brief description of what the patch does",
+  "patches": [
+    {{
+      "type": "modify",
+      "node_id": "uuid-of-node-to-modify",
+      "changes": {{
+        "input_default": {{"field": "new_value"}},
+        "metadata": {{"customized_name": "New Name"}}
+      }}
+    }},
+    {{
+      "type": "add",
+      "new_nodes": [
+        {{
+          "id": "new-uuid",
+          "block_id": "block-uuid",
+          "input_default": {{}},
+          "metadata": {{"position": {{"x": 0, "y": 0}}, "customized_name": "Name"}}
+        }}
+      ],
+      "new_links": [
+        {{
+          "id": "link-uuid",
+          "source_id": "source-node-id",
+          "source_name": "output_field",
+          "sink_id": "sink-node-id",
+          "sink_name": "input_field"
+        }}
+      ]
+    }},
+    {{
+      "type": "remove",
+      "node_ids": ["uuid-of-node-to-remove"],
+      "link_ids": ["uuid-of-link-to-remove"]
+    }}
+  ]
+}}
+```
+
+If you need more information, return:
+```json
+{{
+  "type": "clarifying_questions",
+  "questions": [
+    {{
+      "question": "What specific change do you want?",
+      "keyword": "change_type",
+      "example": "Add error handling"
+    }}
+  ]
+}}
+```
+
+Generate the minimal patch needed. Output ONLY valid JSON.
+"""

autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/utils.py

@@ -0,0 +1,213 @@
+"""Utilities for agent generation."""
+
+import json
+import re
+from typing import Any
+
+from backend.data.block import get_blocks
+
+# UUID validation regex
+UUID_REGEX = re.compile(
+    r"^[a-f0-9]{8}-[a-f0-9]{4}-4[a-f0-9]{3}-[89ab][a-f0-9]{3}-[a-f0-9]{12}$"
+)
+
+# Block IDs for various fixes
+STORE_VALUE_BLOCK_ID = "1ff065e9-88e8-4358-9d82-8dc91f622ba9"
+CONDITION_BLOCK_ID = "715696a0-e1da-45c8-b209-c2fa9c3b0be6"
+ADDTOLIST_BLOCK_ID = "aeb08fc1-2fc1-4141-bc8e-f758f183a822"
+ADDTODICTIONARY_BLOCK_ID = "31d1064e-7446-4693-a7d4-65e5ca1180d1"
+CREATELIST_BLOCK_ID = "a912d5c7-6e00-4542-b2a9-8034136930e4"
+CREATEDICT_BLOCK_ID = "b924ddf4-de4f-4b56-9a85-358930dcbc91"
+CODE_EXECUTION_BLOCK_ID = "0b02b072-abe7-11ef-8372-fb5d162dd712"
+DATA_SAMPLING_BLOCK_ID = "4a448883-71fa-49cf-91cf-70d793bd7d87"
+UNIVERSAL_TYPE_CONVERTER_BLOCK_ID = "95d1b990-ce13-4d88-9737-ba5c2070c97b"
+GET_CURRENT_DATE_BLOCK_ID = "b29c1b50-5d0e-4d9f-8f9d-1b0e6fcbf0b1"
+
+DOUBLE_CURLY_BRACES_BLOCK_IDS = [
+    "44f6c8ad-d75c-4ae1-8209-aad1c0326928",  # FillTextTemplateBlock
+    "6ab085e2-20b3-4055-bc3e-08036e01eca6",
+    "90f8c45e-e983-4644-aa0b-b4ebe2f531bc",
+    "363ae599-353e-4804-937e-b2ee3cef3da4",  # AgentOutputBlock
+    "3b191d9f-356f-482d-8238-ba04b6d18381",
+    "db7d8f02-2f44-4c55-ab7a-eae0941f0c30",
+    "3a7c4b8d-6e2f-4a5d-b9c1-f8d23c5a9b0e",
+    "ed1ae7a0-b770-4089-b520-1f0005fad19a",
+    "a892b8d9-3e4e-4e9c-9c1e-75f8efcf1bfa",
+    "b29c1b50-5d0e-4d9f-8f9d-1b0e6fcbf0b1",
+    "716a67b3-6760-42e7-86dc-18645c6e00fc",
+    "530cf046-2ce0-4854-ae2c-659db17c7a46",
+    "ed55ac19-356e-4243-a6cb-bc599e9b716f",
+    "1f292d4a-41a4-4977-9684-7c8d560b9f91",  # LLM blocks
+    "32a87eab-381e-4dd4-bdb8-4c47151be35a",
+]
+
+
+def is_valid_uuid(value: str) -> bool:
+    """Check if a string is a valid UUID v4."""
+    return isinstance(value, str) and UUID_REGEX.match(value) is not None
+
+
+def _compact_schema(schema: dict) -> dict[str, str]:
+    """Extract compact type info from a JSON schema properties dict.
+
+    Returns a dict of {field_name: type_string} for essential info only.
+    """
+    props = schema.get("properties", {})
+    result = {}
+
+    for name, prop in props.items():
+        # Skip internal/complex fields
+        if name.startswith("_"):
+            continue
+
+        # Get type string
+        type_str = prop.get("type", "any")
+
+        # Handle anyOf/oneOf (optional types)
+        if "anyOf" in prop:
+            types = [t.get("type", "?") for t in prop["anyOf"] if t.get("type")]
+            type_str = "|".join(types) if types else "any"
+        elif "allOf" in prop:
+            type_str = "object"
+
+        # Add array item type if present
+        if type_str == "array" and "items" in prop:
+            items = prop["items"]
+            if isinstance(items, dict):
+                item_type = items.get("type", "any")
+                type_str = f"array[{item_type}]"
+
+        result[name] = type_str
+
+    return result
+
+
+def get_block_summaries(include_schemas: bool = True) -> str:
+    """Generate compact block summaries for prompts.
+
+    Args:
+        include_schemas: Whether to include input/output type info
+
+    Returns:
+        Formatted string of block summaries (compact format)
+    """
+    blocks = get_blocks()
+    summaries = []
+
+    for block_id, block_cls in blocks.items():
+        block = block_cls()
+        name = block.name
+        desc = getattr(block, "description", "") or ""
+
+        # Truncate description
+        if len(desc) > 150:
+            desc = desc[:147] + "..."
+
+        if not include_schemas:
+            summaries.append(f"- {name} (id: {block_id}): {desc}")
+        else:
+            # Compact format with type info only
+            inputs = {}
+            outputs = {}
+            required = []
+
+            if hasattr(block, "input_schema"):
+                try:
+                    schema = block.input_schema.jsonschema()
+                    inputs = _compact_schema(schema)
+                    required = schema.get("required", [])
+                except Exception:
+                    pass
+
+            if hasattr(block, "output_schema"):
+                try:
+                    schema = block.output_schema.jsonschema()
+                    outputs = _compact_schema(schema)
+                except Exception:
+                    pass
+
+            # Build compact line format
+            # Format: NAME (id): desc | in: {field:type, ...} [required] | out: {field:type}
+            in_str = ", ".join(f"{k}:{v}" for k, v in inputs.items())
+            out_str = ", ".join(f"{k}:{v}" for k, v in outputs.items())
+            req_str = f" req=[{','.join(required)}]" if required else ""
+
+            static = " [static]" if getattr(block, "static_output", False) else ""
+
+            line = f"- {name} (id: {block_id}): {desc}"
+            if in_str:
+                line += f"\n  in: {{{in_str}}}{req_str}"
+            if out_str:
+                line += f"\n  out: {{{out_str}}}{static}"
+
+            summaries.append(line)
+
+    return "\n".join(summaries)
+
+
+def get_blocks_info() -> list[dict[str, Any]]:
+    """Get block information with schemas for validation and fixing."""
+    blocks = get_blocks()
+    blocks_info = []
+    for block_id, block_cls in blocks.items():
+        block = block_cls()
+        blocks_info.append(
+            {
+                "id": block_id,
+                "name": block.name,
+                "description": getattr(block, "description", ""),
+                "categories": getattr(block, "categories", []),
+                "staticOutput": getattr(block, "static_output", False),
+                "inputSchema": (
+                    block.input_schema.jsonschema()
+                    if hasattr(block, "input_schema")
+                    else {}
+                ),
+                "outputSchema": (
+                    block.output_schema.jsonschema()
+                    if hasattr(block, "output_schema")
+                    else {}
+                ),
+            }
+        )
+    return blocks_info
+
+
+def parse_json_from_llm(text: str) -> dict[str, Any] | None:
+    """Extract JSON from LLM response (handles markdown code blocks)."""
+    if not text:
+        return None
+
+    # Try fenced code block
+    match = re.search(r"```(?:json)?\s*([\s\S]*?)```", text, re.IGNORECASE)
+    if match:
+        try:
+            return json.loads(match.group(1).strip())
+        except json.JSONDecodeError:
+            pass
+
+    # Try raw text
+    try:
+        return json.loads(text.strip())
+    except json.JSONDecodeError:
+        pass
+
+    # Try finding {...} span
+    start = text.find("{")
+    end = text.rfind("}")
+    if start != -1 and end > start:
+        try:
+            return json.loads(text[start : end + 1])
+        except json.JSONDecodeError:
+            pass
+
+    # Try finding [...] span
+    start = text.find("[")
+    end = text.rfind("]")
+    if start != -1 and end > start:
+        try:
+            return json.loads(text[start : end + 1])
+        except json.JSONDecodeError:
+            pass
+
+    return None

autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/validator.py

@@ -0,0 +1,279 @@
+"""Agent validator - Validates agent structure and connections."""
+
+import logging
+import re
+from typing import Any
+
+from .utils import get_blocks_info
+
+logger = logging.getLogger(__name__)
+
+
+class AgentValidator:
+    """Validator for AutoGPT agents with detailed error reporting."""
+
+    def __init__(self):
+        self.errors: list[str] = []
+
+    def add_error(self, error: str) -> None:
+        """Add an error message."""
+        self.errors.append(error)
+
+    def validate_block_existence(
+        self, agent: dict[str, Any], blocks_info: list[dict[str, Any]]
+    ) -> bool:
+        """Validate all block IDs exist in the blocks library."""
+        valid = True
+        valid_block_ids = {b.get("id") for b in blocks_info if b.get("id")}
+
+        for node in agent.get("nodes", []):
+            block_id = node.get("block_id")
+            node_id = node.get("id")
+
+            if not block_id:
+                self.add_error(f"Node '{node_id}' is missing 'block_id' field.")
+                valid = False
+                continue
+
+            if block_id not in valid_block_ids:
+                self.add_error(
+                    f"Node '{node_id}' references block_id '{block_id}' which does not exist."
+                )
+                valid = False
+
+        return valid
+
+    def validate_link_node_references(self, agent: dict[str, Any]) -> bool:
+        """Validate all node IDs referenced in links exist."""
+        valid = True
+        valid_node_ids = {n.get("id") for n in agent.get("nodes", []) if n.get("id")}
+
+        for link in agent.get("links", []):
+            link_id = link.get("id", "Unknown")
+            source_id = link.get("source_id")
+            sink_id = link.get("sink_id")
+
+            if not source_id:
+                self.add_error(f"Link '{link_id}' is missing 'source_id'.")
+                valid = False
+            elif source_id not in valid_node_ids:
+                self.add_error(
+                    f"Link '{link_id}' references non-existent source_id '{source_id}'."
+                )
+                valid = False
+
+            if not sink_id:
+                self.add_error(f"Link '{link_id}' is missing 'sink_id'.")
+                valid = False
+            elif sink_id not in valid_node_ids:
+                self.add_error(
+                    f"Link '{link_id}' references non-existent sink_id '{sink_id}'."
+                )
+                valid = False
+
+        return valid
+
+    def validate_required_inputs(
+        self, agent: dict[str, Any], blocks_info: list[dict[str, Any]]
+    ) -> bool:
+        """Validate required inputs are provided."""
+        valid = True
+        block_map = {b.get("id"): b for b in blocks_info}
+
+        for node in agent.get("nodes", []):
+            block_id = node.get("block_id")
+            block = block_map.get(block_id)
+
+            if not block:
+                continue
+
+            required_inputs = block.get("inputSchema", {}).get("required", [])
+            input_defaults = node.get("input_default", {})
+            node_id = node.get("id")
+
+            # Get linked inputs
+            linked_inputs = {
+                link["sink_name"]
+                for link in agent.get("links", [])
+                if link.get("sink_id") == node_id
+            }
+
+            for req_input in required_inputs:
+                if (
+                    req_input not in input_defaults
+                    and req_input not in linked_inputs
+                    and req_input != "credentials"
+                ):
+                    block_name = block.get("name", "Unknown Block")
+                    self.add_error(
+                        f"Node '{node_id}' ({block_name}) is missing required input '{req_input}'."
+                    )
+                    valid = False
+
+        return valid
+
+    def validate_data_type_compatibility(
+        self, agent: dict[str, Any], blocks_info: list[dict[str, Any]]
+    ) -> bool:
+        """Validate linked data types are compatible."""
+        valid = True
+        block_map = {b.get("id"): b for b in blocks_info}
+        node_lookup = {n.get("id"): n for n in agent.get("nodes", [])}
+
+        def get_type(schema: dict, name: str) -> str | None:
+            if "_#_" in name:
+                parent, child = name.split("_#_", 1)
+                parent_schema = schema.get(parent, {})
+                if "properties" in parent_schema:
+                    return parent_schema["properties"].get(child, {}).get("type")
+                return None
+            return schema.get(name, {}).get("type")
+
+        def are_compatible(src: str, sink: str) -> bool:
+            if {src, sink} <= {"integer", "number"}:
+                return True
+            return src == sink
+
+        for link in agent.get("links", []):
+            source_node = node_lookup.get(link.get("source_id"))
+            sink_node = node_lookup.get(link.get("sink_id"))
+
+            if not source_node or not sink_node:
+                continue
+
+            source_block = block_map.get(source_node.get("block_id"))
+            sink_block = block_map.get(sink_node.get("block_id"))
+
+            if not source_block or not sink_block:
+                continue
+
+            source_outputs = source_block.get("outputSchema", {}).get("properties", {})
+            sink_inputs = sink_block.get("inputSchema", {}).get("properties", {})
+
+            source_type = get_type(source_outputs, link.get("source_name", ""))
+            sink_type = get_type(sink_inputs, link.get("sink_name", ""))
+
+            if source_type and sink_type and not are_compatible(source_type, sink_type):
+                self.add_error(
+                    f"Type mismatch: {source_block.get('name')} output '{link['source_name']}' "
+                    f"({source_type}) -> {sink_block.get('name')} input '{link['sink_name']}' ({sink_type})."
+                )
+                valid = False
+
+        return valid
+
+    def validate_nested_sink_links(
+        self, agent: dict[str, Any], blocks_info: list[dict[str, Any]]
+    ) -> bool:
+        """Validate nested sink links (with _#_ notation)."""
+        valid = True
+        block_map = {b.get("id"): b for b in blocks_info}
+        node_lookup = {n.get("id"): n for n in agent.get("nodes", [])}
+
+        for link in agent.get("links", []):
+            sink_name = link.get("sink_name", "")
+
+            if "_#_" in sink_name:
+                parent, child = sink_name.split("_#_", 1)
+
+                sink_node = node_lookup.get(link.get("sink_id"))
+                if not sink_node:
+                    continue
+
+                block = block_map.get(sink_node.get("block_id"))
+                if not block:
+                    continue
+
+                input_props = block.get("inputSchema", {}).get("properties", {})
+                parent_schema = input_props.get(parent)
+
+                if not parent_schema:
+                    self.add_error(
+                        f"Invalid nested link '{sink_name}': parent '{parent}' not found."
+                    )
+                    valid = False
+                    continue
+
+                if not parent_schema.get("additionalProperties"):
+                    if not (
+                        isinstance(parent_schema, dict)
+                        and "properties" in parent_schema
+                        and child in parent_schema.get("properties", {})
+                    ):
+                        self.add_error(
+                            f"Invalid nested link '{sink_name}': child '{child}' not found in '{parent}'."
+                        )
+                        valid = False
+
+        return valid
+
+    def validate_prompt_spaces(self, agent: dict[str, Any]) -> bool:
+        """Validate prompts don't have spaces in template variables."""
+        valid = True
+
+        for node in agent.get("nodes", []):
+            input_default = node.get("input_default", {})
+            prompt = input_default.get("prompt", "")
+
+            if not isinstance(prompt, str):
+                continue
+
+            # Find {{...}} with spaces
+            matches = re.finditer(r"\{\{([^}]+)\}\}", prompt)
+            for match in matches:
+                content = match.group(1)
+                if " " in content:
+                    self.add_error(
+                        f"Node '{node.get('id')}' has spaces in template variable: "
+                        f"'{{{{{content}}}}}' should be '{{{{{content.replace(' ', '_')}}}}}'."
+                    )
+                    valid = False
+
+        return valid
+
+    def validate(
+        self, agent: dict[str, Any], blocks_info: list[dict[str, Any]] | None = None
+    ) -> tuple[bool, str | None]:
+        """Run all validations.
+
+        Returns:
+            Tuple of (is_valid, error_message)
+        """
+        self.errors = []
+
+        if blocks_info is None:
+            blocks_info = get_blocks_info()
+
+        checks = [
+            self.validate_block_existence(agent, blocks_info),
+            self.validate_link_node_references(agent),
+            self.validate_required_inputs(agent, blocks_info),
+            self.validate_data_type_compatibility(agent, blocks_info),
+            self.validate_nested_sink_links(agent, blocks_info),
+            self.validate_prompt_spaces(agent),
+        ]
+
+        all_passed = all(checks)
+
+        if all_passed:
+            logger.info("Agent validation successful")
+            return True, None
+
+        error_message = "Agent validation failed:\n"
+        for i, error in enumerate(self.errors, 1):
+            error_message += f"{i}. {error}\n"
+
+        logger.warning(f"Agent validation failed with {len(self.errors)} errors")
+        return False, error_message
+
+
+def validate_agent(
+    agent: dict[str, Any], blocks_info: list[dict[str, Any]] | None = None
+) -> tuple[bool, str | None]:
+    """Convenience function to validate an agent.
+
+    Returns:
+        Tuple of (is_valid, error_message)
+    """
+    validator = AgentValidator()
+    return validator.validate(agent, blocks_info)

autogpt_platform/backend/backend/api/features/chat/tools/agent_output.py

@@ -55,47 +55,56 @@ def parse_time_expression(
     """
     Parse time expression into datetime range (start, end).
 
-    Supports: "latest", "yesterday", "today", "last week", "last 7 days",
-    "last month", "last 30 days", ISO date "YYYY-MM-DD", ISO datetime.
+    Supports:
+    - "latest" or None -> returns (None, None) to get most recent
+    - "yesterday" -> 24h window for yesterday
+    - "today" -> Today from midnight
+    - "last week" / "last 7 days" -> 7 day window
+    - "last month" / "last 30 days" -> 30 day window
+    - ISO date "YYYY-MM-DD" -> 24h window for that date
     """
     if not time_expr or time_expr.lower() == "latest":
         return None, None
 
     now = datetime.now(timezone.utc)
-    today_start = now.replace(hour=0, minute=0, second=0, microsecond=0)
     expr = time_expr.lower().strip()
 
-    # Relative time expressions lookup
-    relative_times: dict[str, tuple[datetime, datetime]] = {
-        "yesterday": (today_start - timedelta(days=1), today_start),
-        "today": (today_start, now),
-        "last week": (now - timedelta(days=7), now),
-        "last 7 days": (now - timedelta(days=7), now),
-        "last month": (now - timedelta(days=30), now),
-        "last 30 days": (now - timedelta(days=30), now),
-    }
-    if expr in relative_times:
-        return relative_times[expr]
+    # Relative expressions
+    if expr == "yesterday":
+        end = now.replace(hour=0, minute=0, second=0, microsecond=0)
+        start = end - timedelta(days=1)
+        return start, end
+
+    if expr in ("last week", "last 7 days"):
+        return now - timedelta(days=7), now
+
+    if expr in ("last month", "last 30 days"):
+        return now - timedelta(days=30), now
+
+    if expr == "today":
+        start = now.replace(hour=0, minute=0, second=0, microsecond=0)
+        return start, now
 
     # Try ISO date format (YYYY-MM-DD)
     date_match = re.match(r"^(\d{4})-(\d{2})-(\d{2})$", expr)
     if date_match:
-        try:
-            year, month, day = map(int, date_match.groups())
-            start = datetime(year, month, day, 0, 0, 0, tzinfo=timezone.utc)
-            return start, start + timedelta(days=1)
-        except ValueError:
-            # Invalid date components (e.g., month=13, day=32)
-            pass
+        year, month, day = map(int, date_match.groups())
+        start = datetime(year, month, day, 0, 0, 0, tzinfo=timezone.utc)
+        end = start + timedelta(days=1)
+        return start, end
 
     # Try ISO datetime
     try:
         parsed = datetime.fromisoformat(expr.replace("Z", "+00:00"))
         if parsed.tzinfo is None:
             parsed = parsed.replace(tzinfo=timezone.utc)
+        # Return +/- 1 hour window around the specified time
         return parsed - timedelta(hours=1), parsed + timedelta(hours=1)
     except ValueError:
-        return None, None
+        pass
+
+    # Fallback: treat as "latest"
+    return None, None
 
 
 class AgentOutputTool(BaseTool):

autogpt_platform/backend/backend/api/features/chat/tools/agent_search.py

@@ -1,151 +0,0 @@
-"""Shared agent search functionality for find_agent and find_library_agent tools."""
-
-import logging
-from typing import Literal
-
-from backend.api.features.library import db as library_db
-from backend.api.features.store import db as store_db
-from backend.util.exceptions import DatabaseError, NotFoundError
-
-from .models import (
-    AgentInfo,
-    AgentsFoundResponse,
-    ErrorResponse,
-    NoResultsResponse,
-    ToolResponseBase,
-)
-
-logger = logging.getLogger(__name__)
-
-SearchSource = Literal["marketplace", "library"]
-
-
-async def search_agents(
-    query: str,
-    source: SearchSource,
-    session_id: str | None,
-    user_id: str | None = None,
-) -> ToolResponseBase:
-    """
-    Search for agents in marketplace or user library.
-
-    Args:
-        query: Search query string
-        source: "marketplace" or "library"
-        session_id: Chat session ID
-        user_id: User ID (required for library search)
-
-    Returns:
-        AgentsFoundResponse, NoResultsResponse, or ErrorResponse
-    """
-    if not query:
-        return ErrorResponse(
-            message="Please provide a search query", session_id=session_id
-        )
-
-    if source == "library" and not user_id:
-        return ErrorResponse(
-            message="User authentication required to search library",
-            session_id=session_id,
-        )
-
-    agents: list[AgentInfo] = []
-    try:
-        if source == "marketplace":
-            logger.info(f"Searching marketplace for: {query}")
-            results = await store_db.get_store_agents(search_query=query, page_size=5)
-            for agent in results.agents:
-                agents.append(
-                    AgentInfo(
-                        id=f"{agent.creator}/{agent.slug}",
-                        name=agent.agent_name,
-                        description=agent.description or "",
-                        source="marketplace",
-                        in_library=False,
-                        creator=agent.creator,
-                        category="general",
-                        rating=agent.rating,
-                        runs=agent.runs,
-                        is_featured=False,
-                    )
-                )
-        else:  # library
-            logger.info(f"Searching user library for: {query}")
-            results = await library_db.list_library_agents(
-                user_id=user_id,  # type: ignore[arg-type]
-                search_term=query,
-                page_size=10,
-            )
-            for agent in results.agents:
-                agents.append(
-                    AgentInfo(
-                        id=agent.id,
-                        name=agent.name,
-                        description=agent.description or "",
-                        source="library",
-                        in_library=True,
-                        creator=agent.creator_name,
-                        status=agent.status.value,
-                        can_access_graph=agent.can_access_graph,
-                        has_external_trigger=agent.has_external_trigger,
-                        new_output=agent.new_output,
-                        graph_id=agent.graph_id,
-                    )
-                )
-        logger.info(f"Found {len(agents)} agents in {source}")
-    except NotFoundError:
-        pass
-    except DatabaseError as e:
-        logger.error(f"Error searching {source}: {e}", exc_info=True)
-        return ErrorResponse(
-            message=f"Failed to search {source}. Please try again.",
-            error=str(e),
-            session_id=session_id,
-        )
-
-    if not agents:
-        suggestions = (
-            [
-                "Try more general terms",
-                "Browse categories in the marketplace",
-                "Check spelling",
-            ]
-            if source == "marketplace"
-            else [
-                "Try different keywords",
-                "Use find_agent to search the marketplace",
-                "Check your library at /library",
-            ]
-        )
-        no_results_msg = (
-            f"No agents found matching '{query}'. Try different keywords or browse the marketplace."
-            if source == "marketplace"
-            else f"No agents matching '{query}' found in your library."
-        )
-        return NoResultsResponse(
-            message=no_results_msg, session_id=session_id, suggestions=suggestions
-        )
-
-    title = f"Found {len(agents)} agent{'s' if len(agents) != 1 else ''} "
-    title += (
-        f"for '{query}'"
-        if source == "marketplace"
-        else f"in your library for '{query}'"
-    )
-
-    message = (
-        "Now you have found some options for the user to choose from. "
-        "You can add a link to a recommended agent at: /marketplace/agent/agent_id "
-        "Please ask the user if they would like to use any of these agents."
-        if source == "marketplace"
-        else "Found agents in the user's library. You can provide a link to view an agent at: "
-        "/library/agents/{agent_id}. Use agent_output to get execution results, or run_agent to execute."
-    )
-
-    return AgentsFoundResponse(
-        message=message,
-        title=title,
-        agents=agents,
-        count=len(agents),
-        session_id=session_id,
-    )

autogpt_platform/backend/backend/api/features/chat/tools/base.py

@@ -6,7 +6,7 @@ from typing import Any
 from openai.types.chat import ChatCompletionToolParam
 
 from backend.api.features.chat.model import ChatSession
-from backend.api.features.chat.response_model import StreamToolOutputAvailable
+from backend.api.features.chat.response_model import StreamToolExecutionResult
 
 from .models import ErrorResponse, NeedLoginResponse, ToolResponseBase
 
@@ -53,7 +53,7 @@ class BaseTool:
         session: ChatSession,
         tool_call_id: str,
         **kwargs,
-    ) -> StreamToolOutputAvailable:
+    ) -> StreamToolExecutionResult:
         """Execute the tool with authentication check.
 
         Args:
@@ -69,10 +69,10 @@ class BaseTool:
             logger.error(
                 f"Attempted tool call for {self.name} but user not authenticated"
             )
-            return StreamToolOutputAvailable(
-                toolCallId=tool_call_id,
-                toolName=self.name,
-                output=NeedLoginResponse(
+            return StreamToolExecutionResult(
+                tool_id=tool_call_id,
+                tool_name=self.name,
+                result=NeedLoginResponse(
                     message=f"Please sign in to use {self.name}",
                     session_id=session.session_id,
                 ).model_dump_json(),
@@ -81,17 +81,17 @@ class BaseTool:
 
         try:
             result = await self._execute(user_id, session, **kwargs)
-            return StreamToolOutputAvailable(
-                toolCallId=tool_call_id,
-                toolName=self.name,
-                output=result.model_dump_json(),
+            return StreamToolExecutionResult(
+                tool_id=tool_call_id,
+                tool_name=self.name,
+                result=result.model_dump_json(),
             )
         except Exception as e:
             logger.error(f"Error in {self.name}: {e}", exc_info=True)
-            return StreamToolOutputAvailable(
-                toolCallId=tool_call_id,
-                toolName=self.name,
-                output=ErrorResponse(
+            return StreamToolExecutionResult(
+                tool_id=tool_call_id,
+                tool_name=self.name,
+                result=ErrorResponse(
                     message=f"An error occurred while executing {self.name}",
                     error=str(e),
                     session_id=session.session_id,

autogpt_platform/backend/backend/api/features/chat/tools/create_agent.py

@@ -0,0 +1,279 @@
+"""CreateAgentTool - Creates agents from natural language descriptions."""
+
+import logging
+from typing import Any
+
+from backend.api.features.chat.model import ChatSession
+
+from .agent_generator import (
+    apply_all_fixes,
+    decompose_goal,
+    generate_agent,
+    get_blocks_info,
+    save_agent_to_library,
+    validate_agent,
+)
+from .base import BaseTool
+from .models import (
+    AgentPreviewResponse,
+    AgentSavedResponse,
+    ClarificationNeededResponse,
+    ClarifyingQuestion,
+    ErrorResponse,
+    ToolResponseBase,
+)
+
+logger = logging.getLogger(__name__)
+
+# Maximum retries for agent generation with validation feedback
+MAX_GENERATION_RETRIES = 2
+
+
+class CreateAgentTool(BaseTool):
+    """Tool for creating agents from natural language descriptions."""
+
+    @property
+    def name(self) -> str:
+        return "create_agent"
+
+    @property
+    def description(self) -> str:
+        return (
+            "Create a new agent workflow from a natural language description. "
+            "First generates a preview, then saves to library if save=true."
+        )
+
+    @property
+    def requires_auth(self) -> bool:
+        return True
+
+    @property
+    def parameters(self) -> dict[str, Any]:
+        return {
+            "type": "object",
+            "properties": {
+                "description": {
+                    "type": "string",
+                    "description": (
+                        "Natural language description of what the agent should do. "
+                        "Be specific about inputs, outputs, and the workflow steps."
+                    ),
+                },
+                "context": {
+                    "type": "string",
+                    "description": (
+                        "Additional context or answers to previous clarifying questions. "
+                        "Include any preferences or constraints mentioned by the user."
+                    ),
+                },
+                "save": {
+                    "type": "boolean",
+                    "description": (
+                        "Whether to save the agent to the user's library. "
+                        "Default is true. Set to false for preview only."
+                    ),
+                    "default": True,
+                },
+            },
+            "required": ["description"],
+        }
+
+    async def _execute(
+        self,
+        user_id: str | None,
+        session: ChatSession,
+        **kwargs,
+    ) -> ToolResponseBase:
+        """Execute the create_agent tool.
+
+        Flow:
+        1. Decompose the description into steps (may return clarifying questions)
+        2. Generate agent JSON from the steps
+        3. Apply fixes to correct common LLM errors
+        4. Preview or save based on the save parameter
+        """
+        description = kwargs.get("description", "").strip()
+        context = kwargs.get("context", "")
+        save = kwargs.get("save", True)
+        session_id = session.session_id if session else None
+
+        if not description:
+            return ErrorResponse(
+                message="Please provide a description of what the agent should do.",
+                error="Missing description parameter",
+                session_id=session_id,
+            )
+
+        # Step 1: Decompose goal into steps
+        try:
+            decomposition_result = await decompose_goal(description, context)
+        except ValueError as e:
+            # Handle missing API key or configuration errors
+            return ErrorResponse(
+                message=f"Agent generation is not configured: {str(e)}",
+                error="configuration_error",
+                session_id=session_id,
+            )
+
+        if decomposition_result is None:
+            return ErrorResponse(
+                message="Failed to analyze the goal. Please try rephrasing.",
+                error="Decomposition failed",
+                session_id=session_id,
+            )
+
+        # Check if LLM returned clarifying questions
+        if decomposition_result.get("type") == "clarifying_questions":
+            questions = decomposition_result.get("questions", [])
+            return ClarificationNeededResponse(
+                message=(
+                    "I need some more information to create this agent. "
+                    "Please answer the following questions:"
+                ),
+                questions=[
+                    ClarifyingQuestion(
+                        question=q.get("question", ""),
+                        keyword=q.get("keyword", ""),
+                        example=q.get("example"),
+                    )
+                    for q in questions
+                ],
+                session_id=session_id,
+            )
+
+        # Check for unachievable/vague goals
+        if decomposition_result.get("type") == "unachievable_goal":
+            suggested = decomposition_result.get("suggested_goal", "")
+            reason = decomposition_result.get("reason", "")
+            return ErrorResponse(
+                message=(
+                    f"This goal cannot be accomplished with the available blocks. "
+                    f"{reason} "
+                    f"Suggestion: {suggested}"
+                ),
+                error="unachievable_goal",
+                details={"suggested_goal": suggested, "reason": reason},
+                session_id=session_id,
+            )
+
+        if decomposition_result.get("type") == "vague_goal":
+            suggested = decomposition_result.get("suggested_goal", "")
+            return ErrorResponse(
+                message=(
+                    f"The goal is too vague to create a specific workflow. "
+                    f"Suggestion: {suggested}"
+                ),
+                error="vague_goal",
+                details={"suggested_goal": suggested},
+                session_id=session_id,
+            )
+
+        # Step 2: Generate agent JSON with retry on validation failure
+        blocks_info = get_blocks_info()
+        agent_json = None
+        validation_errors = None
+
+        for attempt in range(MAX_GENERATION_RETRIES + 1):
+            # Generate agent (include validation errors from previous attempt)
+            if attempt == 0:
+                agent_json = await generate_agent(decomposition_result)
+            else:
+                # Retry with validation error feedback
+                logger.info(
+                    f"Retry {attempt}/{MAX_GENERATION_RETRIES} with validation feedback"
+                )
+                retry_instructions = {
+                    **decomposition_result,
+                    "previous_errors": validation_errors,
+                    "retry_instructions": (
+                        "The previous generation had validation errors. "
+                        "Please fix these issues in the new generation:\n"
+                        f"{validation_errors}"
+                    ),
+                }
+                agent_json = await generate_agent(retry_instructions)
+
+            if agent_json is None:
+                if attempt == MAX_GENERATION_RETRIES:
+                    return ErrorResponse(
+                        message="Failed to generate the agent. Please try again.",
+                        error="Generation failed",
+                        session_id=session_id,
+                    )
+                continue
+
+            # Step 3: Apply fixes to correct common errors
+            agent_json = apply_all_fixes(agent_json, blocks_info)
+
+            # Step 4: Validate the agent
+            is_valid, validation_errors = validate_agent(agent_json, blocks_info)
+
+            if is_valid:
+                logger.info(f"Agent generated successfully on attempt {attempt + 1}")
+                break
+
+            logger.warning(
+                f"Validation failed on attempt {attempt + 1}: {validation_errors}"
+            )
+
+            if attempt == MAX_GENERATION_RETRIES:
+                # Return error with validation details
+                return ErrorResponse(
+                    message=(
+                        f"Generated agent has validation errors after {MAX_GENERATION_RETRIES + 1} attempts. "
+                        f"Please try rephrasing your request or simplify the workflow."
+                    ),
+                    error="validation_failed",
+                    details={"validation_errors": validation_errors},
+                    session_id=session_id,
+                )
+
+        agent_name = agent_json.get("name", "Generated Agent")
+        agent_description = agent_json.get("description", "")
+        node_count = len(agent_json.get("nodes", []))
+        link_count = len(agent_json.get("links", []))
+
+        # Step 4: Preview or save
+        if not save:
+            return AgentPreviewResponse(
+                message=(
+                    f"I've generated an agent called '{agent_name}' with {node_count} blocks. "
+                    f"Review it and call create_agent with save=true to save it to your library."
+                ),
+                agent_json=agent_json,
+                agent_name=agent_name,
+                description=agent_description,
+                node_count=node_count,
+                link_count=link_count,
+                session_id=session_id,
+            )
+
+        # Save to library
+        if not user_id:
+            return ErrorResponse(
+                message="You must be logged in to save agents.",
+                error="auth_required",
+                session_id=session_id,
+            )
+
+        try:
+            created_graph, library_agent = await save_agent_to_library(
+                agent_json, user_id
+            )
+
+            return AgentSavedResponse(
+                message=f"Agent '{created_graph.name}' has been saved to your library!",
+                agent_id=created_graph.id,
+                agent_name=created_graph.name,
+                library_agent_id=library_agent.id,
+                library_agent_link=f"/library/{library_agent.id}",
+                agent_page_link=f"/build?flowID={created_graph.id}",
+                session_id=session_id,
+            )
+        except Exception as e:
+            return ErrorResponse(
+                message=f"Failed to save the agent: {str(e)}",
+                error="save_failed",
+                details={"exception": str(e)},
+                session_id=session_id,
+            )

autogpt_platform/backend/backend/api/features/chat/tools/edit_agent.py

@@ -0,0 +1,294 @@
+"""EditAgentTool - Edits existing agents using natural language."""
+
+import logging
+from typing import Any
+
+from backend.api.features.chat.model import ChatSession
+
+from .agent_generator import (
+    apply_agent_patch,
+    apply_all_fixes,
+    generate_agent_patch,
+    get_agent_as_json,
+    get_blocks_info,
+    save_agent_to_library,
+    validate_agent,
+)
+from .base import BaseTool
+from .models import (
+    AgentPreviewResponse,
+    AgentSavedResponse,
+    ClarificationNeededResponse,
+    ClarifyingQuestion,
+    ErrorResponse,
+    ToolResponseBase,
+)
+
+logger = logging.getLogger(__name__)
+
+# Maximum retries for patch generation with validation feedback
+MAX_GENERATION_RETRIES = 2
+
+
+class EditAgentTool(BaseTool):
+    """Tool for editing existing agents using natural language."""
+
+    @property
+    def name(self) -> str:
+        return "edit_agent"
+
+    @property
+    def description(self) -> str:
+        return (
+            "Edit an existing agent from the user's library using natural language. "
+            "Generates a patch to update the agent while preserving unchanged parts."
+        )
+
+    @property
+    def requires_auth(self) -> bool:
+        return True
+
+    @property
+    def parameters(self) -> dict[str, Any]:
+        return {
+            "type": "object",
+            "properties": {
+                "agent_id": {
+                    "type": "string",
+                    "description": (
+                        "The ID of the agent to edit. "
+                        "Can be a graph ID or library agent ID."
+                    ),
+                },
+                "changes": {
+                    "type": "string",
+                    "description": (
+                        "Natural language description of what changes to make. "
+                        "Be specific about what to add, remove, or modify."
+                    ),
+                },
+                "context": {
+                    "type": "string",
+                    "description": (
+                        "Additional context or answers to previous clarifying questions."
+                    ),
+                },
+                "save": {
+                    "type": "boolean",
+                    "description": (
+                        "Whether to save the changes. "
+                        "Default is true. Set to false for preview only."
+                    ),
+                    "default": True,
+                },
+            },
+            "required": ["agent_id", "changes"],
+        }
+
+    async def _execute(
+        self,
+        user_id: str | None,
+        session: ChatSession,
+        **kwargs,
+    ) -> ToolResponseBase:
+        """Execute the edit_agent tool.
+
+        Flow:
+        1. Fetch the current agent
+        2. Generate a patch based on the requested changes
+        3. Apply the patch to create an updated agent
+        4. Preview or save based on the save parameter
+        """
+        agent_id = kwargs.get("agent_id", "").strip()
+        changes = kwargs.get("changes", "").strip()
+        context = kwargs.get("context", "")
+        save = kwargs.get("save", True)
+        session_id = session.session_id if session else None
+
+        if not agent_id:
+            return ErrorResponse(
+                message="Please provide the agent ID to edit.",
+                error="Missing agent_id parameter",
+                session_id=session_id,
+            )
+
+        if not changes:
+            return ErrorResponse(
+                message="Please describe what changes you want to make.",
+                error="Missing changes parameter",
+                session_id=session_id,
+            )
+
+        # Step 1: Fetch current agent
+        current_agent = await get_agent_as_json(agent_id, user_id)
+
+        if current_agent is None:
+            return ErrorResponse(
+                message=f"Could not find agent with ID '{agent_id}' in your library.",
+                error="agent_not_found",
+                session_id=session_id,
+            )
+
+        # Build the update request with context
+        update_request = changes
+        if context:
+            update_request = f"{changes}\n\nAdditional context:\n{context}"
+
+        # Step 2: Generate patch with retry on validation failure
+        blocks_info = get_blocks_info()
+        updated_agent = None
+        validation_errors = None
+        intent = "Applied requested changes"
+
+        for attempt in range(MAX_GENERATION_RETRIES + 1):
+            # Generate patch (include validation errors from previous attempt)
+            try:
+                if attempt == 0:
+                    patch_result = await generate_agent_patch(
+                        update_request, current_agent
+                    )
+                else:
+                    # Retry with validation error feedback
+                    logger.info(
+                        f"Retry {attempt}/{MAX_GENERATION_RETRIES} with validation feedback"
+                    )
+                    retry_request = (
+                        f"{update_request}\n\n"
+                        f"IMPORTANT: The previous edit had validation errors. "
+                        f"Please fix these issues:\n{validation_errors}"
+                    )
+                    patch_result = await generate_agent_patch(
+                        retry_request, current_agent
+                    )
+            except ValueError as e:
+                # Handle missing API key or configuration errors
+                return ErrorResponse(
+                    message=f"Agent generation is not configured: {str(e)}",
+                    error="configuration_error",
+                    session_id=session_id,
+                )
+
+            if patch_result is None:
+                if attempt == MAX_GENERATION_RETRIES:
+                    return ErrorResponse(
+                        message="Failed to generate changes. Please try rephrasing.",
+                        error="Patch generation failed",
+                        session_id=session_id,
+                    )
+                continue
+
+            # Check if LLM returned clarifying questions
+            if patch_result.get("type") == "clarifying_questions":
+                questions = patch_result.get("questions", [])
+                return ClarificationNeededResponse(
+                    message=(
+                        "I need some more information about the changes. "
+                        "Please answer the following questions:"
+                    ),
+                    questions=[
+                        ClarifyingQuestion(
+                            question=q.get("question", ""),
+                            keyword=q.get("keyword", ""),
+                            example=q.get("example"),
+                        )
+                        for q in questions
+                    ],
+                    session_id=session_id,
+                )
+
+            # Step 3: Apply patch and fixes
+            try:
+                updated_agent = apply_agent_patch(current_agent, patch_result)
+                updated_agent = apply_all_fixes(updated_agent, blocks_info)
+            except Exception as e:
+                if attempt == MAX_GENERATION_RETRIES:
+                    return ErrorResponse(
+                        message=f"Failed to apply changes: {str(e)}",
+                        error="patch_apply_failed",
+                        details={"exception": str(e)},
+                        session_id=session_id,
+                    )
+                validation_errors = str(e)
+                continue
+
+            # Step 4: Validate the updated agent
+            is_valid, validation_errors = validate_agent(updated_agent, blocks_info)
+
+            if is_valid:
+                logger.info(f"Agent edited successfully on attempt {attempt + 1}")
+                intent = patch_result.get("intent", "Applied requested changes")
+                break
+
+            logger.warning(
+                f"Validation failed on attempt {attempt + 1}: {validation_errors}"
+            )
+
+            if attempt == MAX_GENERATION_RETRIES:
+                # Return error with validation details
+                return ErrorResponse(
+                    message=(
+                        f"Updated agent has validation errors after "
+                        f"{MAX_GENERATION_RETRIES + 1} attempts. "
+                        f"Please try rephrasing your request or simplify the changes."
+                    ),
+                    error="validation_failed",
+                    details={"validation_errors": validation_errors},
+                    session_id=session_id,
+                )
+
+        # At this point, updated_agent is guaranteed to be set (we return on all failure paths)
+        assert updated_agent is not None
+
+        agent_name = updated_agent.get("name", "Updated Agent")
+        agent_description = updated_agent.get("description", "")
+        node_count = len(updated_agent.get("nodes", []))
+        link_count = len(updated_agent.get("links", []))
+
+        # Step 5: Preview or save
+        if not save:
+            return AgentPreviewResponse(
+                message=(
+                    f"I've updated the agent. Changes: {intent}. "
+                    f"The agent now has {node_count} blocks. "
+                    f"Review it and call edit_agent with save=true to save the changes."
+                ),
+                agent_json=updated_agent,
+                agent_name=agent_name,
+                description=agent_description,
+                node_count=node_count,
+                link_count=link_count,
+                session_id=session_id,
+            )
+
+        # Save to library (creates a new version)
+        if not user_id:
+            return ErrorResponse(
+                message="You must be logged in to save agents.",
+                error="auth_required",
+                session_id=session_id,
+            )
+
+        try:
+            created_graph, library_agent = await save_agent_to_library(
+                updated_agent, user_id, is_update=True
+            )
+
+            return AgentSavedResponse(
+                message=(
+                    f"Updated agent '{created_graph.name}' has been saved to your library! "
+                    f"Changes: {intent}"
+                ),
+                agent_id=created_graph.id,
+                agent_name=created_graph.name,
+                library_agent_id=library_agent.id,
+                library_agent_link=f"/library/{library_agent.id}",
+                agent_page_link=f"/build?flowID={created_graph.id}",
+                session_id=session_id,
+            )
+        except Exception as e:
+            return ErrorResponse(
+                message=f"Failed to save the updated agent: {str(e)}",
+                error="save_failed",
+                details={"exception": str(e)},
+                session_id=session_id,
+            )

autogpt_platform/backend/backend/api/features/chat/tools/find_agent.py

@@ -1,16 +1,26 @@
-"""Tool for discovering agents from marketplace."""
+"""Tool for discovering agents from marketplace and user library."""
 
+import logging
 from typing import Any
 
 from backend.api.features.chat.model import ChatSession
+from backend.api.features.store import db as store_db
+from backend.util.exceptions import DatabaseError, NotFoundError
 
-from .agent_search import search_agents
 from .base import BaseTool
-from .models import ToolResponseBase
+from .models import (
+    AgentCarouselResponse,
+    AgentInfo,
+    ErrorResponse,
+    NoResultsResponse,
+    ToolResponseBase,
+)
+
+logger = logging.getLogger(__name__)
 
 
 class FindAgentTool(BaseTool):
-    """Tool for discovering agents from the marketplace."""
+    """Tool for discovering agents based on user needs."""
 
     @property
     def name(self) -> str:
@@ -36,11 +46,84 @@ class FindAgentTool(BaseTool):
         }
 
     async def _execute(
-        self, user_id: str | None, session: ChatSession, **kwargs
+        self,
+        user_id: str | None,
+        session: ChatSession,
+        **kwargs,
     ) -> ToolResponseBase:
-        return await search_agents(
-            query=kwargs.get("query", "").strip(),
-            source="marketplace",
-            session_id=session.session_id,
-            user_id=user_id,
+        """Search for agents in the marketplace.
+
+        Args:
+            user_id: User ID (may be anonymous)
+            session_id: Chat session ID
+            query: Search query
+
+        Returns:
+            AgentCarouselResponse: List of agents found in the marketplace
+            NoResultsResponse: No agents found in the marketplace
+            ErrorResponse: Error message
+        """
+        query = kwargs.get("query", "").strip()
+        session_id = session.session_id
+        if not query:
+            return ErrorResponse(
+                message="Please provide a search query",
+                session_id=session_id,
+            )
+        agents = []
+        try:
+            logger.info(f"Searching marketplace for: {query}")
+            store_results = await store_db.get_store_agents(
+                search_query=query,
+                page_size=5,
+            )
+
+            logger.info(f"Find agents tool found {len(store_results.agents)} agents")
+            for agent in store_results.agents:
+                agent_id = f"{agent.creator}/{agent.slug}"
+                logger.info(f"Building agent ID = {agent_id}")
+                agents.append(
+                    AgentInfo(
+                        id=agent_id,
+                        name=agent.agent_name,
+                        description=agent.description or "",
+                        source="marketplace",
+                        in_library=False,
+                        creator=agent.creator,
+                        category="general",
+                        rating=agent.rating,
+                        runs=agent.runs,
+                        is_featured=False,
+                    ),
+                )
+        except NotFoundError:
+            pass
+        except DatabaseError as e:
+            logger.error(f"Error searching agents: {e}", exc_info=True)
+            return ErrorResponse(
+                message="Failed to search for agents. Please try again.",
+                error=str(e),
+                session_id=session_id,
+            )
+        if not agents:
+            return NoResultsResponse(
+                message=f"No agents found matching '{query}'. Try different keywords or browse the marketplace. If you have 3 consecutive find_agent tool calls results and found no agents. Please stop trying and ask the user if there is anything else you can help with.",
+                session_id=session_id,
+                suggestions=[
+                    "Try more general terms",
+                    "Browse categories in the marketplace",
+                    "Check spelling",
+                ],
+            )
+
+        # Return formatted carousel
+        title = (
+            f"Found {len(agents)} agent{'s' if len(agents) != 1 else ''} for '{query}'"
+        )
+        return AgentCarouselResponse(
+            message="Now you have found some options for the user to choose from. You can add a link to a recommended agent at: /marketplace/agent/agent_id Please ask the user if they would like to use any of these agents. If they do, please call the get_agent_details tool for this agent.",
+            title=title,
+            agents=agents,
+            count=len(agents),
+            session_id=session_id,
         )

autogpt_platform/backend/backend/api/features/chat/tools/find_block.py

@@ -0,0 +1,253 @@
+"""Tool for searching available blocks using hybrid search."""
+
+import logging
+from typing import Any
+
+from backend.api.features.chat.model import ChatSession
+from backend.blocks import load_all_blocks
+
+from .base import BaseTool
+from .models import (
+    BlockInfoSummary,
+    BlockListResponse,
+    ErrorResponse,
+    NoResultsResponse,
+    ToolResponseBase,
+)
+from .search_blocks import get_block_search_index
+
+logger = logging.getLogger(__name__)
+
+
+class FindBlockTool(BaseTool):
+    """Tool for searching available blocks."""
+
+    @property
+    def name(self) -> str:
+        return "find_block"
+
+    @property
+    def description(self) -> str:
+        return (
+            "Search for available blocks by name or description. "
+            "Blocks are reusable components that perform specific tasks like "
+            "sending emails, making API calls, processing text, etc. "
+            "Use this to find blocks that can be executed directly."
+        )
+
+    @property
+    def parameters(self) -> dict[str, Any]:
+        return {
+            "type": "object",
+            "properties": {
+                "query": {
+                    "type": "string",
+                    "description": (
+                        "Search query to find blocks by name or description. "
+                        "Use keywords like 'email', 'http', 'text', 'ai', etc."
+                    ),
+                },
+            },
+            "required": ["query"],
+        }
+
+    @property
+    def requires_auth(self) -> bool:
+        return True
+
+    def _matches_query(self, block, query: str) -> tuple[int, bool]:
+        """
+        Check if a block matches the query and return a priority score.
+
+        Returns (priority, matches) where:
+        - priority 0: exact name match
+        - priority 1: name contains query
+        - priority 2: description contains query
+        - priority 3: category contains query
+        """
+        query_lower = query.lower()
+        name_lower = block.name.lower()
+        desc_lower = block.description.lower()
+
+        # Exact name match
+        if query_lower == name_lower:
+            return 0, True
+
+        # Name contains query
+        if query_lower in name_lower:
+            return 1, True
+
+        # Description contains query
+        if query_lower in desc_lower:
+            return 2, True
+
+        # Category contains query
+        for category in block.categories:
+            if query_lower in category.name.lower():
+                return 3, True
+
+        return 4, False
+
+    async def _execute(
+        self,
+        user_id: str | None,
+        session: ChatSession,
+        **kwargs,
+    ) -> ToolResponseBase:
+        """Search for blocks matching the query.
+
+        Args:
+            user_id: User ID (required)
+            session: Chat session
+            query: Search query
+
+        Returns:
+            BlockListResponse: List of matching blocks
+            NoResultsResponse: No blocks found
+            ErrorResponse: Error message
+        """
+        query = kwargs.get("query", "").strip()
+        session_id = session.session_id
+
+        if not query:
+            return ErrorResponse(
+                message="Please provide a search query",
+                session_id=session_id,
+            )
+
+        try:
+            # Try hybrid search first
+            search_results = self._hybrid_search(query)
+
+            if search_results is not None:
+                # Hybrid search succeeded
+                if not search_results:
+                    return NoResultsResponse(
+                        message=f"No blocks found matching '{query}'",
+                        session_id=session_id,
+                        suggestions=[
+                            "Try more general terms",
+                            "Search by category: ai, text, social, search, etc.",
+                            "Check block names like 'SendEmail', 'HttpRequest', etc.",
+                        ],
+                    )
+
+                # Get full block info for each result
+                all_blocks = load_all_blocks()
+                blocks = []
+                for result in search_results:
+                    block_cls = all_blocks.get(result.block_id)
+                    if block_cls:
+                        block = block_cls()
+                        blocks.append(
+                            BlockInfoSummary(
+                                id=block.id,
+                                name=block.name,
+                                description=block.description,
+                                categories=[cat.name for cat in block.categories],
+                                input_schema=block.input_schema.jsonschema(),
+                                output_schema=block.output_schema.jsonschema(),
+                            )
+                        )
+
+                return BlockListResponse(
+                    message=(
+                        f"Found {len(blocks)} block{'s' if len(blocks) != 1 else ''} "
+                        f"matching '{query}'. Use run_block to execute a block with "
+                        "the required inputs."
+                    ),
+                    blocks=blocks,
+                    count=len(blocks),
+                    query=query,
+                    session_id=session_id,
+                )
+
+            # Fallback to simple search if hybrid search failed
+            return self._simple_search(query, session_id)
+
+        except Exception as e:
+            logger.error(f"Error searching blocks: {e}", exc_info=True)
+            return ErrorResponse(
+                message="Failed to search blocks. Please try again.",
+                error=str(e),
+                session_id=session_id,
+            )
+
+    def _hybrid_search(self, query: str) -> list | None:
+        """
+        Perform hybrid search using embeddings and BM25.
+
+        Returns:
+            List of BlockSearchResult if successful, None if index not available
+        """
+        try:
+            index = get_block_search_index()
+            if not index.load():
+                logger.info(
+                    "Block search index not available, falling back to simple search"
+                )
+                return None
+
+            results = index.search(query, top_k=10)
+            logger.info(f"Hybrid search found {len(results)} blocks for: {query}")
+            return results
+
+        except Exception as e:
+            logger.warning(f"Hybrid search failed, falling back to simple: {e}")
+            return None
+
+    def _simple_search(self, query: str, session_id: str) -> ToolResponseBase:
+        """Fallback simple search using substring matching."""
+        all_blocks = load_all_blocks()
+        logger.info(f"Simple searching {len(all_blocks)} blocks for: {query}")
+
+        # Find matching blocks with priority scores
+        matches: list[tuple[int, Any]] = []
+        for block_id, block_cls in all_blocks.items():
+            block = block_cls()
+            priority, is_match = self._matches_query(block, query)
+            if is_match:
+                matches.append((priority, block))
+
+        # Sort by priority (lower is better)
+        matches.sort(key=lambda x: x[0])
+
+        # Take top 10 results
+        top_matches = [block for _, block in matches[:10]]
+
+        if not top_matches:
+            return NoResultsResponse(
+                message=f"No blocks found matching '{query}'",
+                session_id=session_id,
+                suggestions=[
+                    "Try more general terms",
+                    "Search by category: ai, text, social, search, etc.",
+                    "Check block names like 'SendEmail', 'HttpRequest', etc.",
+                ],
+            )
+
+        # Build response
+        blocks = []
+        for block in top_matches:
+            blocks.append(
+                BlockInfoSummary(
+                    id=block.id,
+                    name=block.name,
+                    description=block.description,
+                    categories=[cat.name for cat in block.categories],
+                    input_schema=block.input_schema.jsonschema(),
+                    output_schema=block.output_schema.jsonschema(),
+                )
+            )
+
+        return BlockListResponse(
+            message=(
+                f"Found {len(blocks)} block{'s' if len(blocks) != 1 else ''} "
+                f"matching '{query}'. Use run_block to execute a block with "
+                "the required inputs."
+            ),
+            blocks=blocks,
+            count=len(blocks),
+            query=query,
+            session_id=session_id,
+        )

autogpt_platform/backend/backend/api/features/chat/tools/find_library_agent.py

@@ -1,12 +1,22 @@
 """Tool for searching agents in the user's library."""
 
+import logging
 from typing import Any
 
 from backend.api.features.chat.model import ChatSession
+from backend.api.features.library import db as library_db
+from backend.util.exceptions import DatabaseError
 
-from .agent_search import search_agents
 from .base import BaseTool
-from .models import ToolResponseBase
+from .models import (
+    AgentCarouselResponse,
+    AgentInfo,
+    ErrorResponse,
+    NoResultsResponse,
+    ToolResponseBase,
+)
+
+logger = logging.getLogger(__name__)
 
 
 class FindLibraryAgentTool(BaseTool):
@@ -31,7 +41,10 @@ class FindLibraryAgentTool(BaseTool):
             "properties": {
                 "query": {
                     "type": "string",
-                    "description": "Search query to find agents by name or description.",
+                    "description": (
+                        "Search query to find agents by name or description. "
+                        "Use keywords for best results."
+                    ),
                 },
             },
             "required": ["query"],
@@ -42,11 +55,103 @@ class FindLibraryAgentTool(BaseTool):
         return True
 
     async def _execute(
-        self, user_id: str | None, session: ChatSession, **kwargs
+        self,
+        user_id: str | None,
+        session: ChatSession,
+        **kwargs,
     ) -> ToolResponseBase:
-        return await search_agents(
-            query=kwargs.get("query", "").strip(),
-            source="library",
-            session_id=session.session_id,
-            user_id=user_id,
+        """Search for agents in the user's library.
+
+        Args:
+            user_id: User ID (required)
+            session: Chat session
+            query: Search query
+
+        Returns:
+            AgentCarouselResponse: List of agents found in the library
+            NoResultsResponse: No agents found
+            ErrorResponse: Error message
+        """
+        query = kwargs.get("query", "").strip()
+        session_id = session.session_id
+
+        if not query:
+            return ErrorResponse(
+                message="Please provide a search query",
+                session_id=session_id,
+            )
+
+        if not user_id:
+            return ErrorResponse(
+                message="User authentication required to search library",
+                session_id=session_id,
+            )
+
+        agents = []
+        try:
+            logger.info(f"Searching user library for: {query}")
+            library_results = await library_db.list_library_agents(
+                user_id=user_id,
+                search_term=query,
+                page_size=10,
+            )
+
+            logger.info(
+                f"Find library agents tool found {len(library_results.agents)} agents"
+            )
+
+            for agent in library_results.agents:
+                agents.append(
+                    AgentInfo(
+                        id=agent.id,
+                        name=agent.name,
+                        description=agent.description or "",
+                        source="library",
+                        in_library=True,
+                        creator=agent.creator_name,
+                        status=agent.status.value,
+                        can_access_graph=agent.can_access_graph,
+                        has_external_trigger=agent.has_external_trigger,
+                        new_output=agent.new_output,
+                        graph_id=agent.graph_id,
+                    ),
+                )
+
+        except DatabaseError as e:
+            logger.error(f"Error searching library agents: {e}", exc_info=True)
+            return ErrorResponse(
+                message="Failed to search library. Please try again.",
+                error=str(e),
+                session_id=session_id,
+            )
+
+        if not agents:
+            return NoResultsResponse(
+                message=(
+                    f"No agents found matching '{query}' in your library. "
+                    "Try different keywords or use find_agent to search the marketplace."
+                ),
+                session_id=session_id,
+                suggestions=[
+                    "Try more general terms",
+                    "Use find_agent to search the marketplace",
+                    "Check your library at /library",
+                ],
+            )
+
+        title = (
+            f"Found {len(agents)} agent{'s' if len(agents) != 1 else ''} "
+            f"in your library for '{query}'"
+        )
+
+        return AgentCarouselResponse(
+            message=(
+                "Found agents in the user's library. You can provide a link to "
+                "view an agent at: /library/agents/{agent_id}. "
+                "Use agent_output to get execution results, or run_agent to execute."
+            ),
+            title=title,
+            agents=agents,
+            count=len(agents),
+            session_id=session_id,
         )

autogpt_platform/backend/backend/api/features/chat/tools/index_blocks.py

@@ -0,0 +1,483 @@
+#!/usr/bin/env python3
+"""
+Block Indexer for Hybrid Search
+
+Creates a hybrid search index from blocks:
+- OpenAI embeddings (text-embedding-3-small)
+- BM25 index for lexical search
+- Name index for title matching boost
+
+Supports incremental updates by tracking content hashes.
+
+Usage:
+    python -m backend.server.v2.chat.tools.index_blocks [--force]
+"""
+
+import argparse
+import base64
+import hashlib
+import json
+import logging
+import os
+import re
+import sys
+from collections import defaultdict
+from datetime import datetime, timezone
+from pathlib import Path
+from typing import Any
+
+import numpy as np
+
+logger = logging.getLogger(__name__)
+
+# Check for OpenAI availability
+try:
+    import openai  # noqa: F401
+
+    HAS_OPENAI = True
+except ImportError:
+    HAS_OPENAI = False
+    print("Warning: openai not installed. Run: pip install openai")
+
+# Default embedding model (OpenAI)
+DEFAULT_EMBEDDING_MODEL = "text-embedding-3-small"
+DEFAULT_EMBEDDING_DIM = 1536
+
+# Output path (relative to this file)
+INDEX_PATH = Path(__file__).parent / "blocks_index.json"
+
+# Stopwords for tokenization
+STOPWORDS = {
+    "the",
+    "a",
+    "an",
+    "is",
+    "are",
+    "was",
+    "were",
+    "be",
+    "been",
+    "being",
+    "have",
+    "has",
+    "had",
+    "do",
+    "does",
+    "did",
+    "will",
+    "would",
+    "could",
+    "should",
+    "may",
+    "might",
+    "must",
+    "shall",
+    "can",
+    "need",
+    "dare",
+    "ought",
+    "used",
+    "to",
+    "of",
+    "in",
+    "for",
+    "on",
+    "with",
+    "at",
+    "by",
+    "from",
+    "as",
+    "into",
+    "through",
+    "during",
+    "before",
+    "after",
+    "above",
+    "below",
+    "between",
+    "under",
+    "again",
+    "further",
+    "then",
+    "once",
+    "and",
+    "but",
+    "or",
+    "nor",
+    "so",
+    "yet",
+    "both",
+    "either",
+    "neither",
+    "not",
+    "only",
+    "own",
+    "same",
+    "than",
+    "too",
+    "very",
+    "just",
+    "also",
+    "now",
+    "here",
+    "there",
+    "when",
+    "where",
+    "why",
+    "how",
+    "all",
+    "each",
+    "every",
+    "few",
+    "more",
+    "most",
+    "other",
+    "some",
+    "such",
+    "no",
+    "any",
+    "this",
+    "that",
+    "these",
+    "those",
+    "it",
+    "its",
+    "block",  # Too common in block context
+}
+
+
+def tokenize(text: str) -> list[str]:
+    """Simple tokenizer for BM25."""
+    text = text.lower()
+    # Remove code blocks if any
+    text = re.sub(r"```[\s\S]*?```", "", text)
+    text = re.sub(r"`[^`]+`", "", text)
+    # Extract words (including camelCase split)
+    # First, split camelCase
+    text = re.sub(r"([a-z])([A-Z])", r"\1 \2", text)
+    # Extract words
+    words = re.findall(r"\b[a-z][a-z0-9_-]*\b", text)
+    # Remove very short words and stopwords
+    return [w for w in words if len(w) > 2 and w not in STOPWORDS]
+
+
+def build_searchable_text(block: Any) -> str:
+    """Build searchable text from block attributes."""
+    parts = []
+
+    # Block name (split camelCase for better tokenization)
+    name = block.name
+    # Split camelCase: GetCurrentTimeBlock -> Get Current Time Block
+    name_split = re.sub(r"([a-z])([A-Z])", r"\1 \2", name)
+    parts.append(name_split)
+
+    # Description
+    if block.description:
+        parts.append(block.description)
+
+    # Categories
+    for category in block.categories:
+        parts.append(category.name)
+
+    # Input schema field names and descriptions
+    try:
+        input_schema = block.input_schema.jsonschema()
+        if "properties" in input_schema:
+            for field_name, field_info in input_schema["properties"].items():
+                parts.append(field_name)
+                if "description" in field_info:
+                    parts.append(field_info["description"])
+    except Exception:
+        pass
+
+    # Output schema field names
+    try:
+        output_schema = block.output_schema.jsonschema()
+        if "properties" in output_schema:
+            for field_name in output_schema["properties"]:
+                parts.append(field_name)
+    except Exception:
+        pass
+
+    return " ".join(parts)
+
+
+def compute_content_hash(text: str) -> str:
+    """Compute MD5 hash of text for change detection."""
+    return hashlib.md5(text.encode()).hexdigest()
+
+
+def load_existing_index(index_path: Path) -> dict[str, Any] | None:
+    """Load existing index if present."""
+    if not index_path.exists():
+        return None
+
+    try:
+        with open(index_path, "r", encoding="utf-8") as f:
+            return json.load(f)
+    except Exception as e:
+        logger.warning(f"Failed to load existing index: {e}")
+        return None
+
+
+def create_embeddings(
+    texts: list[str],
+    model_name: str = DEFAULT_EMBEDDING_MODEL,
+    batch_size: int = 100,
+) -> np.ndarray:
+    """Create embeddings using OpenAI API."""
+    if not HAS_OPENAI:
+        raise RuntimeError("openai not installed. Run: pip install openai")
+
+    # Import here to satisfy type checker
+    from openai import OpenAI
+
+    # Check for API key
+    api_key = os.getenv("OPENAI_API_KEY")
+    if not api_key:
+        raise RuntimeError("OPENAI_API_KEY environment variable not set")
+
+    client = OpenAI(api_key=api_key)
+    embeddings = []
+
+    print(f"Creating embeddings for {len(texts)} texts using {model_name}...")
+
+    for i in range(0, len(texts), batch_size):
+        batch = texts[i : i + batch_size]
+        # Truncate texts to max token limit (8191 tokens for text-embedding-3-small)
+        # Roughly 4 chars per token, so ~32000 chars max
+        batch = [text[:32000] for text in batch]
+
+        response = client.embeddings.create(
+            model=model_name,
+            input=batch,
+        )
+
+        for embedding_data in response.data:
+            embeddings.append(embedding_data.embedding)
+
+        print(f"  Processed {min(i + batch_size, len(texts))}/{len(texts)} texts")
+
+    return np.array(embeddings, dtype=np.float32)
+
+
+def build_bm25_data(
+    blocks_data: list[dict[str, Any]],
+) -> dict[str, Any]:
+    """Build BM25 metadata from block data."""
+    # Tokenize all searchable texts
+    tokenized_docs = []
+    for block in blocks_data:
+        tokens = tokenize(block["searchable_text"])
+        tokenized_docs.append(tokens)
+
+    # Calculate document frequencies
+    doc_freq: dict[str, int] = {}
+    for tokens in tokenized_docs:
+        seen = set()
+        for token in tokens:
+            if token not in seen:
+                doc_freq[token] = doc_freq.get(token, 0) + 1
+                seen.add(token)
+
+    n_docs = len(tokenized_docs)
+    doc_lens = [len(d) for d in tokenized_docs]
+    avgdl = sum(doc_lens) / max(n_docs, 1)
+
+    return {
+        "n_docs": n_docs,
+        "avgdl": avgdl,
+        "df": doc_freq,
+        "doc_lens": doc_lens,
+    }
+
+
+def build_name_index(
+    blocks_data: list[dict[str, Any]],
+) -> dict[str, list[list[int | float]]]:
+    """Build inverted index for name search boost."""
+    index: dict[str, list[list[int | float]]] = defaultdict(list)
+
+    for idx, block in enumerate(blocks_data):
+        # Tokenize block name
+        name_tokens = tokenize(block["name"])
+        seen = set()
+
+        for i, token in enumerate(name_tokens):
+            if token in seen:
+                continue
+            seen.add(token)
+
+            # Score: first token gets higher weight
+            score = 1.5 if i == 0 else 1.0
+            index[token].append([idx, score])
+
+    return dict(index)
+
+
+def build_block_index(
+    force_rebuild: bool = False,
+    output_path: Path = INDEX_PATH,
+) -> dict[str, Any]:
+    """
+    Build the block search index.
+
+    Args:
+        force_rebuild: If True, rebuild all embeddings even if unchanged
+        output_path: Path to save the index
+
+    Returns:
+        The generated index dictionary
+    """
+    # Import here to avoid circular imports
+    from backend.blocks import load_all_blocks
+
+    print("Loading all blocks...")
+    all_blocks = load_all_blocks()
+    print(f"Found {len(all_blocks)} blocks")
+
+    # Load existing index for incremental updates
+    existing_index = None if force_rebuild else load_existing_index(output_path)
+    existing_blocks: dict[str, dict[str, Any]] = {}
+
+    if existing_index:
+        print(
+            f"Loaded existing index with {len(existing_index.get('blocks', []))} blocks"
+        )
+        for block in existing_index.get("blocks", []):
+            existing_blocks[block["id"]] = block
+
+    # Process each block
+    blocks_data: list[dict[str, Any]] = []
+    blocks_needing_embedding: list[tuple[int, str]] = []  # (index, searchable_text)
+
+    for block_id, block_cls in all_blocks.items():
+        try:
+            block = block_cls()
+
+            # Skip disabled blocks
+            if block.disabled:
+                continue
+
+            searchable_text = build_searchable_text(block)
+            content_hash = compute_content_hash(searchable_text)
+
+            block_data = {
+                "id": block.id,
+                "name": block.name,
+                "description": block.description,
+                "categories": [cat.name for cat in block.categories],
+                "searchable_text": searchable_text,
+                "content_hash": content_hash,
+                "emb": None,  # Will be filled later
+            }
+
+            # Check if we can reuse existing embedding
+            if (
+                block.id in existing_blocks
+                and existing_blocks[block.id].get("content_hash") == content_hash
+                and existing_blocks[block.id].get("emb")
+            ):
+                # Reuse existing embedding
+                block_data["emb"] = existing_blocks[block.id]["emb"]
+            else:
+                # Need new embedding
+                blocks_needing_embedding.append((len(blocks_data), searchable_text))
+
+            blocks_data.append(block_data)
+
+        except Exception as e:
+            logger.warning(f"Failed to process block {block_id}: {e}")
+            continue
+
+    print(f"Processed {len(blocks_data)} blocks")
+    print(f"Blocks needing new embeddings: {len(blocks_needing_embedding)}")
+
+    # Create embeddings for new/changed blocks
+    if blocks_needing_embedding and HAS_OPENAI:
+        texts_to_embed = [text for _, text in blocks_needing_embedding]
+        try:
+            embeddings = create_embeddings(texts_to_embed)
+
+            # Assign embeddings to blocks
+            for i, (block_idx, _) in enumerate(blocks_needing_embedding):
+                emb = embeddings[i].astype(np.float32)
+                # Encode as base64
+                blocks_data[block_idx]["emb"] = base64.b64encode(emb.tobytes()).decode(
+                    "ascii"
+                )
+        except Exception as e:
+            print(f"Warning: Failed to create embeddings: {e}")
+    elif blocks_needing_embedding:
+        print(
+            "Warning: Cannot create embeddings (openai not installed or OPENAI_API_KEY not set)"
+        )
+
+    # Build BM25 data
+    print("Building BM25 index...")
+    bm25_data = build_bm25_data(blocks_data)
+
+    # Build name index
+    print("Building name index...")
+    name_index = build_name_index(blocks_data)
+
+    # Build final index
+    index = {
+        "version": "1.0.0",
+        "embedding_model": DEFAULT_EMBEDDING_MODEL,
+        "embedding_dim": DEFAULT_EMBEDDING_DIM,
+        "generated_at": datetime.now(timezone.utc).isoformat(),
+        "blocks": blocks_data,
+        "bm25": bm25_data,
+        "name_index": name_index,
+    }
+
+    # Save index
+    print(f"Saving index to {output_path}...")
+    with open(output_path, "w", encoding="utf-8") as f:
+        json.dump(index, f, separators=(",", ":"))
+
+    size_kb = output_path.stat().st_size / 1024
+    print(f"Index saved ({size_kb:.1f} KB)")
+
+    # Print statistics
+    print("\nIndex Statistics:")
+    print(f"  Blocks indexed: {len(blocks_data)}")
+    print(f"  BM25 vocabulary size: {len(bm25_data['df'])}")
+    print(f"  Name index terms: {len(name_index)}")
+    print(f"  Embeddings: {'Yes' if any(b.get('emb') for b in blocks_data) else 'No'}")
+
+    return index
+
+
+def main():
+    parser = argparse.ArgumentParser(description="Build hybrid search index for blocks")
+    parser.add_argument(
+        "--force",
+        action="store_true",
+        help="Force rebuild all embeddings even if unchanged",
+    )
+    parser.add_argument(
+        "--output",
+        type=Path,
+        default=INDEX_PATH,
+        help=f"Output index file path (default: {INDEX_PATH})",
+    )
+
+    args = parser.parse_args()
+
+    try:
+        build_block_index(
+            force_rebuild=args.force,
+            output_path=args.output,
+        )
+    except Exception as e:
+        print(f"Error building index: {e}")
+        import traceback
+
+        traceback.print_exc()
+        sys.exit(1)
+
+
+if __name__ == "__main__":
+    main()

autogpt_platform/backend/backend/api/features/chat/tools/models.py

@@ -12,15 +12,23 @@ from backend.data.model import CredentialsMetaInput
 class ResponseType(str, Enum):
     """Types of tool responses."""
 
-    AGENTS_FOUND = "agents_found"
+    AGENT_CAROUSEL = "agent_carousel"
     AGENT_DETAILS = "agent_details"
     SETUP_REQUIREMENTS = "setup_requirements"
     EXECUTION_STARTED = "execution_started"
     NEED_LOGIN = "need_login"
     ERROR = "error"
     NO_RESULTS = "no_results"
+    SUCCESS = "success"
+    DOC_SEARCH_RESULTS = "doc_search_results"
     AGENT_OUTPUT = "agent_output"
+    BLOCK_LIST = "block_list"
+    BLOCK_OUTPUT = "block_output"
     UNDERSTANDING_UPDATED = "understanding_updated"
+    # Agent generation responses
+    AGENT_PREVIEW = "agent_preview"
+    AGENT_SAVED = "agent_saved"
+    CLARIFICATION_NEEDED = "clarification_needed"
 
 
 # Base response model
@@ -53,14 +61,14 @@ class AgentInfo(BaseModel):
     graph_id: str | None = None
 
 
-class AgentsFoundResponse(ToolResponseBase):
+class AgentCarouselResponse(ToolResponseBase):
     """Response for find_agent tool."""
 
-    type: ResponseType = ResponseType.AGENTS_FOUND
+    type: ResponseType = ResponseType.AGENT_CAROUSEL
     title: str = "Available Agents"
     agents: list[AgentInfo]
     count: int
-    name: str = "agents_found"
+    name: str = "agent_carousel"
 
 
 class NoResultsResponse(ToolResponseBase):
@@ -177,6 +185,28 @@ class ErrorResponse(ToolResponseBase):
     details: dict[str, Any] | None = None
 
 
+# Documentation search models
+class DocSearchResult(BaseModel):
+    """A single documentation search result."""
+
+    title: str
+    path: str
+    section: str
+    snippet: str  # Short excerpt for UI display
+    content: str  # Full text content for LLM to read and understand
+    score: float
+    doc_url: str | None = None
+
+
+class DocSearchResultsResponse(ToolResponseBase):
+    """Response for search_docs tool."""
+
+    type: ResponseType = ResponseType.DOC_SEARCH_RESULTS
+    results: list[DocSearchResult]
+    count: int
+    query: str
+
+
 # Agent output models
 class ExecutionOutputInfo(BaseModel):
     """Summary of a single execution's outputs."""
@@ -202,6 +232,37 @@ class AgentOutputResponse(ToolResponseBase):
     total_executions: int = 0
 
 
+# Block models
+class BlockInfoSummary(BaseModel):
+    """Summary of a block for search results."""
+
+    id: str
+    name: str
+    description: str
+    categories: list[str]
+    input_schema: dict[str, Any]
+    output_schema: dict[str, Any]
+
+
+class BlockListResponse(ToolResponseBase):
+    """Response for find_block tool."""
+
+    type: ResponseType = ResponseType.BLOCK_LIST
+    blocks: list[BlockInfoSummary]
+    count: int
+    query: str
+
+
+class BlockOutputResponse(ToolResponseBase):
+    """Response for run_block tool."""
+
+    type: ResponseType = ResponseType.BLOCK_OUTPUT
+    block_id: str
+    block_name: str
+    outputs: dict[str, list[Any]]
+    success: bool = True
+
+
 # Business understanding models
 class UnderstandingUpdatedResponse(ToolResponseBase):
     """Response for add_understanding tool."""
@@ -209,3 +270,41 @@ class UnderstandingUpdatedResponse(ToolResponseBase):
     type: ResponseType = ResponseType.UNDERSTANDING_UPDATED
     updated_fields: list[str] = Field(default_factory=list)
     current_understanding: dict[str, Any] = Field(default_factory=dict)
+
+
+# Agent generation models
+class ClarifyingQuestion(BaseModel):
+    """A question that needs user clarification."""
+
+    question: str
+    keyword: str
+    example: str | None = None
+
+
+class AgentPreviewResponse(ToolResponseBase):
+    """Response for previewing a generated agent before saving."""
+
+    type: ResponseType = ResponseType.AGENT_PREVIEW
+    agent_json: dict[str, Any]
+    agent_name: str
+    description: str
+    node_count: int
+    link_count: int = 0
+
+
+class AgentSavedResponse(ToolResponseBase):
+    """Response when an agent is saved to the library."""
+
+    type: ResponseType = ResponseType.AGENT_SAVED
+    agent_id: str
+    agent_name: str
+    library_agent_id: str
+    library_agent_link: str
+    agent_page_link: str  # Link to the agent builder/editor page
+
+
+class ClarificationNeededResponse(ToolResponseBase):
+    """Response when the LLM needs more information from the user."""
+
+    type: ResponseType = ResponseType.CLARIFICATION_NEEDED
+    questions: list[ClarifyingQuestion] = Field(default_factory=list)

autogpt_platform/backend/backend/api/features/chat/tools/run_agent_test.py

@@ -1,5 +1,4 @@
 import uuid
-from unittest.mock import AsyncMock, patch
 
 import orjson
 import pytest
@@ -18,17 +17,6 @@ setup_test_data = setup_test_data
 setup_firecrawl_test_data = setup_firecrawl_test_data
 
 
-@pytest.fixture(scope="session", autouse=True)
-def mock_embedding_functions():
-    """Mock embedding functions for all tests to avoid database/API dependencies."""
-    with patch(
-        "backend.api.features.store.db.ensure_embedding",
-        new_callable=AsyncMock,
-        return_value=True,
-    ):
-        yield
-
-
 @pytest.mark.asyncio(scope="session")
 async def test_run_agent(setup_test_data):
     """Test that the run_agent tool successfully executes an approved agent"""
@@ -58,11 +46,11 @@ async def test_run_agent(setup_test_data):
 
     # Verify the response
     assert response is not None
-    assert hasattr(response, "output")
+    assert hasattr(response, "result")
     # Parse the result JSON to verify the execution started
 
-    assert isinstance(response.output, str)
-    result_data = orjson.loads(response.output)
+    assert isinstance(response.result, str)
+    result_data = orjson.loads(response.result)
     assert "execution_id" in result_data
     assert "graph_id" in result_data
     assert result_data["graph_id"] == graph.id
@@ -98,11 +86,11 @@ async def test_run_agent_missing_inputs(setup_test_data):
 
     # Verify that we get an error response
     assert response is not None
-    assert hasattr(response, "output")
+    assert hasattr(response, "result")
     # The tool should return an ErrorResponse when setup info indicates not ready
 
-    assert isinstance(response.output, str)
-    result_data = orjson.loads(response.output)
+    assert isinstance(response.result, str)
+    result_data = orjson.loads(response.result)
     assert "message" in result_data
 
 
@@ -130,10 +118,10 @@ async def test_run_agent_invalid_agent_id(setup_test_data):
 
     # Verify that we get an error response
     assert response is not None
-    assert hasattr(response, "output")
+    assert hasattr(response, "result")
 
-    assert isinstance(response.output, str)
-    result_data = orjson.loads(response.output)
+    assert isinstance(response.result, str)
+    result_data = orjson.loads(response.result)
     assert "message" in result_data
     # Should get an error about failed setup or not found
     assert any(
@@ -170,12 +158,12 @@ async def test_run_agent_with_llm_credentials(setup_llm_test_data):
 
     # Verify the response
     assert response is not None
-    assert hasattr(response, "output")
+    assert hasattr(response, "result")
 
     # Parse the result JSON to verify the execution started
 
-    assert isinstance(response.output, str)
-    result_data = orjson.loads(response.output)
+    assert isinstance(response.result, str)
+    result_data = orjson.loads(response.result)
 
     # Should successfully start execution since credentials are available
     assert "execution_id" in result_data
@@ -207,9 +195,9 @@ async def test_run_agent_shows_available_inputs_when_none_provided(setup_test_da
     )
 
     assert response is not None
-    assert hasattr(response, "output")
-    assert isinstance(response.output, str)
-    result_data = orjson.loads(response.output)
+    assert hasattr(response, "result")
+    assert isinstance(response.result, str)
+    result_data = orjson.loads(response.result)
 
     # Should return agent_details type showing available inputs
     assert result_data.get("type") == "agent_details"
@@ -242,9 +230,9 @@ async def test_run_agent_with_use_defaults(setup_test_data):
     )
 
     assert response is not None
-    assert hasattr(response, "output")
-    assert isinstance(response.output, str)
-    result_data = orjson.loads(response.output)
+    assert hasattr(response, "result")
+    assert isinstance(response.result, str)
+    result_data = orjson.loads(response.result)
 
     # Should execute successfully
     assert "execution_id" in result_data
@@ -272,9 +260,9 @@ async def test_run_agent_missing_credentials(setup_firecrawl_test_data):
     )
 
     assert response is not None
-    assert hasattr(response, "output")
-    assert isinstance(response.output, str)
-    result_data = orjson.loads(response.output)
+    assert hasattr(response, "result")
+    assert isinstance(response.result, str)
+    result_data = orjson.loads(response.result)
 
     # Should return setup_requirements type with missing credentials
     assert result_data.get("type") == "setup_requirements"
@@ -304,9 +292,9 @@ async def test_run_agent_invalid_slug_format(setup_test_data):
     )
 
     assert response is not None
-    assert hasattr(response, "output")
-    assert isinstance(response.output, str)
-    result_data = orjson.loads(response.output)
+    assert hasattr(response, "result")
+    assert isinstance(response.result, str)
+    result_data = orjson.loads(response.result)
 
     # Should return error
     assert result_data.get("type") == "error"
@@ -317,10 +305,9 @@ async def test_run_agent_invalid_slug_format(setup_test_data):
 async def test_run_agent_unauthenticated():
     """Test that run_agent returns need_login for unauthenticated users."""
     tool = RunAgentTool()
-    # Session has a user_id (session owner), but we test tool execution without user_id
-    session = make_session(user_id="test-session-owner")
+    session = make_session(user_id=None)
 
-    # Execute without user_id to test unauthenticated behavior
+    # Execute without user_id
     response = await tool.execute(
         user_id=None,
         session_id=str(uuid.uuid4()),
@@ -331,9 +318,9 @@ async def test_run_agent_unauthenticated():
     )
 
     assert response is not None
-    assert hasattr(response, "output")
-    assert isinstance(response.output, str)
-    result_data = orjson.loads(response.output)
+    assert hasattr(response, "result")
+    assert isinstance(response.result, str)
+    result_data = orjson.loads(response.result)
 
     # Base tool returns need_login type for unauthenticated users
     assert result_data.get("type") == "need_login"
@@ -363,9 +350,9 @@ async def test_run_agent_schedule_without_cron(setup_test_data):
     )
 
     assert response is not None
-    assert hasattr(response, "output")
-    assert isinstance(response.output, str)
-    result_data = orjson.loads(response.output)
+    assert hasattr(response, "result")
+    assert isinstance(response.result, str)
+    result_data = orjson.loads(response.result)
 
     # Should return error about missing cron
     assert result_data.get("type") == "error"
@@ -395,9 +382,9 @@ async def test_run_agent_schedule_without_name(setup_test_data):
     )
 
     assert response is not None
-    assert hasattr(response, "output")
-    assert isinstance(response.output, str)
-    result_data = orjson.loads(response.output)
+    assert hasattr(response, "result")
+    assert isinstance(response.result, str)
+    result_data = orjson.loads(response.result)
 
     # Should return error about missing schedule_name
     assert result_data.get("type") == "error"

autogpt_platform/backend/backend/api/features/chat/tools/run_block.py

@@ -0,0 +1,287 @@
+"""Tool for executing blocks directly."""
+
+import logging
+from collections import defaultdict
+from typing import Any
+
+from backend.api.features.chat.model import ChatSession
+from backend.data.block import get_block
+from backend.data.model import CredentialsMetaInput
+from backend.integrations.creds_manager import IntegrationCredentialsManager
+from backend.util.exceptions import BlockError
+
+from .base import BaseTool
+from .models import (
+    BlockOutputResponse,
+    ErrorResponse,
+    SetupInfo,
+    SetupRequirementsResponse,
+    ToolResponseBase,
+    UserReadiness,
+)
+
+logger = logging.getLogger(__name__)
+
+
+class RunBlockTool(BaseTool):
+    """Tool for executing a block and returning its outputs."""
+
+    @property
+    def name(self) -> str:
+        return "run_block"
+
+    @property
+    def description(self) -> str:
+        return (
+            "Execute a specific block with the provided input data. "
+            "Use find_block to discover available blocks and their input schemas. "
+            "The block will run and return its outputs once complete."
+        )
+
+    @property
+    def parameters(self) -> dict[str, Any]:
+        return {
+            "type": "object",
+            "properties": {
+                "block_id": {
+                    "type": "string",
+                    "description": "The UUID of the block to execute",
+                },
+                "input_data": {
+                    "type": "object",
+                    "description": (
+                        "Input values for the block. Must match the block's input schema. "
+                        "Check the block's input_schema from find_block for required fields."
+                    ),
+                },
+            },
+            "required": ["block_id", "input_data"],
+        }
+
+    @property
+    def requires_auth(self) -> bool:
+        return True
+
+    async def _check_block_credentials(
+        self,
+        user_id: str,
+        block: Any,
+    ) -> tuple[dict[str, CredentialsMetaInput], list[CredentialsMetaInput]]:
+        """
+        Check if user has required credentials for a block.
+
+        Returns:
+            tuple[matched_credentials, missing_credentials]
+        """
+        matched_credentials: dict[str, CredentialsMetaInput] = {}
+        missing_credentials: list[CredentialsMetaInput] = []
+
+        # Get credential field info from block's input schema
+        credentials_fields_info = block.input_schema.get_credentials_fields_info()
+
+        if not credentials_fields_info:
+            return matched_credentials, missing_credentials
+
+        # Get user's available credentials
+        creds_manager = IntegrationCredentialsManager()
+        available_creds = await creds_manager.store.get_all_creds(user_id)
+
+        for field_name, field_info in credentials_fields_info.items():
+            # field_info.provider is a frozenset of acceptable providers
+            # field_info.supported_types is a frozenset of acceptable types
+            matching_cred = next(
+                (
+                    cred
+                    for cred in available_creds
+                    if cred.provider in field_info.provider
+                    and cred.type in field_info.supported_types
+                ),
+                None,
+            )
+
+            if matching_cred:
+                matched_credentials[field_name] = CredentialsMetaInput(
+                    id=matching_cred.id,
+                    provider=matching_cred.provider,  # type: ignore
+                    type=matching_cred.type,
+                    title=matching_cred.title,
+                )
+            else:
+                # Create a placeholder for the missing credential
+                provider = next(iter(field_info.provider), "unknown")
+                cred_type = next(iter(field_info.supported_types), "api_key")
+                missing_credentials.append(
+                    CredentialsMetaInput(
+                        id=field_name,
+                        provider=provider,  # type: ignore
+                        type=cred_type,  # type: ignore
+                        title=field_name.replace("_", " ").title(),
+                    )
+                )
+
+        return matched_credentials, missing_credentials
+
+    async def _execute(
+        self,
+        user_id: str | None,
+        session: ChatSession,
+        **kwargs,
+    ) -> ToolResponseBase:
+        """Execute a block with the given input data.
+
+        Args:
+            user_id: User ID (required)
+            session: Chat session
+            block_id: Block UUID to execute
+            input_data: Input values for the block
+
+        Returns:
+            BlockOutputResponse: Block execution outputs
+            SetupRequirementsResponse: Missing credentials
+            ErrorResponse: Error message
+        """
+        block_id = kwargs.get("block_id", "").strip()
+        input_data = kwargs.get("input_data", {})
+        session_id = session.session_id
+
+        if not block_id:
+            return ErrorResponse(
+                message="Please provide a block_id",
+                session_id=session_id,
+            )
+
+        if not isinstance(input_data, dict):
+            return ErrorResponse(
+                message="input_data must be an object",
+                session_id=session_id,
+            )
+
+        if not user_id:
+            return ErrorResponse(
+                message="Authentication required",
+                session_id=session_id,
+            )
+
+        # Get the block
+        block = get_block(block_id)
+        if not block:
+            return ErrorResponse(
+                message=f"Block '{block_id}' not found",
+                session_id=session_id,
+            )
+
+        logger.info(f"Executing block {block.name} ({block_id}) for user {user_id}")
+
+        # Check credentials
+        creds_manager = IntegrationCredentialsManager()
+        matched_credentials, missing_credentials = await self._check_block_credentials(
+            user_id, block
+        )
+
+        if missing_credentials:
+            # Return setup requirements response with missing credentials
+            missing_creds_dict = {c.id: c.model_dump() for c in missing_credentials}
+
+            return SetupRequirementsResponse(
+                message=(
+                    f"Block '{block.name}' requires credentials that are not configured. "
+                    "Please set up the required credentials before running this block."
+                ),
+                session_id=session_id,
+                setup_info=SetupInfo(
+                    agent_id=block_id,
+                    agent_name=block.name,
+                    user_readiness=UserReadiness(
+                        has_all_credentials=False,
+                        missing_credentials=missing_creds_dict,
+                        ready_to_run=False,
+                    ),
+                    requirements={
+                        "credentials": [c.model_dump() for c in missing_credentials],
+                        "inputs": self._get_inputs_list(block),
+                        "execution_modes": ["immediate"],
+                    },
+                ),
+                graph_id=None,
+                graph_version=None,
+            )
+
+        try:
+            # Fetch actual credentials and prepare kwargs for block execution
+            exec_kwargs: dict[str, Any] = {"user_id": user_id}
+
+            for field_name, cred_meta in matched_credentials.items():
+                # Inject metadata into input_data (for validation)
+                if field_name not in input_data:
+                    input_data[field_name] = cred_meta.model_dump()
+
+                # Fetch actual credentials and pass as kwargs (for execution)
+                actual_credentials = await creds_manager.get(
+                    user_id, cred_meta.id, lock=False
+                )
+                if actual_credentials:
+                    exec_kwargs[field_name] = actual_credentials
+                else:
+                    return ErrorResponse(
+                        message=f"Failed to retrieve credentials for {field_name}",
+                        session_id=session_id,
+                    )
+
+            # Execute the block and collect outputs
+            outputs: dict[str, list[Any]] = defaultdict(list)
+            async for output_name, output_data in block.execute(
+                input_data,
+                **exec_kwargs,
+            ):
+                outputs[output_name].append(output_data)
+
+            return BlockOutputResponse(
+                message=f"Block '{block.name}' executed successfully",
+                block_id=block_id,
+                block_name=block.name,
+                outputs=dict(outputs),
+                success=True,
+                session_id=session_id,
+            )
+
+        except BlockError as e:
+            logger.warning(f"Block execution failed: {e}")
+            return ErrorResponse(
+                message=f"Block execution failed: {e}",
+                error=str(e),
+                session_id=session_id,
+            )
+        except Exception as e:
+            logger.error(f"Unexpected error executing block: {e}", exc_info=True)
+            return ErrorResponse(
+                message=f"Failed to execute block: {str(e)}",
+                error=str(e),
+                session_id=session_id,
+            )
+
+    def _get_inputs_list(self, block: Any) -> list[dict[str, Any]]:
+        """Extract non-credential inputs from block schema."""
+        inputs_list = []
+        schema = block.input_schema.jsonschema()
+        properties = schema.get("properties", {})
+        required_fields = set(schema.get("required", []))
+
+        # Get credential field names to exclude
+        credentials_fields = set(block.input_schema.get_credentials_fields().keys())
+
+        for field_name, field_schema in properties.items():
+            # Skip credential fields
+            if field_name in credentials_fields:
+                continue
+
+            inputs_list.append(
+                {
+                    "name": field_name,
+                    "title": field_schema.get("title", field_name),
+                    "type": field_schema.get("type", "string"),
+                    "description": field_schema.get("description", ""),
+                    "required": field_name in required_fields,
+                }
+            )
+
+        return inputs_list

autogpt_platform/backend/backend/api/features/chat/tools/search_blocks.py

@@ -0,0 +1,460 @@
+"""
+Block Hybrid Search
+
+Combines multiple ranking signals for block search:
+- Semantic search (OpenAI embeddings + cosine similarity)
+- Lexical search (BM25)
+- Name matching (boost for block name matches)
+- Category matching (boost for category matches)
+
+Based on the docs search implementation.
+"""
+
+import base64
+import json
+import logging
+import math
+import os
+import re
+from dataclasses import dataclass
+from pathlib import Path
+from typing import Any, Optional
+
+import numpy as np
+
+logger = logging.getLogger(__name__)
+
+# OpenAI embedding model
+EMBEDDING_MODEL = "text-embedding-3-small"
+
+# Path to the JSON index file
+INDEX_PATH = Path(__file__).parent / "blocks_index.json"
+
+# Stopwords for tokenization (same as index_blocks.py)
+STOPWORDS = {
+    "the",
+    "a",
+    "an",
+    "is",
+    "are",
+    "was",
+    "were",
+    "be",
+    "been",
+    "being",
+    "have",
+    "has",
+    "had",
+    "do",
+    "does",
+    "did",
+    "will",
+    "would",
+    "could",
+    "should",
+    "may",
+    "might",
+    "must",
+    "shall",
+    "can",
+    "need",
+    "dare",
+    "ought",
+    "used",
+    "to",
+    "of",
+    "in",
+    "for",
+    "on",
+    "with",
+    "at",
+    "by",
+    "from",
+    "as",
+    "into",
+    "through",
+    "during",
+    "before",
+    "after",
+    "above",
+    "below",
+    "between",
+    "under",
+    "again",
+    "further",
+    "then",
+    "once",
+    "and",
+    "but",
+    "or",
+    "nor",
+    "so",
+    "yet",
+    "both",
+    "either",
+    "neither",
+    "not",
+    "only",
+    "own",
+    "same",
+    "than",
+    "too",
+    "very",
+    "just",
+    "also",
+    "now",
+    "here",
+    "there",
+    "when",
+    "where",
+    "why",
+    "how",
+    "all",
+    "each",
+    "every",
+    "few",
+    "more",
+    "most",
+    "other",
+    "some",
+    "such",
+    "no",
+    "any",
+    "this",
+    "that",
+    "these",
+    "those",
+    "it",
+    "its",
+    "block",
+}
+
+
+def tokenize(text: str) -> list[str]:
+    """Simple tokenizer for search."""
+    text = text.lower()
+    # Remove code blocks if any
+    text = re.sub(r"```[\s\S]*?```", "", text)
+    text = re.sub(r"`[^`]+`", "", text)
+    # Split camelCase
+    text = re.sub(r"([a-z])([A-Z])", r"\1 \2", text)
+    # Extract words
+    words = re.findall(r"\b[a-z][a-z0-9_-]*\b", text)
+    # Remove very short words and stopwords
+    return [w for w in words if len(w) > 2 and w not in STOPWORDS]
+
+
+@dataclass
+class SearchWeights:
+    """Configuration for hybrid search signal weights."""
+
+    semantic: float = 0.40  # Embedding similarity
+    bm25: float = 0.25  # Lexical matching
+    name_match: float = 0.25  # Block name matches
+    category_match: float = 0.10  # Category matches
+
+
+@dataclass
+class BlockSearchResult:
+    """A single block search result."""
+
+    block_id: str
+    name: str
+    description: str
+    categories: list[str]
+    score: float
+
+    # Individual signal scores (for debugging)
+    semantic_score: float = 0.0
+    bm25_score: float = 0.0
+    name_score: float = 0.0
+    category_score: float = 0.0
+
+
+class BlockSearchIndex:
+    """Hybrid search index for blocks combining BM25 + embeddings."""
+
+    def __init__(self, index_path: Path = INDEX_PATH):
+        self.blocks: list[dict[str, Any]] = []
+        self.bm25_data: dict[str, Any] = {}
+        self.name_index: dict[str, list[list[int | float]]] = {}
+        self.embeddings: Optional[np.ndarray] = None
+        self.normalized_embeddings: Optional[np.ndarray] = None
+        self._loaded = False
+        self._index_path = index_path
+        self._embedding_model: Any = None
+
+    def load(self) -> bool:
+        """Load the index from JSON file."""
+        if self._loaded:
+            return True
+
+        if not self._index_path.exists():
+            logger.warning(f"Block index not found at {self._index_path}")
+            return False
+
+        try:
+            with open(self._index_path, "r", encoding="utf-8") as f:
+                data = json.load(f)
+
+            self.blocks = data.get("blocks", [])
+            self.bm25_data = data.get("bm25", {})
+            self.name_index = data.get("name_index", {})
+
+            # Decode embeddings from base64
+            embeddings_list = []
+            for block in self.blocks:
+                if block.get("emb"):
+                    emb_bytes = base64.b64decode(block["emb"])
+                    emb = np.frombuffer(emb_bytes, dtype=np.float32)
+                    embeddings_list.append(emb)
+                else:
+                    # No embedding, use zeros
+                    dim = data.get("embedding_dim", 384)
+                    embeddings_list.append(np.zeros(dim, dtype=np.float32))
+
+            if embeddings_list:
+                self.embeddings = np.stack(embeddings_list)
+                # Precompute normalized embeddings for cosine similarity
+                norms = np.linalg.norm(self.embeddings, axis=1, keepdims=True)
+                self.normalized_embeddings = self.embeddings / (norms + 1e-10)
+
+            self._loaded = True
+            logger.info(f"Loaded block index with {len(self.blocks)} blocks")
+            return True
+
+        except Exception as e:
+            logger.error(f"Failed to load block index: {e}")
+            return False
+
+    def _get_openai_client(self) -> Any:
+        """Get OpenAI client for query embedding."""
+        if self._embedding_model is None:
+            try:
+                from openai import OpenAI
+
+                api_key = os.getenv("OPENAI_API_KEY")
+                if not api_key:
+                    logger.warning("OPENAI_API_KEY not set")
+                    return None
+                self._embedding_model = OpenAI(api_key=api_key)
+            except ImportError:
+                logger.warning("openai not installed")
+                return None
+        return self._embedding_model
+
+    def _embed_query(self, query: str) -> Optional[np.ndarray]:
+        """Embed the search query using OpenAI."""
+        client = self._get_openai_client()
+        if client is None:
+            return None
+
+        try:
+            response = client.embeddings.create(
+                model=EMBEDDING_MODEL,
+                input=query,
+            )
+            embedding = response.data[0].embedding
+            return np.array(embedding, dtype=np.float32)
+        except Exception as e:
+            logger.warning(f"Failed to embed query: {e}")
+            return None
+
+    def _compute_semantic_scores(self, query_embedding: np.ndarray) -> np.ndarray:
+        """Compute cosine similarity between query and all blocks."""
+        if self.normalized_embeddings is None:
+            return np.zeros(len(self.blocks))
+
+        # Normalize query embedding
+        query_norm = query_embedding / (np.linalg.norm(query_embedding) + 1e-10)
+
+        # Cosine similarity via dot product
+        similarities = self.normalized_embeddings @ query_norm
+
+        # Scale to [0, 1] (cosine ranges from -1 to 1)
+        return (similarities + 1) / 2
+
+    def _compute_bm25_scores(self, query_tokens: list[str]) -> np.ndarray:
+        """Compute BM25 scores for all blocks."""
+        scores = np.zeros(len(self.blocks))
+
+        if not self.bm25_data or not query_tokens:
+            return scores
+
+        # BM25 parameters
+        k1 = 1.5
+        b = 0.75
+        n_docs = self.bm25_data.get("n_docs", len(self.blocks))
+        avgdl = self.bm25_data.get("avgdl", 100)
+        df = self.bm25_data.get("df", {})
+        doc_lens = self.bm25_data.get("doc_lens", [100] * len(self.blocks))
+
+        for i, block in enumerate(self.blocks):
+            # Tokenize block's searchable text
+            block_tokens = tokenize(block.get("searchable_text", ""))
+            doc_len = doc_lens[i] if i < len(doc_lens) else len(block_tokens)
+
+            # Calculate BM25 score
+            score = 0.0
+            for token in query_tokens:
+                if token not in df:
+                    continue
+
+                # Term frequency in this document
+                tf = block_tokens.count(token)
+                if tf == 0:
+                    continue
+
+                # IDF
+                doc_freq = df.get(token, 0)
+                idf = math.log((n_docs - doc_freq + 0.5) / (doc_freq + 0.5) + 1)
+
+                # BM25 score component
+                numerator = tf * (k1 + 1)
+                denominator = tf + k1 * (1 - b + b * doc_len / avgdl)
+                score += idf * numerator / denominator
+
+            scores[i] = score
+
+        # Normalize to [0, 1]
+        max_score = scores.max()
+        if max_score > 0:
+            scores = scores / max_score
+
+        return scores
+
+    def _compute_name_scores(self, query_tokens: list[str]) -> np.ndarray:
+        """Compute name match scores using the name index."""
+        scores = np.zeros(len(self.blocks))
+
+        if not self.name_index or not query_tokens:
+            return scores
+
+        for token in query_tokens:
+            if token in self.name_index:
+                for block_idx, weight in self.name_index[token]:
+                    if block_idx < len(scores):
+                        scores[int(block_idx)] += weight
+
+        # Also check for partial matches in block names
+        for i, block in enumerate(self.blocks):
+            name_lower = block.get("name", "").lower()
+            for token in query_tokens:
+                if token in name_lower:
+                    scores[i] += 0.5
+
+        # Normalize to [0, 1]
+        max_score = scores.max()
+        if max_score > 0:
+            scores = scores / max_score
+
+        return scores
+
+    def _compute_category_scores(self, query_tokens: list[str]) -> np.ndarray:
+        """Compute category match scores."""
+        scores = np.zeros(len(self.blocks))
+
+        if not query_tokens:
+            return scores
+
+        for i, block in enumerate(self.blocks):
+            categories = block.get("categories", [])
+            category_text = " ".join(categories).lower()
+
+            for token in query_tokens:
+                if token in category_text:
+                    scores[i] += 1.0
+
+        # Normalize to [0, 1]
+        max_score = scores.max()
+        if max_score > 0:
+            scores = scores / max_score
+
+        return scores
+
+    def search(
+        self,
+        query: str,
+        top_k: int = 10,
+        weights: Optional[SearchWeights] = None,
+    ) -> list[BlockSearchResult]:
+        """
+        Perform hybrid search combining multiple signals.
+
+        Args:
+            query: Search query string
+            top_k: Number of results to return
+            weights: Optional custom weights for signals
+
+        Returns:
+            List of BlockSearchResult sorted by score
+        """
+        if not self._loaded and not self.load():
+            return []
+
+        if weights is None:
+            weights = SearchWeights()
+
+        # Tokenize query
+        query_tokens = tokenize(query)
+        if not query_tokens:
+            # Fallback: try raw query words
+            query_tokens = query.lower().split()
+
+        # Compute semantic scores
+        semantic_scores = np.zeros(len(self.blocks))
+        if self.normalized_embeddings is not None:
+            query_embedding = self._embed_query(query)
+            if query_embedding is not None:
+                semantic_scores = self._compute_semantic_scores(query_embedding)
+
+        # Compute other scores
+        bm25_scores = self._compute_bm25_scores(query_tokens)
+        name_scores = self._compute_name_scores(query_tokens)
+        category_scores = self._compute_category_scores(query_tokens)
+
+        # Combine scores using weights
+        combined_scores = (
+            weights.semantic * semantic_scores
+            + weights.bm25 * bm25_scores
+            + weights.name_match * name_scores
+            + weights.category_match * category_scores
+        )
+
+        # Get top-k indices
+        top_indices = np.argsort(combined_scores)[::-1][:top_k]
+
+        # Build results
+        results = []
+        for idx in top_indices:
+            if combined_scores[idx] <= 0:
+                continue
+
+            block = self.blocks[idx]
+            results.append(
+                BlockSearchResult(
+                    block_id=block["id"],
+                    name=block["name"],
+                    description=block["description"],
+                    categories=block.get("categories", []),
+                    score=float(combined_scores[idx]),
+                    semantic_score=float(semantic_scores[idx]),
+                    bm25_score=float(bm25_scores[idx]),
+                    name_score=float(name_scores[idx]),
+                    category_score=float(category_scores[idx]),
+                )
+            )
+
+        return results
+
+
+# Global index instance (lazy loaded)
+_block_search_index: Optional[BlockSearchIndex] = None
+
+
+def get_block_search_index() -> BlockSearchIndex:
+    """Get or create the block search index singleton."""
+    global _block_search_index
+    if _block_search_index is None:
+        _block_search_index = BlockSearchIndex(INDEX_PATH)
+    return _block_search_index

autogpt_platform/backend/backend/api/features/chat/tools/search_docs.py

@@ -0,0 +1,386 @@
+"""Tool for searching platform documentation."""
+
+import json
+import logging
+import math
+import re
+from pathlib import Path
+from typing import Any
+
+from backend.api.features.chat.model import ChatSession
+
+from .base import BaseTool
+from .models import (
+    DocSearchResult,
+    DocSearchResultsResponse,
+    ErrorResponse,
+    NoResultsResponse,
+    ToolResponseBase,
+)
+
+logger = logging.getLogger(__name__)
+
+# Documentation base URL
+DOCS_BASE_URL = "https://docs.agpt.co/platform"
+
+# Path to the JSON index file (relative to this file)
+INDEX_PATH = Path(__file__).parent / "docs_index.json"
+
+
+def tokenize(text: str) -> list[str]:
+    """Simple tokenizer for BM25."""
+    text = text.lower()
+    # Remove code blocks
+    text = re.sub(r"```[\s\S]*?```", "", text)
+    text = re.sub(r"`[^`]+`", "", text)
+    # Extract words
+    words = re.findall(r"\b[a-z][a-z0-9_-]*\b", text)
+    # Remove very short words and stopwords
+    stopwords = {
+        "the",
+        "a",
+        "an",
+        "is",
+        "are",
+        "was",
+        "were",
+        "be",
+        "been",
+        "being",
+        "have",
+        "has",
+        "had",
+        "do",
+        "does",
+        "did",
+        "will",
+        "would",
+        "could",
+        "should",
+        "may",
+        "might",
+        "must",
+        "shall",
+        "can",
+        "need",
+        "dare",
+        "ought",
+        "used",
+        "to",
+        "of",
+        "in",
+        "for",
+        "on",
+        "with",
+        "at",
+        "by",
+        "from",
+        "as",
+        "into",
+        "through",
+        "during",
+        "before",
+        "after",
+        "above",
+        "below",
+        "between",
+        "under",
+        "again",
+        "further",
+        "then",
+        "once",
+        "and",
+        "but",
+        "or",
+        "nor",
+        "so",
+        "yet",
+        "both",
+        "either",
+        "neither",
+        "not",
+        "only",
+        "own",
+        "same",
+        "than",
+        "too",
+        "very",
+        "just",
+        "also",
+        "now",
+        "here",
+        "there",
+        "when",
+        "where",
+        "why",
+        "how",
+        "all",
+        "each",
+        "every",
+        "both",
+        "few",
+        "more",
+        "most",
+        "other",
+        "some",
+        "such",
+        "no",
+        "any",
+        "this",
+        "that",
+        "these",
+        "those",
+        "it",
+        "its",
+    }
+    return [w for w in words if len(w) > 2 and w not in stopwords]
+
+
+class DocSearchIndex:
+    """Lightweight documentation search index using BM25."""
+
+    def __init__(self, index_path: Path):
+        self.chunks: list[dict] = []
+        self.bm25_data: dict = {}
+        self._loaded = False
+        self._index_path = index_path
+
+    def load(self) -> bool:
+        """Load the index from JSON file."""
+        if self._loaded:
+            return True
+
+        if not self._index_path.exists():
+            logger.warning(f"Documentation index not found at {self._index_path}")
+            return False
+
+        try:
+            with open(self._index_path, "r", encoding="utf-8") as f:
+                data = json.load(f)
+
+            self.chunks = data.get("chunks", [])
+            self.bm25_data = data.get("bm25", {})
+            self._loaded = True
+            logger.info(f"Loaded documentation index with {len(self.chunks)} chunks")
+            return True
+        except Exception as e:
+            logger.error(f"Failed to load documentation index: {e}")
+            return False
+
+    def search(self, query: str, top_k: int = 5) -> list[dict]:
+        """Search the index using BM25."""
+        if not self._loaded and not self.load():
+            return []
+
+        query_tokens = tokenize(query)
+        if not query_tokens:
+            return []
+
+        # BM25 parameters
+        k1 = 1.5
+        b = 0.75
+        n_docs = self.bm25_data.get("n_docs", len(self.chunks))
+        avgdl = self.bm25_data.get("avgdl", 100)
+        df = self.bm25_data.get("df", {})
+        doc_lens = self.bm25_data.get("doc_lens", [100] * len(self.chunks))
+
+        scores = []
+        for i, chunk in enumerate(self.chunks):
+            # Tokenize chunk text
+            chunk_tokens = tokenize(chunk.get("text", ""))
+            doc_len = doc_lens[i] if i < len(doc_lens) else len(chunk_tokens)
+
+            # Calculate BM25 score
+            score = 0.0
+            for token in query_tokens:
+                if token not in df:
+                    continue
+
+                # Term frequency in this document
+                tf = chunk_tokens.count(token)
+                if tf == 0:
+                    continue
+
+                # IDF
+                doc_freq = df.get(token, 0)
+                idf = math.log((n_docs - doc_freq + 0.5) / (doc_freq + 0.5) + 1)
+
+                # BM25 score component
+                numerator = tf * (k1 + 1)
+                denominator = tf + k1 * (1 - b + b * doc_len / avgdl)
+                score += idf * numerator / denominator
+
+            # Boost for title/heading matches
+            title = chunk.get("title", "").lower()
+            heading = chunk.get("heading", "").lower()
+            for token in query_tokens:
+                if token in title:
+                    score *= 1.5
+                if token in heading:
+                    score *= 1.2
+
+            scores.append((i, score))
+
+        # Sort by score and return top_k
+        scores.sort(key=lambda x: x[1], reverse=True)
+
+        results = []
+        seen_sections = set()
+        for idx, score in scores:
+            if score <= 0:
+                continue
+
+            chunk = self.chunks[idx]
+            section_key = (chunk.get("doc", ""), chunk.get("heading", ""))
+
+            # Deduplicate by section
+            if section_key in seen_sections:
+                continue
+            seen_sections.add(section_key)
+
+            results.append(
+                {
+                    "title": chunk.get("title", ""),
+                    "path": chunk.get("doc", ""),
+                    "heading": chunk.get("heading", ""),
+                    "text": chunk.get("text", ""),  # Full text for LLM comprehension
+                    "score": score,
+                }
+            )
+
+            if len(results) >= top_k:
+                break
+
+        return results
+
+
+# Global index instance (lazy loaded)
+_search_index: DocSearchIndex | None = None
+
+
+def get_search_index() -> DocSearchIndex:
+    """Get or create the search index singleton."""
+    global _search_index
+    if _search_index is None:
+        _search_index = DocSearchIndex(INDEX_PATH)
+    return _search_index
+
+
+class SearchDocsTool(BaseTool):
+    """Tool for searching AutoGPT platform documentation."""
+
+    @property
+    def name(self) -> str:
+        return "search_platform_docs"
+
+    @property
+    def description(self) -> str:
+        return (
+            "Search the AutoGPT platform documentation and support Q&A for information about "
+            "how to use the platform, create agents, configure blocks, "
+            "set up integrations, troubleshoot issues, and more. Use this when users ask "
+            "support questions or want to learn how to do something with AutoGPT."
+        )
+
+    @property
+    def parameters(self) -> dict[str, Any]:
+        return {
+            "type": "object",
+            "properties": {
+                "query": {
+                    "type": "string",
+                    "description": (
+                        "Search query describing what the user wants to learn about. "
+                        "Use keywords like 'blocks', 'agents', 'credentials', 'API', etc."
+                    ),
+                },
+            },
+            "required": ["query"],
+        }
+
+    async def _execute(
+        self,
+        user_id: str | None,
+        session: ChatSession,
+        **kwargs,
+    ) -> ToolResponseBase:
+        """Search documentation for the query.
+
+        Args:
+            user_id: User ID (may be anonymous)
+            session: Chat session
+            query: Search query
+
+        Returns:
+            DocSearchResultsResponse: List of matching documentation sections
+            NoResultsResponse: No results found
+            ErrorResponse: Error message
+        """
+        query = kwargs.get("query", "").strip()
+        session_id = session.session_id
+
+        if not query:
+            return ErrorResponse(
+                message="Please provide a search query",
+                session_id=session_id,
+            )
+
+        try:
+            index = get_search_index()
+            results = index.search(query, top_k=5)
+
+            if not results:
+                return NoResultsResponse(
+                    message=f"No documentation found for '{query}'. Try different keywords.",
+                    session_id=session_id,
+                    suggestions=[
+                        "Try more general terms like 'blocks', 'agents', 'setup'",
+                        "Check the documentation at docs.agpt.co",
+                    ],
+                )
+
+            # Convert to response format
+            doc_results = []
+            for r in results:
+                # Build documentation URL
+                path = r["path"]
+                if path.endswith(".md"):
+                    path = path[:-3]  # Remove .md extension
+                doc_url = f"{DOCS_BASE_URL}/{path}"
+
+                full_text = r["text"]
+                doc_results.append(
+                    DocSearchResult(
+                        title=r["title"],
+                        path=r["path"],
+                        section=r["heading"],
+                        snippet=(
+                            full_text[:300] + "..."
+                            if len(full_text) > 300
+                            else full_text
+                        ),
+                        content=full_text,  # Full text for LLM to read and understand
+                        score=round(r["score"], 3),
+                        doc_url=doc_url,
+                    )
+                )
+
+            return DocSearchResultsResponse(
+                message=(
+                    f"Found {len(doc_results)} relevant documentation sections. "
+                    "Use these to help answer the user's question. "
+                    "Include links to the documentation when helpful."
+                ),
+                results=doc_results,
+                count=len(doc_results),
+                query=query,
+                session_id=session_id,
+            )
+
+        except Exception as e:
+            logger.error(f"Error searching documentation: {e}", exc_info=True)
+            return ErrorResponse(
+                message="Failed to search documentation. Please try again.",
+                error=str(e),
+                session_id=session_id,
+            )

autogpt_platform/backend/backend/api/features/integrations/router.py

@@ -35,7 +35,11 @@ from backend.data.model import (
     OAuth2Credentials,
     UserIntegrations,
 )
-from backend.data.onboarding import OnboardingStep, complete_onboarding_step
+from backend.data.onboarding import (
+    OnboardingStep,
+    complete_onboarding_step,
+    increment_runs,
+)
 from backend.data.user import get_user_integrations
 from backend.executor.utils import add_graph_execution
 from backend.integrations.ayrshare import AyrshareClient, SocialPlatform
@@ -171,7 +175,6 @@ async def callback(
         f"Successfully processed OAuth callback for user {user_id} "
         f"and provider {provider.value}"
     )
-
     return CredentialsMetaResponse(
         id=credentials.id,
         provider=credentials.provider,
@@ -190,7 +193,6 @@ async def list_credentials(
     user_id: Annotated[str, Security(get_user_id)],
 ) -> list[CredentialsMetaResponse]:
     credentials = await creds_manager.store.get_all_creds(user_id)
-
     return [
         CredentialsMetaResponse(
             id=cred.id,
@@ -213,7 +215,6 @@ async def list_credentials_by_provider(
     user_id: Annotated[str, Security(get_user_id)],
 ) -> list[CredentialsMetaResponse]:
     credentials = await creds_manager.store.get_creds_by_provider(user_id, provider)
-
     return [
         CredentialsMetaResponse(
             id=cred.id,
@@ -377,6 +378,7 @@ async def webhook_ingress_generic(
         return
 
     await complete_onboarding_step(user_id, OnboardingStep.TRIGGER_WEBHOOK)
+    await increment_runs(user_id)
 
     # Execute all triggers concurrently for better performance
     tasks = []
@@ -829,18 +831,6 @@ async def list_providers() -> List[str]:
     return all_providers
 
 
-@router.get("/providers/system", response_model=List[str])
-async def list_system_providers() -> List[str]:
-    """
-    Get a list of providers that have platform credits (system credentials) available.
-
-    These providers can be used without the user providing their own API keys.
-    """
-    from backend.integrations.credentials_store import SYSTEM_PROVIDERS
-
-    return list(SYSTEM_PROVIDERS)
-
-
 @router.get("/providers/names", response_model=ProviderNamesResponse)
 async def get_provider_names() -> ProviderNamesResponse:
     """

autogpt_platform/backend/backend/api/features/library/db.py

@@ -489,7 +489,7 @@ async def update_agent_version_in_library(
     agent_graph_version: int,
 ) -> library_model.LibraryAgent:
     """
-    Updates the agent version in the library for any agent owned by the user.
+    Updates the agent version in the library if useGraphIsActiveVersion is True.
 
     Args:
         user_id: Owner of the LibraryAgent.
@@ -498,31 +498,20 @@ async def update_agent_version_in_library(
 
     Raises:
         DatabaseError: If there's an error with the update.
-        NotFoundError: If no library agent is found for this user and agent.
     """
     logger.debug(
         f"Updating agent version in library for user #{user_id}, "
         f"agent #{agent_graph_id} v{agent_graph_version}"
     )
-    async with transaction() as tx:
-        library_agent = await prisma.models.LibraryAgent.prisma(tx).find_first_or_raise(
+    try:
+        library_agent = await prisma.models.LibraryAgent.prisma().find_first_or_raise(
             where={
                 "userId": user_id,
                 "agentGraphId": agent_graph_id,
+                "useGraphIsActiveVersion": True,
             },
         )
-
-        # Delete any conflicting LibraryAgent for the target version
-        await prisma.models.LibraryAgent.prisma(tx).delete_many(
-            where={
-                "userId": user_id,
-                "agentGraphId": agent_graph_id,
-                "agentGraphVersion": agent_graph_version,
-                "id": {"not": library_agent.id},
-            }
-        )
-
-        lib = await prisma.models.LibraryAgent.prisma(tx).update(
+        lib = await prisma.models.LibraryAgent.prisma().update(
             where={"id": library_agent.id},
             data={
                 "AgentGraph": {
@@ -536,13 +525,13 @@ async def update_agent_version_in_library(
             },
             include={"AgentGraph": True},
         )
+        if lib is None:
+            raise NotFoundError(f"Library agent {library_agent.id} not found")
 
-    if lib is None:
-        raise NotFoundError(
-            f"Failed to update library agent for {agent_graph_id} v{agent_graph_version}"
-        )
-
-    return library_model.LibraryAgent.from_db(lib)
+        return library_model.LibraryAgent.from_db(lib)
+    except prisma.errors.PrismaError as e:
+        logger.error(f"Database error updating agent version in library: {e}")
+        raise DatabaseError("Failed to update agent version in library") from e
 
 
 async def update_library_agent(
@@ -836,7 +825,6 @@ async def add_store_agent_to_library(
                     }
                 },
                 "isCreatedByUser": False,
-                "useGraphIsActiveVersion": False,
                 "settings": SafeJson(
                     _initialize_graph_settings(graph_model).model_dump()
                 ),

autogpt_platform/backend/backend/api/features/library/model.py

@@ -48,7 +48,6 @@ class LibraryAgent(pydantic.BaseModel):
     id: str
     graph_id: str
     graph_version: int
-    owner_user_id: str  # ID of user who owns/created this agent graph
 
     image_url: str | None
 
@@ -164,7 +163,6 @@ class LibraryAgent(pydantic.BaseModel):
             id=agent.id,
             graph_id=agent.agentGraphId,
             graph_version=agent.agentGraphVersion,
-            owner_user_id=agent.userId,
             image_url=agent.imageUrl,
             creator_name=creator_name,
             creator_image_url=creator_image_url,

autogpt_platform/backend/backend/api/features/library/routes/presets.py

@@ -8,6 +8,7 @@ from backend.data.execution import GraphExecutionMeta
 from backend.data.graph import get_graph
 from backend.data.integrations import get_webhook
 from backend.data.model import CredentialsMetaInput
+from backend.data.onboarding import increment_runs
 from backend.executor.utils import add_graph_execution, make_node_credentials_input_map
 from backend.integrations.creds_manager import IntegrationCredentialsManager
 from backend.integrations.webhooks import get_webhook_manager
@@ -402,6 +403,8 @@ async def execute_preset(
     merged_node_input = preset.inputs | inputs
     merged_credential_inputs = preset.credentials | credential_inputs
 
+    await increment_runs(user_id)
+
     return await add_graph_execution(
         user_id=user_id,
         graph_id=preset.graph_id,

autogpt_platform/backend/backend/api/features/library/routes_test.py

@@ -42,7 +42,6 @@ async def test_get_library_agents_success(
                 id="test-agent-1",
                 graph_id="test-agent-1",
                 graph_version=1,
-                owner_user_id=test_user_id,
                 name="Test Agent 1",
                 description="Test Description 1",
                 image_url=None,
@@ -65,7 +64,6 @@ async def test_get_library_agents_success(
                 id="test-agent-2",
                 graph_id="test-agent-2",
                 graph_version=1,
-                owner_user_id=test_user_id,
                 name="Test Agent 2",
                 description="Test Description 2",
                 image_url=None,
@@ -140,7 +138,6 @@ async def test_get_favorite_library_agents_success(
                 id="test-agent-1",
                 graph_id="test-agent-1",
                 graph_version=1,
-                owner_user_id=test_user_id,
                 name="Favorite Agent 1",
                 description="Test Favorite Description 1",
                 image_url=None,
@@ -208,7 +205,6 @@ def test_add_agent_to_library_success(
         id="test-library-agent-id",
         graph_id="test-agent-1",
         graph_version=1,
-        owner_user_id=test_user_id,
         name="Test Agent 1",
         description="Test Description 1",
         image_url=None,

autogpt_platform/backend/backend/api/features/store/backfill_embeddings.py

@@ -0,0 +1,72 @@
+#!/usr/bin/env python3
+"""
+CLI script to backfill embeddings for store agents.
+
+Usage:
+    poetry run python -m backend.server.v2.store.backfill_embeddings [--batch-size N]
+"""
+
+import argparse
+import asyncio
+import sys
+
+import prisma
+
+
+async def main(batch_size: int = 100) -> int:
+    """Run the backfill process."""
+    # Initialize Prisma client
+    client = prisma.Prisma()
+    await client.connect()
+    prisma.register(client)
+
+    try:
+        from backend.api.features.store.embeddings import (
+            backfill_missing_embeddings,
+            get_embedding_stats,
+        )
+
+        # Get current stats
+        print("Current embedding stats:")
+        stats = await get_embedding_stats()
+        print(f"  Total approved: {stats['total_approved']}")
+        print(f"  With embeddings: {stats['with_embeddings']}")
+        print(f"  Without embeddings: {stats['without_embeddings']}")
+        print(f"  Coverage: {stats['coverage_percent']}%")
+
+        if stats["without_embeddings"] == 0:
+            print("\nAll agents already have embeddings. Nothing to do.")
+            return 0
+
+        # Run backfill
+        print(f"\nBackfilling up to {batch_size} embeddings...")
+        result = await backfill_missing_embeddings(batch_size=batch_size)
+        print(f"  Processed: {result['processed']}")
+        print(f"  Success: {result['success']}")
+        print(f"  Failed: {result['failed']}")
+
+        # Get final stats
+        print("\nFinal embedding stats:")
+        stats = await get_embedding_stats()
+        print(f"  Total approved: {stats['total_approved']}")
+        print(f"  With embeddings: {stats['with_embeddings']}")
+        print(f"  Without embeddings: {stats['without_embeddings']}")
+        print(f"  Coverage: {stats['coverage_percent']}%")
+
+        return 0 if result["failed"] == 0 else 1
+
+    finally:
+        await client.disconnect()
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Backfill embeddings for store agents")
+    parser.add_argument(
+        "--batch-size",
+        type=int,
+        default=100,
+        help="Number of embeddings to generate (default: 100)",
+    )
+    args = parser.parse_args()
+
+    sys.exit(asyncio.run(main(batch_size=args.batch_size)))

autogpt_platform/backend/backend/api/features/store/content_handlers.py

@@ -1,431 +0,0 @@
-"""
-Content Type Handlers for Unified Embeddings
-
-Pluggable system for different content sources (store agents, blocks, docs).
-Each handler knows how to fetch and process its content type for embedding.
-"""
-
-import logging
-from abc import ABC, abstractmethod
-from dataclasses import dataclass
-from pathlib import Path
-from typing import Any
-
-from prisma.enums import ContentType
-
-from backend.data.db import query_raw_with_schema
-
-logger = logging.getLogger(__name__)
-
-
-@dataclass
-class ContentItem:
-    """Represents a piece of content to be embedded."""
-
-    content_id: str  # Unique identifier (DB ID or file path)
-    content_type: ContentType
-    searchable_text: str  # Combined text for embedding
-    metadata: dict[str, Any]  # Content-specific metadata
-    user_id: str | None = None  # For user-scoped content
-
-
-class ContentHandler(ABC):
-    """Base handler for fetching and processing content for embeddings."""
-
-    @property
-    @abstractmethod
-    def content_type(self) -> ContentType:
-        """The ContentType this handler manages."""
-        pass
-
-    @abstractmethod
-    async def get_missing_items(self, batch_size: int) -> list[ContentItem]:
-        """
-        Fetch items that don't have embeddings yet.
-
-        Args:
-            batch_size: Maximum number of items to return
-
-        Returns:
-            List of ContentItem objects ready for embedding
-        """
-        pass
-
-    @abstractmethod
-    async def get_stats(self) -> dict[str, int]:
-        """
-        Get statistics about embedding coverage.
-
-        Returns:
-            Dict with keys: total, with_embeddings, without_embeddings
-        """
-        pass
-
-
-class StoreAgentHandler(ContentHandler):
-    """Handler for marketplace store agent listings."""
-
-    @property
-    def content_type(self) -> ContentType:
-        return ContentType.STORE_AGENT
-
-    async def get_missing_items(self, batch_size: int) -> list[ContentItem]:
-        """Fetch approved store listings without embeddings."""
-        from backend.api.features.store.embeddings import build_searchable_text
-
-        missing = await query_raw_with_schema(
-            """
-            SELECT
-                slv.id,
-                slv.name,
-                slv.description,
-                slv."subHeading",
-                slv.categories
-            FROM {schema_prefix}"StoreListingVersion" slv
-            LEFT JOIN {schema_prefix}"UnifiedContentEmbedding" uce
-                ON slv.id = uce."contentId" AND uce."contentType" = 'STORE_AGENT'::{schema_prefix}"ContentType"
-            WHERE slv."submissionStatus" = 'APPROVED'
-            AND slv."isDeleted" = false
-            AND uce."contentId" IS NULL
-            LIMIT $1
-            """,
-            batch_size,
-        )
-
-        return [
-            ContentItem(
-                content_id=row["id"],
-                content_type=ContentType.STORE_AGENT,
-                searchable_text=build_searchable_text(
-                    name=row["name"],
-                    description=row["description"],
-                    sub_heading=row["subHeading"],
-                    categories=row["categories"] or [],
-                ),
-                metadata={
-                    "name": row["name"],
-                    "categories": row["categories"] or [],
-                },
-                user_id=None,  # Store agents are public
-            )
-            for row in missing
-        ]
-
-    async def get_stats(self) -> dict[str, int]:
-        """Get statistics about store agent embedding coverage."""
-        # Count approved versions
-        approved_result = await query_raw_with_schema(
-            """
-            SELECT COUNT(*) as count
-            FROM {schema_prefix}"StoreListingVersion"
-            WHERE "submissionStatus" = 'APPROVED'
-            AND "isDeleted" = false
-            """
-        )
-        total_approved = approved_result[0]["count"] if approved_result else 0
-
-        # Count versions with embeddings
-        embedded_result = await query_raw_with_schema(
-            """
-            SELECT COUNT(*) as count
-            FROM {schema_prefix}"StoreListingVersion" slv
-            JOIN {schema_prefix}"UnifiedContentEmbedding" uce ON slv.id = uce."contentId" AND uce."contentType" = 'STORE_AGENT'::{schema_prefix}"ContentType"
-            WHERE slv."submissionStatus" = 'APPROVED'
-            AND slv."isDeleted" = false
-            """
-        )
-        with_embeddings = embedded_result[0]["count"] if embedded_result else 0
-
-        return {
-            "total": total_approved,
-            "with_embeddings": with_embeddings,
-            "without_embeddings": total_approved - with_embeddings,
-        }
-
-
-class BlockHandler(ContentHandler):
-    """Handler for block definitions (Python classes)."""
-
-    @property
-    def content_type(self) -> ContentType:
-        return ContentType.BLOCK
-
-    async def get_missing_items(self, batch_size: int) -> list[ContentItem]:
-        """Fetch blocks without embeddings."""
-        from backend.data.block import get_blocks
-
-        # Get all available blocks
-        all_blocks = get_blocks()
-
-        # Check which ones have embeddings
-        if not all_blocks:
-            return []
-
-        block_ids = list(all_blocks.keys())
-
-        # Query for existing embeddings
-        placeholders = ",".join([f"${i+1}" for i in range(len(block_ids))])
-        existing_result = await query_raw_with_schema(
-            f"""
-            SELECT "contentId"
-            FROM {{schema_prefix}}"UnifiedContentEmbedding"
-            WHERE "contentType" = 'BLOCK'::{{schema_prefix}}"ContentType"
-            AND "contentId" = ANY(ARRAY[{placeholders}])
-            """,
-            *block_ids,
-        )
-
-        existing_ids = {row["contentId"] for row in existing_result}
-        missing_blocks = [
-            (block_id, block_cls)
-            for block_id, block_cls in all_blocks.items()
-            if block_id not in existing_ids
-        ]
-
-        # Convert to ContentItem
-        items = []
-        for block_id, block_cls in missing_blocks[:batch_size]:
-            try:
-                block_instance = block_cls()
-
-                # Build searchable text from block metadata
-                parts = []
-                if hasattr(block_instance, "name") and block_instance.name:
-                    parts.append(block_instance.name)
-                if (
-                    hasattr(block_instance, "description")
-                    and block_instance.description
-                ):
-                    parts.append(block_instance.description)
-                if hasattr(block_instance, "categories") and block_instance.categories:
-                    # Convert BlockCategory enum to strings
-                    parts.append(
-                        " ".join(str(cat.value) for cat in block_instance.categories)
-                    )
-
-                # Add input/output schema info
-                if hasattr(block_instance, "input_schema"):
-                    schema = block_instance.input_schema
-                    if hasattr(schema, "model_json_schema"):
-                        schema_dict = schema.model_json_schema()
-                        if "properties" in schema_dict:
-                            for prop_name, prop_info in schema_dict[
-                                "properties"
-                            ].items():
-                                if "description" in prop_info:
-                                    parts.append(
-                                        f"{prop_name}: {prop_info['description']}"
-                                    )
-
-                searchable_text = " ".join(parts)
-
-                # Convert categories set of enums to list of strings for JSON serialization
-                categories = getattr(block_instance, "categories", set())
-                categories_list = (
-                    [cat.value for cat in categories] if categories else []
-                )
-
-                items.append(
-                    ContentItem(
-                        content_id=block_id,
-                        content_type=ContentType.BLOCK,
-                        searchable_text=searchable_text,
-                        metadata={
-                            "name": getattr(block_instance, "name", ""),
-                            "categories": categories_list,
-                        },
-                        user_id=None,  # Blocks are public
-                    )
-                )
-            except Exception as e:
-                logger.warning(f"Failed to process block {block_id}: {e}")
-                continue
-
-        return items
-
-    async def get_stats(self) -> dict[str, int]:
-        """Get statistics about block embedding coverage."""
-        from backend.data.block import get_blocks
-
-        all_blocks = get_blocks()
-        total_blocks = len(all_blocks)
-
-        if total_blocks == 0:
-            return {"total": 0, "with_embeddings": 0, "without_embeddings": 0}
-
-        block_ids = list(all_blocks.keys())
-        placeholders = ",".join([f"${i+1}" for i in range(len(block_ids))])
-
-        embedded_result = await query_raw_with_schema(
-            f"""
-            SELECT COUNT(*) as count
-            FROM {{schema_prefix}}"UnifiedContentEmbedding"
-            WHERE "contentType" = 'BLOCK'::{{schema_prefix}}"ContentType"
-            AND "contentId" = ANY(ARRAY[{placeholders}])
-            """,
-            *block_ids,
-        )
-
-        with_embeddings = embedded_result[0]["count"] if embedded_result else 0
-
-        return {
-            "total": total_blocks,
-            "with_embeddings": with_embeddings,
-            "without_embeddings": total_blocks - with_embeddings,
-        }
-
-
-class DocumentationHandler(ContentHandler):
-    """Handler for documentation files (.md/.mdx)."""
-
-    @property
-    def content_type(self) -> ContentType:
-        return ContentType.DOCUMENTATION
-
-    def _get_docs_root(self) -> Path:
-        """Get the documentation root directory."""
-        # content_handlers.py is at: backend/backend/api/features/store/content_handlers.py
-        # Need to go up to project root then into docs/
-        # In container: /app/autogpt_platform/backend/backend/api/features/store -> /app/docs
-        # In development: /repo/autogpt_platform/backend/backend/api/features/store -> /repo/docs
-        this_file = Path(
-            __file__
-        )  # .../backend/backend/api/features/store/content_handlers.py
-        project_root = (
-            this_file.parent.parent.parent.parent.parent.parent.parent
-        )  # -> /app or /repo
-        docs_root = project_root / "docs"
-        return docs_root
-
-    def _extract_title_and_content(self, file_path: Path) -> tuple[str, str]:
-        """Extract title and content from markdown file."""
-        try:
-            content = file_path.read_text(encoding="utf-8")
-
-            # Try to extract title from first # heading
-            lines = content.split("\n")
-            title = ""
-            body_lines = []
-
-            for line in lines:
-                if line.startswith("# ") and not title:
-                    title = line[2:].strip()
-                else:
-                    body_lines.append(line)
-
-            # If no title found, use filename
-            if not title:
-                title = file_path.stem.replace("-", " ").replace("_", " ").title()
-
-            body = "\n".join(body_lines)
-
-            return title, body
-        except Exception as e:
-            logger.warning(f"Failed to read {file_path}: {e}")
-            return file_path.stem, ""
-
-    async def get_missing_items(self, batch_size: int) -> list[ContentItem]:
-        """Fetch documentation files without embeddings."""
-        docs_root = self._get_docs_root()
-
-        if not docs_root.exists():
-            logger.warning(f"Documentation root not found: {docs_root}")
-            return []
-
-        # Find all .md and .mdx files
-        all_docs = list(docs_root.rglob("*.md")) + list(docs_root.rglob("*.mdx"))
-
-        # Get relative paths for content IDs
-        doc_paths = [str(doc.relative_to(docs_root)) for doc in all_docs]
-
-        if not doc_paths:
-            return []
-
-        # Check which ones have embeddings
-        placeholders = ",".join([f"${i+1}" for i in range(len(doc_paths))])
-        existing_result = await query_raw_with_schema(
-            f"""
-            SELECT "contentId"
-            FROM {{schema_prefix}}"UnifiedContentEmbedding"
-            WHERE "contentType" = 'DOCUMENTATION'::{{schema_prefix}}"ContentType"
-            AND "contentId" = ANY(ARRAY[{placeholders}])
-            """,
-            *doc_paths,
-        )
-
-        existing_ids = {row["contentId"] for row in existing_result}
-        missing_docs = [
-            (doc_path, doc_file)
-            for doc_path, doc_file in zip(doc_paths, all_docs)
-            if doc_path not in existing_ids
-        ]
-
-        # Convert to ContentItem
-        items = []
-        for doc_path, doc_file in missing_docs[:batch_size]:
-            try:
-                title, content = self._extract_title_and_content(doc_file)
-
-                # Build searchable text
-                searchable_text = f"{title} {content}"
-
-                items.append(
-                    ContentItem(
-                        content_id=doc_path,
-                        content_type=ContentType.DOCUMENTATION,
-                        searchable_text=searchable_text,
-                        metadata={
-                            "title": title,
-                            "path": doc_path,
-                        },
-                        user_id=None,  # Documentation is public
-                    )
-                )
-            except Exception as e:
-                logger.warning(f"Failed to process doc {doc_path}: {e}")
-                continue
-
-        return items
-
-    async def get_stats(self) -> dict[str, int]:
-        """Get statistics about documentation embedding coverage."""
-        docs_root = self._get_docs_root()
-
-        if not docs_root.exists():
-            return {"total": 0, "with_embeddings": 0, "without_embeddings": 0}
-
-        # Count all .md and .mdx files
-        all_docs = list(docs_root.rglob("*.md")) + list(docs_root.rglob("*.mdx"))
-        total_docs = len(all_docs)
-
-        if total_docs == 0:
-            return {"total": 0, "with_embeddings": 0, "without_embeddings": 0}
-
-        doc_paths = [str(doc.relative_to(docs_root)) for doc in all_docs]
-        placeholders = ",".join([f"${i+1}" for i in range(len(doc_paths))])
-
-        embedded_result = await query_raw_with_schema(
-            f"""
-            SELECT COUNT(*) as count
-            FROM {{schema_prefix}}"UnifiedContentEmbedding"
-            WHERE "contentType" = 'DOCUMENTATION'::{{schema_prefix}}"ContentType"
-            AND "contentId" = ANY(ARRAY[{placeholders}])
-            """,
-            *doc_paths,
-        )
-
-        with_embeddings = embedded_result[0]["count"] if embedded_result else 0
-
-        return {
-            "total": total_docs,
-            "with_embeddings": with_embeddings,
-            "without_embeddings": total_docs - with_embeddings,
-        }
-
-
-# Content handler registry
-CONTENT_HANDLERS: dict[ContentType, ContentHandler] = {
-    ContentType.STORE_AGENT: StoreAgentHandler(),
-    ContentType.BLOCK: BlockHandler(),
-    ContentType.DOCUMENTATION: DocumentationHandler(),
-}

autogpt_platform/backend/backend/api/features/store/content_handlers_integration_test.py

@@ -1,215 +0,0 @@
-"""
-Integration tests for content handlers using real DB.
-
-Run with: poetry run pytest backend/api/features/store/content_handlers_integration_test.py -xvs
-
-These tests use the real database but mock OpenAI calls.
-"""
-
-from unittest.mock import patch
-
-import pytest
-
-from backend.api.features.store.content_handlers import (
-    CONTENT_HANDLERS,
-    BlockHandler,
-    DocumentationHandler,
-    StoreAgentHandler,
-)
-from backend.api.features.store.embeddings import (
-    EMBEDDING_DIM,
-    backfill_all_content_types,
-    ensure_content_embedding,
-    get_embedding_stats,
-)
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_store_agent_handler_real_db():
-    """Test StoreAgentHandler with real database queries."""
-    handler = StoreAgentHandler()
-
-    # Get stats from real DB
-    stats = await handler.get_stats()
-
-    # Stats should have correct structure
-    assert "total" in stats
-    assert "with_embeddings" in stats
-    assert "without_embeddings" in stats
-    assert stats["total"] >= 0
-    assert stats["with_embeddings"] >= 0
-    assert stats["without_embeddings"] >= 0
-
-    # Get missing items (max 1 to keep test fast)
-    items = await handler.get_missing_items(batch_size=1)
-
-    # Items should be list (may be empty if all have embeddings)
-    assert isinstance(items, list)
-
-    if items:
-        item = items[0]
-        assert item.content_id is not None
-        assert item.content_type.value == "STORE_AGENT"
-        assert item.searchable_text != ""
-        assert item.user_id is None
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_block_handler_real_db():
-    """Test BlockHandler with real database queries."""
-    handler = BlockHandler()
-
-    # Get stats from real DB
-    stats = await handler.get_stats()
-
-    # Stats should have correct structure
-    assert "total" in stats
-    assert "with_embeddings" in stats
-    assert "without_embeddings" in stats
-    assert stats["total"] >= 0  # Should have at least some blocks
-    assert stats["with_embeddings"] >= 0
-    assert stats["without_embeddings"] >= 0
-
-    # Get missing items (max 1 to keep test fast)
-    items = await handler.get_missing_items(batch_size=1)
-
-    # Items should be list
-    assert isinstance(items, list)
-
-    if items:
-        item = items[0]
-        assert item.content_id is not None  # Should be block UUID
-        assert item.content_type.value == "BLOCK"
-        assert item.searchable_text != ""
-        assert item.user_id is None
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_documentation_handler_real_fs():
-    """Test DocumentationHandler with real filesystem."""
-    handler = DocumentationHandler()
-
-    # Get stats from real filesystem
-    stats = await handler.get_stats()
-
-    # Stats should have correct structure
-    assert "total" in stats
-    assert "with_embeddings" in stats
-    assert "without_embeddings" in stats
-    assert stats["total"] >= 0
-    assert stats["with_embeddings"] >= 0
-    assert stats["without_embeddings"] >= 0
-
-    # Get missing items (max 1 to keep test fast)
-    items = await handler.get_missing_items(batch_size=1)
-
-    # Items should be list
-    assert isinstance(items, list)
-
-    if items:
-        item = items[0]
-        assert item.content_id is not None  # Should be relative path
-        assert item.content_type.value == "DOCUMENTATION"
-        assert item.searchable_text != ""
-        assert item.user_id is None
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_get_embedding_stats_all_types():
-    """Test get_embedding_stats aggregates all content types."""
-    stats = await get_embedding_stats()
-
-    # Should have structure with by_type and totals
-    assert "by_type" in stats
-    assert "totals" in stats
-
-    # Check each content type is present
-    by_type = stats["by_type"]
-    assert "STORE_AGENT" in by_type
-    assert "BLOCK" in by_type
-    assert "DOCUMENTATION" in by_type
-
-    # Check totals are aggregated
-    totals = stats["totals"]
-    assert totals["total"] >= 0
-    assert totals["with_embeddings"] >= 0
-    assert totals["without_embeddings"] >= 0
-    assert "coverage_percent" in totals
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@patch("backend.api.features.store.embeddings.generate_embedding")
-async def test_ensure_content_embedding_blocks(mock_generate):
-    """Test creating embeddings for blocks (mocked OpenAI)."""
-    # Mock OpenAI to return fake embedding
-    mock_generate.return_value = [0.1] * EMBEDDING_DIM
-
-    # Get one block without embedding
-    handler = BlockHandler()
-    items = await handler.get_missing_items(batch_size=1)
-
-    if not items:
-        pytest.skip("No blocks without embeddings")
-
-    item = items[0]
-
-    # Try to create embedding (OpenAI mocked)
-    result = await ensure_content_embedding(
-        content_type=item.content_type,
-        content_id=item.content_id,
-        searchable_text=item.searchable_text,
-        metadata=item.metadata,
-        user_id=item.user_id,
-    )
-
-    # Should succeed with mocked OpenAI
-    assert result is True
-    mock_generate.assert_called_once()
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@patch("backend.api.features.store.embeddings.generate_embedding")
-async def test_backfill_all_content_types_dry_run(mock_generate):
-    """Test backfill_all_content_types processes all handlers in order."""
-    # Mock OpenAI to return fake embedding
-    mock_generate.return_value = [0.1] * EMBEDDING_DIM
-
-    # Run backfill with batch_size=1 to process max 1 per type
-    result = await backfill_all_content_types(batch_size=1)
-
-    # Should have results for all content types
-    assert "by_type" in result
-    assert "totals" in result
-
-    by_type = result["by_type"]
-    assert "BLOCK" in by_type
-    assert "STORE_AGENT" in by_type
-    assert "DOCUMENTATION" in by_type
-
-    # Each type should have correct structure
-    for content_type, type_result in by_type.items():
-        assert "processed" in type_result
-        assert "success" in type_result
-        assert "failed" in type_result
-
-    # Totals should aggregate
-    totals = result["totals"]
-    assert totals["processed"] >= 0
-    assert totals["success"] >= 0
-    assert totals["failed"] >= 0
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_content_handler_registry():
-    """Test all handlers are registered in correct order."""
-    from prisma.enums import ContentType
-
-    # All three types should be registered
-    assert ContentType.STORE_AGENT in CONTENT_HANDLERS
-    assert ContentType.BLOCK in CONTENT_HANDLERS
-    assert ContentType.DOCUMENTATION in CONTENT_HANDLERS
-
-    # Check handler types
-    assert isinstance(CONTENT_HANDLERS[ContentType.STORE_AGENT], StoreAgentHandler)
-    assert isinstance(CONTENT_HANDLERS[ContentType.BLOCK], BlockHandler)
-    assert isinstance(CONTENT_HANDLERS[ContentType.DOCUMENTATION], DocumentationHandler)

autogpt_platform/backend/backend/api/features/store/content_handlers_test.py

@@ -1,324 +0,0 @@
-"""
-E2E tests for content handlers (blocks, store agents, documentation).
-
-Tests the full flow: discovering content → generating embeddings → storing.
-"""
-
-from pathlib import Path
-from unittest.mock import MagicMock, patch
-
-import pytest
-from prisma.enums import ContentType
-
-from backend.api.features.store.content_handlers import (
-    CONTENT_HANDLERS,
-    BlockHandler,
-    DocumentationHandler,
-    StoreAgentHandler,
-)
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_store_agent_handler_get_missing_items(mocker):
-    """Test StoreAgentHandler fetches approved agents without embeddings."""
-    handler = StoreAgentHandler()
-
-    # Mock database query
-    mock_missing = [
-        {
-            "id": "agent-1",
-            "name": "Test Agent",
-            "description": "A test agent",
-            "subHeading": "Test heading",
-            "categories": ["AI", "Testing"],
-        }
-    ]
-
-    with patch(
-        "backend.api.features.store.content_handlers.query_raw_with_schema",
-        return_value=mock_missing,
-    ):
-        items = await handler.get_missing_items(batch_size=10)
-
-        assert len(items) == 1
-        assert items[0].content_id == "agent-1"
-        assert items[0].content_type == ContentType.STORE_AGENT
-        assert "Test Agent" in items[0].searchable_text
-        assert "A test agent" in items[0].searchable_text
-        assert items[0].metadata["name"] == "Test Agent"
-        assert items[0].user_id is None
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_store_agent_handler_get_stats(mocker):
-    """Test StoreAgentHandler returns correct stats."""
-    handler = StoreAgentHandler()
-
-    # Mock approved count query
-    mock_approved = [{"count": 50}]
-    # Mock embedded count query
-    mock_embedded = [{"count": 30}]
-
-    with patch(
-        "backend.api.features.store.content_handlers.query_raw_with_schema",
-        side_effect=[mock_approved, mock_embedded],
-    ):
-        stats = await handler.get_stats()
-
-        assert stats["total"] == 50
-        assert stats["with_embeddings"] == 30
-        assert stats["without_embeddings"] == 20
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_block_handler_get_missing_items(mocker):
-    """Test BlockHandler discovers blocks without embeddings."""
-    handler = BlockHandler()
-
-    # Mock get_blocks to return test blocks
-    mock_block_class = MagicMock()
-    mock_block_instance = MagicMock()
-    mock_block_instance.name = "Calculator Block"
-    mock_block_instance.description = "Performs calculations"
-    mock_block_instance.categories = [MagicMock(value="MATH")]
-    mock_block_instance.input_schema.model_json_schema.return_value = {
-        "properties": {"expression": {"description": "Math expression to evaluate"}}
-    }
-    mock_block_class.return_value = mock_block_instance
-
-    mock_blocks = {"block-uuid-1": mock_block_class}
-
-    # Mock existing embeddings query (no embeddings exist)
-    mock_existing = []
-
-    with patch(
-        "backend.data.block.get_blocks",
-        return_value=mock_blocks,
-    ):
-        with patch(
-            "backend.api.features.store.content_handlers.query_raw_with_schema",
-            return_value=mock_existing,
-        ):
-            items = await handler.get_missing_items(batch_size=10)
-
-            assert len(items) == 1
-            assert items[0].content_id == "block-uuid-1"
-            assert items[0].content_type == ContentType.BLOCK
-            assert "Calculator Block" in items[0].searchable_text
-            assert "Performs calculations" in items[0].searchable_text
-            assert "MATH" in items[0].searchable_text
-            assert "expression: Math expression" in items[0].searchable_text
-            assert items[0].user_id is None
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_block_handler_get_stats(mocker):
-    """Test BlockHandler returns correct stats."""
-    handler = BlockHandler()
-
-    # Mock get_blocks
-    mock_blocks = {
-        "block-1": MagicMock(),
-        "block-2": MagicMock(),
-        "block-3": MagicMock(),
-    }
-
-    # Mock embedded count query (2 blocks have embeddings)
-    mock_embedded = [{"count": 2}]
-
-    with patch(
-        "backend.data.block.get_blocks",
-        return_value=mock_blocks,
-    ):
-        with patch(
-            "backend.api.features.store.content_handlers.query_raw_with_schema",
-            return_value=mock_embedded,
-        ):
-            stats = await handler.get_stats()
-
-            assert stats["total"] == 3
-            assert stats["with_embeddings"] == 2
-            assert stats["without_embeddings"] == 1
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_documentation_handler_get_missing_items(tmp_path, mocker):
-    """Test DocumentationHandler discovers docs without embeddings."""
-    handler = DocumentationHandler()
-
-    # Create temporary docs directory with test files
-    docs_root = tmp_path / "docs"
-    docs_root.mkdir()
-
-    (docs_root / "guide.md").write_text("# Getting Started\n\nThis is a guide.")
-    (docs_root / "api.mdx").write_text("# API Reference\n\nAPI documentation.")
-
-    # Mock _get_docs_root to return temp dir
-    with patch.object(handler, "_get_docs_root", return_value=docs_root):
-        # Mock existing embeddings query (no embeddings exist)
-        with patch(
-            "backend.api.features.store.content_handlers.query_raw_with_schema",
-            return_value=[],
-        ):
-            items = await handler.get_missing_items(batch_size=10)
-
-            assert len(items) == 2
-
-            # Check guide.md
-            guide_item = next(
-                (item for item in items if item.content_id == "guide.md"), None
-            )
-            assert guide_item is not None
-            assert guide_item.content_type == ContentType.DOCUMENTATION
-            assert "Getting Started" in guide_item.searchable_text
-            assert "This is a guide" in guide_item.searchable_text
-            assert guide_item.metadata["title"] == "Getting Started"
-            assert guide_item.user_id is None
-
-            # Check api.mdx
-            api_item = next(
-                (item for item in items if item.content_id == "api.mdx"), None
-            )
-            assert api_item is not None
-            assert "API Reference" in api_item.searchable_text
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_documentation_handler_get_stats(tmp_path, mocker):
-    """Test DocumentationHandler returns correct stats."""
-    handler = DocumentationHandler()
-
-    # Create temporary docs directory
-    docs_root = tmp_path / "docs"
-    docs_root.mkdir()
-    (docs_root / "doc1.md").write_text("# Doc 1")
-    (docs_root / "doc2.md").write_text("# Doc 2")
-    (docs_root / "doc3.mdx").write_text("# Doc 3")
-
-    # Mock embedded count query (1 doc has embedding)
-    mock_embedded = [{"count": 1}]
-
-    with patch.object(handler, "_get_docs_root", return_value=docs_root):
-        with patch(
-            "backend.api.features.store.content_handlers.query_raw_with_schema",
-            return_value=mock_embedded,
-        ):
-            stats = await handler.get_stats()
-
-            assert stats["total"] == 3
-            assert stats["with_embeddings"] == 1
-            assert stats["without_embeddings"] == 2
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_documentation_handler_title_extraction(tmp_path):
-    """Test DocumentationHandler extracts title from markdown heading."""
-    handler = DocumentationHandler()
-
-    # Test with heading
-    doc_with_heading = tmp_path / "with_heading.md"
-    doc_with_heading.write_text("# My Title\n\nContent here")
-    title, content = handler._extract_title_and_content(doc_with_heading)
-    assert title == "My Title"
-    assert "# My Title" not in content
-    assert "Content here" in content
-
-    # Test without heading
-    doc_without_heading = tmp_path / "no-heading.md"
-    doc_without_heading.write_text("Just content, no heading")
-    title, content = handler._extract_title_and_content(doc_without_heading)
-    assert title == "No Heading"  # Uses filename
-    assert "Just content" in content
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_content_handlers_registry():
-    """Test all content types are registered."""
-    assert ContentType.STORE_AGENT in CONTENT_HANDLERS
-    assert ContentType.BLOCK in CONTENT_HANDLERS
-    assert ContentType.DOCUMENTATION in CONTENT_HANDLERS
-
-    assert isinstance(CONTENT_HANDLERS[ContentType.STORE_AGENT], StoreAgentHandler)
-    assert isinstance(CONTENT_HANDLERS[ContentType.BLOCK], BlockHandler)
-    assert isinstance(CONTENT_HANDLERS[ContentType.DOCUMENTATION], DocumentationHandler)
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_block_handler_handles_missing_attributes():
-    """Test BlockHandler gracefully handles blocks with missing attributes."""
-    handler = BlockHandler()
-
-    # Mock block with minimal attributes
-    mock_block_class = MagicMock()
-    mock_block_instance = MagicMock()
-    mock_block_instance.name = "Minimal Block"
-    # No description, categories, or schema
-    del mock_block_instance.description
-    del mock_block_instance.categories
-    del mock_block_instance.input_schema
-    mock_block_class.return_value = mock_block_instance
-
-    mock_blocks = {"block-minimal": mock_block_class}
-
-    with patch(
-        "backend.data.block.get_blocks",
-        return_value=mock_blocks,
-    ):
-        with patch(
-            "backend.api.features.store.content_handlers.query_raw_with_schema",
-            return_value=[],
-        ):
-            items = await handler.get_missing_items(batch_size=10)
-
-            assert len(items) == 1
-            assert items[0].searchable_text == "Minimal Block"
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_block_handler_skips_failed_blocks():
-    """Test BlockHandler skips blocks that fail to instantiate."""
-    handler = BlockHandler()
-
-    # Mock one good block and one bad block
-    good_block = MagicMock()
-    good_instance = MagicMock()
-    good_instance.name = "Good Block"
-    good_instance.description = "Works fine"
-    good_instance.categories = []
-    good_block.return_value = good_instance
-
-    bad_block = MagicMock()
-    bad_block.side_effect = Exception("Instantiation failed")
-
-    mock_blocks = {"good-block": good_block, "bad-block": bad_block}
-
-    with patch(
-        "backend.data.block.get_blocks",
-        return_value=mock_blocks,
-    ):
-        with patch(
-            "backend.api.features.store.content_handlers.query_raw_with_schema",
-            return_value=[],
-        ):
-            items = await handler.get_missing_items(batch_size=10)
-
-            # Should only get the good block
-            assert len(items) == 1
-            assert items[0].content_id == "good-block"
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_documentation_handler_missing_docs_directory():
-    """Test DocumentationHandler handles missing docs directory gracefully."""
-    handler = DocumentationHandler()
-
-    # Mock _get_docs_root to return non-existent path
-    fake_path = Path("/nonexistent/docs")
-    with patch.object(handler, "_get_docs_root", return_value=fake_path):
-        items = await handler.get_missing_items(batch_size=10)
-        assert items == []
-
-        stats = await handler.get_stats()
-        assert stats["total"] == 0
-        assert stats["with_embeddings"] == 0
-        assert stats["without_embeddings"] == 0

autogpt_platform/backend/backend/api/features/store/db.py

@@ -1,7 +1,7 @@
 import asyncio
 import logging
 from datetime import datetime, timezone
-from typing import Any, Literal
+from typing import Literal
 
 import fastapi
 import prisma.enums
@@ -29,8 +29,6 @@ from backend.util.settings import Settings
 
 from . import exceptions as store_exceptions
 from . import model as store_model
-from .embeddings import ensure_embedding
-from .hybrid_search import hybrid_search
 
 logger = logging.getLogger(__name__)
 settings = Settings()
@@ -51,77 +49,54 @@ async def get_store_agents(
     page_size: int = 20,
 ) -> store_model.StoreAgentsResponse:
     """
-    Get PUBLIC store agents from the StoreAgent view.
-
-    Search behavior:
-    - With search_query: Uses hybrid search (semantic + lexical)
-    - Fallback: If embeddings unavailable, gracefully degrades to lexical-only
-    - Rationale: User-facing endpoint prioritizes availability over accuracy
-
-    Note: Admin operations (approval) use fail-fast to prevent inconsistent state.
+    Get PUBLIC store agents from the StoreAgent view
     """
     logger.debug(
         f"Getting store agents. featured={featured}, creators={creators}, sorted_by={sorted_by}, search={search_query}, category={category}, page={page}"
     )
 
-    search_used_hybrid = False
-    store_agents: list[store_model.StoreAgent] = []
-    agents: list[dict[str, Any]] = []
-    total = 0
-    total_pages = 0
-
     try:
         # If search_query is provided, use hybrid search (embeddings + tsvector)
         if search_query:
-            # Try hybrid search combining semantic and lexical signals
-            # Falls back to lexical-only if OpenAI unavailable (user-facing, high SLA)
-            try:
-                agents, total = await hybrid_search(
-                    query=search_query,
-                    featured=featured,
-                    creators=creators,
-                    category=category,
-                    sorted_by="relevance",  # Use hybrid scoring for relevance
-                    page=page,
-                    page_size=page_size,
-                )
-                search_used_hybrid = True
-            except Exception as e:
-                # Log error but fall back to lexical search for better UX
-                logger.error(
-                    f"Hybrid search failed (likely OpenAI unavailable), "
-                    f"falling back to lexical search: {e}"
-                )
-                # search_used_hybrid remains False, will use fallback path below
+            from backend.api.features.store.hybrid_search import hybrid_search
 
-            # Convert hybrid search results (dict format) if hybrid succeeded
-            if search_used_hybrid:
-                total_pages = (total + page_size - 1) // page_size
-                store_agents: list[store_model.StoreAgent] = []
-                for agent in agents:
-                    try:
-                        store_agent = store_model.StoreAgent(
-                            slug=agent["slug"],
-                            agent_name=agent["agent_name"],
-                            agent_image=(
-                                agent["agent_image"][0] if agent["agent_image"] else ""
-                            ),
-                            creator=agent["creator_username"] or "Needs Profile",
-                            creator_avatar=agent["creator_avatar"] or "",
-                            sub_heading=agent["sub_heading"],
-                            description=agent["description"],
-                            runs=agent["runs"],
-                            rating=agent["rating"],
-                        )
-                        store_agents.append(store_agent)
-                    except Exception as e:
-                        logger.error(
-                            f"Error parsing Store agent from hybrid search results: {e}"
-                        )
-                        continue
+            # Use hybrid search combining semantic and lexical signals
+            agents, total = await hybrid_search(
+                query=search_query,
+                featured=featured,
+                creators=creators,
+                category=category,
+                sorted_by="relevance",  # Use hybrid scoring for relevance
+                page=page,
+                page_size=page_size,
+            )
 
-        if not search_used_hybrid:
-            # Fallback path - use basic search or no search
+            total_pages = (total + page_size - 1) // page_size
+
+            # Convert raw results to StoreAgent models
+            store_agents: list[store_model.StoreAgent] = []
+            for agent in agents:
+                try:
+                    store_agent = store_model.StoreAgent(
+                        slug=agent["slug"],
+                        agent_name=agent["agent_name"],
+                        agent_image=(
+                            agent["agent_image"][0] if agent["agent_image"] else ""
+                        ),
+                        creator=agent["creator_username"] or "Needs Profile",
+                        creator_avatar=agent["creator_avatar"] or "",
+                        sub_heading=agent["sub_heading"],
+                        description=agent["description"],
+                        runs=agent["runs"],
+                        rating=agent["rating"],
+                    )
+                    store_agents.append(store_agent)
+                except Exception as e:
+                    logger.error(f"Error parsing Store agent from search results: {e}")
+                    continue
+
+        else:
+            # Non-search query path (original logic)
             where_clause: prisma.types.StoreAgentWhereInput = {"is_available": True}
             if featured:
                 where_clause["featured"] = featured
@@ -130,14 +105,6 @@ async def get_store_agents(
             if category:
                 where_clause["categories"] = {"has": category}
 
-            # Add basic text search if search_query provided but hybrid failed
-            if search_query:
-                where_clause["OR"] = [
-                    {"agent_name": {"contains": search_query, "mode": "insensitive"}},
-                    {"sub_heading": {"contains": search_query, "mode": "insensitive"}},
-                    {"description": {"contains": search_query, "mode": "insensitive"}},
-                ]
-
             order_by = []
             if sorted_by == "rating":
                 order_by.append({"rating": "desc"})
@@ -146,7 +113,7 @@ async def get_store_agents(
             elif sorted_by == "name":
                 order_by.append({"agent_name": "asc"})
 
-            db_agents = await prisma.models.StoreAgent.prisma().find_many(
+            agents = await prisma.models.StoreAgent.prisma().find_many(
                 where=where_clause,
                 order=order_by,
                 skip=(page - 1) * page_size,
@@ -157,7 +124,7 @@ async def get_store_agents(
             total_pages = (total + page_size - 1) // page_size
 
             store_agents: list[store_model.StoreAgent] = []
-            for agent in db_agents:
+            for agent in agents:
                 try:
                     # Create the StoreAgent object safely
                     store_agent = store_model.StoreAgent(
@@ -572,7 +539,6 @@ async def get_store_submissions(
         submission_models = []
         for sub in submissions:
             submission_model = store_model.StoreSubmission(
-                listing_id=sub.listing_id,
                 agent_id=sub.agent_id,
                 agent_version=sub.agent_version,
                 name=sub.name,
@@ -626,48 +592,35 @@ async def delete_store_submission(
     submission_id: str,
 ) -> bool:
     """
-    Delete a store submission version as the submitting user.
+    Delete a store listing submission as the submitting user.
 
     Args:
         user_id: ID of the authenticated user
-        submission_id: StoreListingVersion ID to delete
+        submission_id: ID of the submission to be deleted
 
     Returns:
-        bool: True if successfully deleted
+        bool: True if the submission was successfully deleted, False otherwise
     """
+    logger.debug(f"Deleting store submission {submission_id} for user {user_id}")
+
     try:
-        # Find the submission version with ownership check
-        version = await prisma.models.StoreListingVersion.prisma().find_first(
-            where={"id": submission_id}, include={"StoreListing": True}
+        # Verify the submission belongs to this user
+        submission = await prisma.models.StoreListing.prisma().find_first(
+            where={"agentGraphId": submission_id, "owningUserId": user_id}
         )
 
-        if (
-            not version
-            or not version.StoreListing
-            or version.StoreListing.owningUserId != user_id
-        ):
-            raise store_exceptions.SubmissionNotFoundError("Submission not found")
-
-        # Prevent deletion of approved submissions
-        if version.submissionStatus == prisma.enums.SubmissionStatus.APPROVED:
-            raise store_exceptions.InvalidOperationError(
-                "Cannot delete approved submissions"
+        if not submission:
+            logger.warning(f"Submission not found for user {user_id}: {submission_id}")
+            raise store_exceptions.SubmissionNotFoundError(
+                f"Submission not found for this user. User ID: {user_id}, Submission ID: {submission_id}"
             )
 
-        # Delete the version
-        await prisma.models.StoreListingVersion.prisma().delete(
-            where={"id": version.id}
-        )
+        # Delete the submission
+        await prisma.models.StoreListing.prisma().delete(where={"id": submission.id})
 
-        # Clean up empty listing if this was the last version
-        remaining = await prisma.models.StoreListingVersion.prisma().count(
-            where={"storeListingId": version.storeListingId}
+        logger.debug(
+            f"Successfully deleted submission {submission_id} for user {user_id}"
         )
-        if remaining == 0:
-            await prisma.models.StoreListing.prisma().delete(
-                where={"id": version.storeListingId}
-            )
-
         return True
 
     except Exception as e:
@@ -731,15 +684,9 @@ async def create_store_submission(
             logger.warning(
                 f"Agent not found for user {user_id}: {agent_id} v{agent_version}"
             )
-            # Provide more user-friendly error message when agent_id is empty
-            if not agent_id or agent_id.strip() == "":
-                raise store_exceptions.AgentNotFoundError(
-                    "No agent selected. Please select an agent before submitting to the store."
-                )
-            else:
-                raise store_exceptions.AgentNotFoundError(
-                    f"Agent not found for this user. User ID: {user_id}, Agent ID: {agent_id}, Version: {agent_version}"
-                )
+            raise store_exceptions.AgentNotFoundError(
+                f"Agent not found for this user. User ID: {user_id}, Agent ID: {agent_id}, Version: {agent_version}"
+            )
 
         # Check if listing already exists for this agent
         existing_listing = await prisma.models.StoreListing.prisma().find_first(
@@ -811,7 +758,6 @@ async def create_store_submission(
         logger.debug(f"Created store listing for agent {agent_id}")
         # Return submission details
         return store_model.StoreSubmission(
-            listing_id=listing.id,
             agent_id=agent_id,
             agent_version=agent_version,
             name=name,
@@ -923,56 +869,81 @@ async def edit_store_submission(
         # Currently we are not allowing user to update the agent associated with a submission
         # If we allow it in future, then we need a check here to verify the agent belongs to this user.
 
-        # Only allow editing of PENDING submissions
-        if current_version.submissionStatus != prisma.enums.SubmissionStatus.PENDING:
+        # Check if we can edit this submission
+        if current_version.submissionStatus == prisma.enums.SubmissionStatus.REJECTED:
             raise store_exceptions.InvalidOperationError(
-                f"Cannot edit a {current_version.submissionStatus.value.lower()} submission. Only pending submissions can be edited."
+                "Cannot edit a rejected submission"
+            )
+
+        # For APPROVED submissions, we need to create a new version
+        if current_version.submissionStatus == prisma.enums.SubmissionStatus.APPROVED:
+            # Create a new version for the existing listing
+            return await create_store_version(
+                user_id=user_id,
+                agent_id=current_version.agentGraphId,
+                agent_version=current_version.agentGraphVersion,
+                store_listing_id=current_version.storeListingId,
+                name=name,
+                video_url=video_url,
+                agent_output_demo_url=agent_output_demo_url,
+                image_urls=image_urls,
+                description=description,
+                sub_heading=sub_heading,
+                categories=categories,
+                changes_summary=changes_summary,
+                recommended_schedule_cron=recommended_schedule_cron,
+                instructions=instructions,
             )
 
         # For PENDING submissions, we can update the existing version
-        # Update the existing version
-        updated_version = await prisma.models.StoreListingVersion.prisma().update(
-            where={"id": store_listing_version_id},
-            data=prisma.types.StoreListingVersionUpdateInput(
+        elif current_version.submissionStatus == prisma.enums.SubmissionStatus.PENDING:
+            # Update the existing version
+            updated_version = await prisma.models.StoreListingVersion.prisma().update(
+                where={"id": store_listing_version_id},
+                data=prisma.types.StoreListingVersionUpdateInput(
+                    name=name,
+                    videoUrl=video_url,
+                    agentOutputDemoUrl=agent_output_demo_url,
+                    imageUrls=image_urls,
+                    description=description,
+                    categories=categories,
+                    subHeading=sub_heading,
+                    changesSummary=changes_summary,
+                    recommendedScheduleCron=recommended_schedule_cron,
+                    instructions=instructions,
+                ),
+            )
+
+            logger.debug(
+                f"Updated existing version {store_listing_version_id} for agent {current_version.agentGraphId}"
+            )
+
+            if not updated_version:
+                raise DatabaseError("Failed to update store listing version")
+            return store_model.StoreSubmission(
+                agent_id=current_version.agentGraphId,
+                agent_version=current_version.agentGraphVersion,
                 name=name,
-                videoUrl=video_url,
-                agentOutputDemoUrl=agent_output_demo_url,
-                imageUrls=image_urls,
+                sub_heading=sub_heading,
+                slug=current_version.StoreListing.slug,
                 description=description,
-                categories=categories,
-                subHeading=sub_heading,
-                changesSummary=changes_summary,
-                recommendedScheduleCron=recommended_schedule_cron,
                 instructions=instructions,
-            ),
-        )
+                image_urls=image_urls,
+                date_submitted=updated_version.submittedAt or updated_version.createdAt,
+                status=updated_version.submissionStatus,
+                runs=0,
+                rating=0.0,
+                store_listing_version_id=updated_version.id,
+                changes_summary=changes_summary,
+                video_url=video_url,
+                categories=categories,
+                version=updated_version.version,
+            )
 
-        logger.debug(
-            f"Updated existing version {store_listing_version_id} for agent {current_version.agentGraphId}"
-        )
-
-        if not updated_version:
-            raise DatabaseError("Failed to update store listing version")
-        return store_model.StoreSubmission(
-            listing_id=current_version.StoreListing.id,
-            agent_id=current_version.agentGraphId,
-            agent_version=current_version.agentGraphVersion,
-            name=name,
-            sub_heading=sub_heading,
-            slug=current_version.StoreListing.slug,
-            description=description,
-            instructions=instructions,
-            image_urls=image_urls,
-            date_submitted=updated_version.submittedAt or updated_version.createdAt,
-            status=updated_version.submissionStatus,
-            runs=0,
-            rating=0.0,
-            store_listing_version_id=updated_version.id,
-            changes_summary=changes_summary,
-            video_url=video_url,
-            categories=categories,
-            version=updated_version.version,
-        )
+        else:
+            raise store_exceptions.InvalidOperationError(
+                f"Cannot edit submission with status: {current_version.submissionStatus}"
+            )
 
     except (
         store_exceptions.SubmissionNotFoundError,
@@ -1051,78 +1022,38 @@ async def create_store_version(
                 f"Agent not found for this user. User ID: {user_id}, Agent ID: {agent_id}, Version: {agent_version}"
             )
 
-        # Check if there's already a PENDING submission for this agent (any version)
-        existing_pending_submission = (
-            await prisma.models.StoreListingVersion.prisma().find_first(
-                where=prisma.types.StoreListingVersionWhereInput(
-                    storeListingId=store_listing_id,
-                    agentGraphId=agent_id,
-                    submissionStatus=prisma.enums.SubmissionStatus.PENDING,
-                    isDeleted=False,
-                )
+        # Get the latest version number
+        latest_version = listing.Versions[0] if listing.Versions else None
+
+        next_version = (latest_version.version + 1) if latest_version else 1
+
+        # Create a new version for the existing listing
+        new_version = await prisma.models.StoreListingVersion.prisma().create(
+            data=prisma.types.StoreListingVersionCreateInput(
+                version=next_version,
+                agentGraphId=agent_id,
+                agentGraphVersion=agent_version,
+                name=name,
+                videoUrl=video_url,
+                agentOutputDemoUrl=agent_output_demo_url,
+                imageUrls=image_urls,
+                description=description,
+                instructions=instructions,
+                categories=categories,
+                subHeading=sub_heading,
+                submissionStatus=prisma.enums.SubmissionStatus.PENDING,
+                submittedAt=datetime.now(),
+                changesSummary=changes_summary,
+                recommendedScheduleCron=recommended_schedule_cron,
+                storeListingId=store_listing_id,
             )
         )
 
-        # Handle existing pending submission and create new one atomically
-        async with transaction() as tx:
-            # Get the latest version number first
-            latest_listing = await prisma.models.StoreListing.prisma(tx).find_first(
-                where=prisma.types.StoreListingWhereInput(
-                    id=store_listing_id, owningUserId=user_id
-                ),
-                include={"Versions": {"order_by": {"version": "desc"}, "take": 1}},
-            )
-
-            if not latest_listing:
-                raise store_exceptions.ListingNotFoundError(
-                    f"Store listing not found. User ID: {user_id}, Listing ID: {store_listing_id}"
-                )
-
-            latest_version = (
-                latest_listing.Versions[0] if latest_listing.Versions else None
-            )
-            next_version = (latest_version.version + 1) if latest_version else 1
-
-            # If there's an existing pending submission, delete it atomically before creating new one
-            if existing_pending_submission:
-                logger.info(
-                    f"Found existing PENDING submission for agent {agent_id} (was v{existing_pending_submission.agentGraphVersion}, now v{agent_version}), replacing existing submission instead of creating duplicate"
-                )
-                await prisma.models.StoreListingVersion.prisma(tx).delete(
-                    where={"id": existing_pending_submission.id}
-                )
-                logger.debug(
-                    f"Deleted existing pending submission {existing_pending_submission.id}"
-                )
-
-            # Create a new version for the existing listing
-            new_version = await prisma.models.StoreListingVersion.prisma(tx).create(
-                data=prisma.types.StoreListingVersionCreateInput(
-                    version=next_version,
-                    agentGraphId=agent_id,
-                    agentGraphVersion=agent_version,
-                    name=name,
-                    videoUrl=video_url,
-                    agentOutputDemoUrl=agent_output_demo_url,
-                    imageUrls=image_urls,
-                    description=description,
-                    instructions=instructions,
-                    categories=categories,
-                    subHeading=sub_heading,
-                    submissionStatus=prisma.enums.SubmissionStatus.PENDING,
-                    submittedAt=datetime.now(),
-                    changesSummary=changes_summary,
-                    recommendedScheduleCron=recommended_schedule_cron,
-                    storeListingId=store_listing_id,
-                )
-            )
-
         logger.debug(
             f"Created new version for listing {store_listing_id} of agent {agent_id}"
         )
         # Return submission details
         return store_model.StoreSubmission(
-            listing_id=listing.id,
             agent_id=agent_id,
             agent_version=agent_version,
             name=name,
@@ -1535,7 +1466,7 @@ async def review_store_submission(
                 )
 
                 # Update the AgentGraph with store listing data
-                await prisma.models.AgentGraph.prisma(tx).update(
+                await prisma.models.AgentGraph.prisma().update(
                     where={
                         "graphVersionId": {
                             "id": store_listing_version.agentGraphId,
@@ -1550,23 +1481,6 @@ async def review_store_submission(
                     },
                 )
 
-                # Generate embedding for approved listing (blocking - admin operation)
-                # Inside transaction: if embedding fails, entire transaction rolls back
-                embedding_success = await ensure_embedding(
-                    version_id=store_listing_version_id,
-                    name=store_listing_version.name,
-                    description=store_listing_version.description,
-                    sub_heading=store_listing_version.subHeading,
-                    categories=store_listing_version.categories or [],
-                    tx=tx,
-                )
-                if not embedding_success:
-                    raise ValueError(
-                        f"Failed to generate embedding for listing {store_listing_version_id}. "
-                        "This is likely due to OpenAI API being unavailable. "
-                        "Please try again later or contact support if the issue persists."
-                    )
-
                 await prisma.models.StoreListing.prisma(tx).update(
                     where={"id": store_listing_version.StoreListing.id},
                     data={
@@ -1575,6 +1489,24 @@ async def review_store_submission(
                     },
                 )
 
+            # Generate embedding for approved listing (non-blocking)
+            try:
+                from backend.api.features.store.embeddings import ensure_embedding
+
+                await ensure_embedding(
+                    version_id=store_listing_version_id,
+                    name=store_listing_version.name,
+                    description=store_listing_version.description,
+                    sub_heading=store_listing_version.subHeading,
+                    categories=store_listing_version.categories or [],
+                )
+            except Exception as e:
+                # Don't fail approval if embedding generation fails
+                logger.warning(
+                    f"Failed to generate embedding for approved listing "
+                    f"{store_listing_version_id}: {e}"
+                )
+
         # If rejecting an approved agent, update the StoreListing accordingly
         if is_rejecting_approved:
             # Check if there are other approved versions
@@ -1719,12 +1651,15 @@ async def review_store_submission(
 
         # Convert to Pydantic model for consistency
         return store_model.StoreSubmission(
-            listing_id=(submission.StoreListing.id if submission.StoreListing else ""),
             agent_id=submission.agentGraphId,
             agent_version=submission.agentGraphVersion,
             name=submission.name,
             sub_heading=submission.subHeading,
-            slug=(submission.StoreListing.slug if submission.StoreListing else ""),
+            slug=(
+                submission.StoreListing.slug
+                if hasattr(submission, "storeListing") and submission.StoreListing
+                else ""
+            ),
             description=submission.description,
             instructions=submission.instructions,
             image_urls=submission.imageUrls or [],
@@ -1826,7 +1761,9 @@ async def get_admin_listings_with_versions(
         where = prisma.types.StoreListingWhereInput(**where_dict)
         include = prisma.types.StoreListingInclude(
             Versions=prisma.types.FindManyStoreListingVersionArgsFromStoreListing(
-                order_by={"version": "desc"}
+                order_by=prisma.types._StoreListingVersion_version_OrderByInput(
+                    version="desc"
+                )
             ),
             OwningUser=True,
         )
@@ -1851,7 +1788,6 @@ async def get_admin_listings_with_versions(
             # If we have versions, turn them into StoreSubmission models
             for version in listing.Versions or []:
                 version_model = store_model.StoreSubmission(
-                    listing_id=listing.id,
                     agent_id=version.agentGraphId,
                     agent_version=version.agentGraphVersion,
                     name=version.name,

autogpt_platform/backend/backend/api/features/store/embeddings_e2e_test.py

@@ -1,666 +0,0 @@
-"""
-End-to-end database tests for embeddings and hybrid search.
-
-These tests hit the actual database to verify SQL queries work correctly.
-Tests cover:
-1. Embedding storage (store_content_embedding)
-2. Embedding retrieval (get_content_embedding)
-3. Embedding deletion (delete_content_embedding)
-4. Unified hybrid search across content types
-5. Store agent hybrid search
-"""
-
-import uuid
-from typing import AsyncGenerator
-
-import pytest
-from prisma.enums import ContentType
-
-from backend.api.features.store import embeddings
-from backend.api.features.store.embeddings import EMBEDDING_DIM
-from backend.api.features.store.hybrid_search import (
-    hybrid_search,
-    unified_hybrid_search,
-)
-
-# ============================================================================
-# Test Fixtures
-# ============================================================================
-
-
-@pytest.fixture
-def test_content_id() -> str:
-    """Generate unique content ID for test isolation."""
-    return f"test-content-{uuid.uuid4()}"
-
-
-@pytest.fixture
-def test_user_id() -> str:
-    """Generate unique user ID for test isolation."""
-    return f"test-user-{uuid.uuid4()}"
-
-
-@pytest.fixture
-def mock_embedding() -> list[float]:
-    """Generate a mock embedding vector."""
-    # Create a normalized embedding vector
-    import math
-
-    raw = [float(i % 10) / 10.0 for i in range(EMBEDDING_DIM)]
-    # Normalize to unit length (required for cosine similarity)
-    magnitude = math.sqrt(sum(x * x for x in raw))
-    return [x / magnitude for x in raw]
-
-
-@pytest.fixture
-def similar_embedding() -> list[float]:
-    """Generate an embedding similar to mock_embedding."""
-    import math
-
-    # Similar but slightly different values
-    raw = [float(i % 10) / 10.0 + 0.01 for i in range(EMBEDDING_DIM)]
-    magnitude = math.sqrt(sum(x * x for x in raw))
-    return [x / magnitude for x in raw]
-
-
-@pytest.fixture
-def different_embedding() -> list[float]:
-    """Generate an embedding very different from mock_embedding."""
-    import math
-
-    # Reversed pattern to be maximally different
-    raw = [float((EMBEDDING_DIM - i) % 10) / 10.0 for i in range(EMBEDDING_DIM)]
-    magnitude = math.sqrt(sum(x * x for x in raw))
-    return [x / magnitude for x in raw]
-
-
-@pytest.fixture
-async def cleanup_embeddings(
-    server,
-) -> AsyncGenerator[list[tuple[ContentType, str, str | None]], None]:
-    """
-    Fixture that tracks created embeddings and cleans them up after tests.
-
-    Yields a list to which tests can append (content_type, content_id, user_id) tuples.
-    """
-    created_embeddings: list[tuple[ContentType, str, str | None]] = []
-    yield created_embeddings
-
-    # Cleanup all created embeddings
-    for content_type, content_id, user_id in created_embeddings:
-        try:
-            await embeddings.delete_content_embedding(content_type, content_id, user_id)
-        except Exception:
-            pass  # Ignore cleanup errors
-
-
-# ============================================================================
-# store_content_embedding Tests
-# ============================================================================
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_store_content_embedding_store_agent(
-    server,
-    test_content_id: str,
-    mock_embedding: list[float],
-    cleanup_embeddings: list,
-):
-    """Test storing embedding for STORE_AGENT content type."""
-    # Track for cleanup
-    cleanup_embeddings.append((ContentType.STORE_AGENT, test_content_id, None))
-
-    result = await embeddings.store_content_embedding(
-        content_type=ContentType.STORE_AGENT,
-        content_id=test_content_id,
-        embedding=mock_embedding,
-        searchable_text="AI assistant for productivity tasks",
-        metadata={"name": "Test Agent", "categories": ["productivity"]},
-        user_id=None,  # Store agents are public
-    )
-
-    assert result is True
-
-    # Verify it was stored
-    stored = await embeddings.get_content_embedding(
-        ContentType.STORE_AGENT, test_content_id, user_id=None
-    )
-    assert stored is not None
-    assert stored["contentId"] == test_content_id
-    assert stored["contentType"] == "STORE_AGENT"
-    assert stored["searchableText"] == "AI assistant for productivity tasks"
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_store_content_embedding_block(
-    server,
-    test_content_id: str,
-    mock_embedding: list[float],
-    cleanup_embeddings: list,
-):
-    """Test storing embedding for BLOCK content type."""
-    cleanup_embeddings.append((ContentType.BLOCK, test_content_id, None))
-
-    result = await embeddings.store_content_embedding(
-        content_type=ContentType.BLOCK,
-        content_id=test_content_id,
-        embedding=mock_embedding,
-        searchable_text="HTTP request block for API calls",
-        metadata={"name": "HTTP Request Block"},
-        user_id=None,  # Blocks are public
-    )
-
-    assert result is True
-
-    stored = await embeddings.get_content_embedding(
-        ContentType.BLOCK, test_content_id, user_id=None
-    )
-    assert stored is not None
-    assert stored["contentType"] == "BLOCK"
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_store_content_embedding_documentation(
-    server,
-    test_content_id: str,
-    mock_embedding: list[float],
-    cleanup_embeddings: list,
-):
-    """Test storing embedding for DOCUMENTATION content type."""
-    cleanup_embeddings.append((ContentType.DOCUMENTATION, test_content_id, None))
-
-    result = await embeddings.store_content_embedding(
-        content_type=ContentType.DOCUMENTATION,
-        content_id=test_content_id,
-        embedding=mock_embedding,
-        searchable_text="Getting started guide for AutoGPT platform",
-        metadata={"title": "Getting Started", "url": "/docs/getting-started"},
-        user_id=None,  # Docs are public
-    )
-
-    assert result is True
-
-    stored = await embeddings.get_content_embedding(
-        ContentType.DOCUMENTATION, test_content_id, user_id=None
-    )
-    assert stored is not None
-    assert stored["contentType"] == "DOCUMENTATION"
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_store_content_embedding_upsert(
-    server,
-    test_content_id: str,
-    mock_embedding: list[float],
-    cleanup_embeddings: list,
-):
-    """Test that storing embedding twice updates instead of duplicates."""
-    cleanup_embeddings.append((ContentType.BLOCK, test_content_id, None))
-
-    # Store first time
-    result1 = await embeddings.store_content_embedding(
-        content_type=ContentType.BLOCK,
-        content_id=test_content_id,
-        embedding=mock_embedding,
-        searchable_text="Original text",
-        metadata={"version": 1},
-        user_id=None,
-    )
-    assert result1 is True
-
-    # Store again with different text (upsert)
-    result2 = await embeddings.store_content_embedding(
-        content_type=ContentType.BLOCK,
-        content_id=test_content_id,
-        embedding=mock_embedding,
-        searchable_text="Updated text",
-        metadata={"version": 2},
-        user_id=None,
-    )
-    assert result2 is True
-
-    # Verify only one record with updated text
-    stored = await embeddings.get_content_embedding(
-        ContentType.BLOCK, test_content_id, user_id=None
-    )
-    assert stored is not None
-    assert stored["searchableText"] == "Updated text"
-
-
-# ============================================================================
-# get_content_embedding Tests
-# ============================================================================
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_get_content_embedding_not_found(server):
-    """Test retrieving non-existent embedding returns None."""
-    result = await embeddings.get_content_embedding(
-        ContentType.STORE_AGENT, "non-existent-id", user_id=None
-    )
-    assert result is None
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_get_content_embedding_with_metadata(
-    server,
-    test_content_id: str,
-    mock_embedding: list[float],
-    cleanup_embeddings: list,
-):
-    """Test that metadata is correctly stored and retrieved."""
-    cleanup_embeddings.append((ContentType.STORE_AGENT, test_content_id, None))
-
-    metadata = {
-        "name": "Test Agent",
-        "subHeading": "A test agent",
-        "categories": ["ai", "productivity"],
-        "customField": 123,
-    }
-
-    await embeddings.store_content_embedding(
-        content_type=ContentType.STORE_AGENT,
-        content_id=test_content_id,
-        embedding=mock_embedding,
-        searchable_text="test",
-        metadata=metadata,
-        user_id=None,
-    )
-
-    stored = await embeddings.get_content_embedding(
-        ContentType.STORE_AGENT, test_content_id, user_id=None
-    )
-
-    assert stored is not None
-    assert stored["metadata"]["name"] == "Test Agent"
-    assert stored["metadata"]["categories"] == ["ai", "productivity"]
-    assert stored["metadata"]["customField"] == 123
-
-
-# ============================================================================
-# delete_content_embedding Tests
-# ============================================================================
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_delete_content_embedding(
-    server,
-    test_content_id: str,
-    mock_embedding: list[float],
-):
-    """Test deleting embedding removes it from database."""
-    # Store embedding
-    await embeddings.store_content_embedding(
-        content_type=ContentType.BLOCK,
-        content_id=test_content_id,
-        embedding=mock_embedding,
-        searchable_text="To be deleted",
-        metadata=None,
-        user_id=None,
-    )
-
-    # Verify it exists
-    stored = await embeddings.get_content_embedding(
-        ContentType.BLOCK, test_content_id, user_id=None
-    )
-    assert stored is not None
-
-    # Delete it
-    result = await embeddings.delete_content_embedding(
-        ContentType.BLOCK, test_content_id, user_id=None
-    )
-    assert result is True
-
-    # Verify it's gone
-    stored = await embeddings.get_content_embedding(
-        ContentType.BLOCK, test_content_id, user_id=None
-    )
-    assert stored is None
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_delete_content_embedding_not_found(server):
-    """Test deleting non-existent embedding doesn't error."""
-    result = await embeddings.delete_content_embedding(
-        ContentType.BLOCK, "non-existent-id", user_id=None
-    )
-    # Should succeed even if nothing to delete
-    assert result is True
-
-
-# ============================================================================
-# unified_hybrid_search Tests
-# ============================================================================
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_unified_hybrid_search_finds_matching_content(
-    server,
-    mock_embedding: list[float],
-    cleanup_embeddings: list,
-):
-    """Test unified search finds content matching the query."""
-    # Create unique content IDs
-    agent_id = f"test-agent-{uuid.uuid4()}"
-    block_id = f"test-block-{uuid.uuid4()}"
-    doc_id = f"test-doc-{uuid.uuid4()}"
-
-    cleanup_embeddings.append((ContentType.STORE_AGENT, agent_id, None))
-    cleanup_embeddings.append((ContentType.BLOCK, block_id, None))
-    cleanup_embeddings.append((ContentType.DOCUMENTATION, doc_id, None))
-
-    # Store embeddings for different content types
-    await embeddings.store_content_embedding(
-        content_type=ContentType.STORE_AGENT,
-        content_id=agent_id,
-        embedding=mock_embedding,
-        searchable_text="AI writing assistant for blog posts",
-        metadata={"name": "Writing Assistant"},
-        user_id=None,
-    )
-
-    await embeddings.store_content_embedding(
-        content_type=ContentType.BLOCK,
-        content_id=block_id,
-        embedding=mock_embedding,
-        searchable_text="Text generation block for creative writing",
-        metadata={"name": "Text Generator"},
-        user_id=None,
-    )
-
-    await embeddings.store_content_embedding(
-        content_type=ContentType.DOCUMENTATION,
-        content_id=doc_id,
-        embedding=mock_embedding,
-        searchable_text="How to use writing blocks in AutoGPT",
-        metadata={"title": "Writing Guide"},
-        user_id=None,
-    )
-
-    # Search for "writing" - should find all three
-    results, total = await unified_hybrid_search(
-        query="writing",
-        page=1,
-        page_size=20,
-    )
-
-    # Should find at least our test content (may find others too)
-    content_ids = [r["content_id"] for r in results]
-    assert agent_id in content_ids or total >= 1  # Lexical search should find it
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_unified_hybrid_search_filter_by_content_type(
-    server,
-    mock_embedding: list[float],
-    cleanup_embeddings: list,
-):
-    """Test unified search can filter by content type."""
-    agent_id = f"test-agent-{uuid.uuid4()}"
-    block_id = f"test-block-{uuid.uuid4()}"
-
-    cleanup_embeddings.append((ContentType.STORE_AGENT, agent_id, None))
-    cleanup_embeddings.append((ContentType.BLOCK, block_id, None))
-
-    # Store both types with same searchable text
-    await embeddings.store_content_embedding(
-        content_type=ContentType.STORE_AGENT,
-        content_id=agent_id,
-        embedding=mock_embedding,
-        searchable_text="unique_search_term_xyz123",
-        metadata={},
-        user_id=None,
-    )
-
-    await embeddings.store_content_embedding(
-        content_type=ContentType.BLOCK,
-        content_id=block_id,
-        embedding=mock_embedding,
-        searchable_text="unique_search_term_xyz123",
-        metadata={},
-        user_id=None,
-    )
-
-    # Search only for BLOCK type
-    results, total = await unified_hybrid_search(
-        query="unique_search_term_xyz123",
-        content_types=[ContentType.BLOCK],
-        page=1,
-        page_size=20,
-    )
-
-    # All results should be BLOCK type
-    for r in results:
-        assert r["content_type"] == "BLOCK"
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_unified_hybrid_search_empty_query(server):
-    """Test unified search with empty query returns empty results."""
-    results, total = await unified_hybrid_search(
-        query="",
-        page=1,
-        page_size=20,
-    )
-
-    assert results == []
-    assert total == 0
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_unified_hybrid_search_pagination(
-    server,
-    mock_embedding: list[float],
-    cleanup_embeddings: list,
-):
-    """Test unified search pagination works correctly."""
-    # Create multiple items
-    content_ids = []
-    for i in range(5):
-        content_id = f"test-pagination-{uuid.uuid4()}"
-        content_ids.append(content_id)
-        cleanup_embeddings.append((ContentType.BLOCK, content_id, None))
-
-        await embeddings.store_content_embedding(
-            content_type=ContentType.BLOCK,
-            content_id=content_id,
-            embedding=mock_embedding,
-            searchable_text=f"pagination test item number {i}",
-            metadata={"index": i},
-            user_id=None,
-        )
-
-    # Get first page
-    page1_results, total1 = await unified_hybrid_search(
-        query="pagination test",
-        content_types=[ContentType.BLOCK],
-        page=1,
-        page_size=2,
-    )
-
-    # Get second page
-    page2_results, total2 = await unified_hybrid_search(
-        query="pagination test",
-        content_types=[ContentType.BLOCK],
-        page=2,
-        page_size=2,
-    )
-
-    # Total should be consistent
-    assert total1 == total2
-
-    # Pages should have different content (if we have enough results)
-    if len(page1_results) > 0 and len(page2_results) > 0:
-        page1_ids = {r["content_id"] for r in page1_results}
-        page2_ids = {r["content_id"] for r in page2_results}
-        # No overlap between pages
-        assert page1_ids.isdisjoint(page2_ids)
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_unified_hybrid_search_min_score_filtering(
-    server,
-    mock_embedding: list[float],
-    cleanup_embeddings: list,
-):
-    """Test unified search respects min_score threshold."""
-    content_id = f"test-minscore-{uuid.uuid4()}"
-    cleanup_embeddings.append((ContentType.BLOCK, content_id, None))
-
-    await embeddings.store_content_embedding(
-        content_type=ContentType.BLOCK,
-        content_id=content_id,
-        embedding=mock_embedding,
-        searchable_text="completely unrelated content about bananas",
-        metadata={},
-        user_id=None,
-    )
-
-    # Search with very high min_score - should filter out low relevance
-    results_high, _ = await unified_hybrid_search(
-        query="quantum computing algorithms",
-        content_types=[ContentType.BLOCK],
-        min_score=0.9,  # Very high threshold
-        page=1,
-        page_size=20,
-    )
-
-    # Search with low min_score
-    results_low, _ = await unified_hybrid_search(
-        query="quantum computing algorithms",
-        content_types=[ContentType.BLOCK],
-        min_score=0.01,  # Very low threshold
-        page=1,
-        page_size=20,
-    )
-
-    # High threshold should have fewer or equal results
-    assert len(results_high) <= len(results_low)
-
-
-# ============================================================================
-# hybrid_search (Store Agents) Tests
-# ============================================================================
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_hybrid_search_store_agents_sql_valid(server):
-    """Test that hybrid_search SQL executes without errors."""
-    # This test verifies the SQL is syntactically correct
-    # even if no results are found
-    results, total = await hybrid_search(
-        query="test agent",
-        page=1,
-        page_size=20,
-    )
-
-    # Should not raise - verifies SQL is valid
-    assert isinstance(results, list)
-    assert isinstance(total, int)
-    assert total >= 0
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_hybrid_search_with_filters(server):
-    """Test hybrid_search with various filter options."""
-    # Test with all filter types
-    results, total = await hybrid_search(
-        query="productivity",
-        featured=True,
-        creators=["test-creator"],
-        category="productivity",
-        page=1,
-        page_size=10,
-    )
-
-    # Should not raise - verifies filter SQL is valid
-    assert isinstance(results, list)
-    assert isinstance(total, int)
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_hybrid_search_pagination(server):
-    """Test hybrid_search pagination."""
-    # Page 1
-    results1, total1 = await hybrid_search(
-        query="agent",
-        page=1,
-        page_size=5,
-    )
-
-    # Page 2
-    results2, total2 = await hybrid_search(
-        query="agent",
-        page=2,
-        page_size=5,
-    )
-
-    # Verify SQL executes without error
-    assert isinstance(results1, list)
-    assert isinstance(results2, list)
-    assert isinstance(total1, int)
-    assert isinstance(total2, int)
-
-    # If page 1 has results, total should be > 0
-    # Note: total from page 2 may be 0 if no results on that page (COUNT(*) OVER limitation)
-    if results1:
-        assert total1 > 0
-
-
-# ============================================================================
-# SQL Validity Tests (verify queries don't break)
-# ============================================================================
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_all_content_types_searchable(server):
-    """Test that all content types can be searched without SQL errors."""
-    for content_type in [
-        ContentType.STORE_AGENT,
-        ContentType.BLOCK,
-        ContentType.DOCUMENTATION,
-    ]:
-        results, total = await unified_hybrid_search(
-            query="test",
-            content_types=[content_type],
-            page=1,
-            page_size=10,
-        )
-
-        # Should not raise
-        assert isinstance(results, list)
-        assert isinstance(total, int)
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_multiple_content_types_searchable(server):
-    """Test searching multiple content types at once."""
-    results, total = await unified_hybrid_search(
-        query="test",
-        content_types=[ContentType.BLOCK, ContentType.DOCUMENTATION],
-        page=1,
-        page_size=20,
-    )
-
-    # Should not raise
-    assert isinstance(results, list)
-    assert isinstance(total, int)
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_search_all_content_types_default(server):
-    """Test searching all content types (default behavior)."""
-    results, total = await unified_hybrid_search(
-        query="test",
-        content_types=None,  # Should search all
-        page=1,
-        page_size=20,
-    )
-
-    # Should not raise
-    assert isinstance(results, list)
-    assert isinstance(total, int)
-
-
-if __name__ == "__main__":
-    pytest.main([__file__, "-v", "-s"])

autogpt_platform/backend/backend/api/features/store/embeddings_schema_test.py

@@ -1,315 +0,0 @@
-"""
-Integration tests for embeddings with schema handling.
-
-These tests verify that embeddings operations work correctly across different database schemas.
-"""
-
-from unittest.mock import AsyncMock, MagicMock, patch
-
-import pytest
-from prisma.enums import ContentType
-
-from backend.api.features.store import embeddings
-from backend.api.features.store.embeddings import EMBEDDING_DIM
-
-# Schema prefix tests removed - functionality moved to db.raw_with_schema() helper
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_store_content_embedding_with_schema():
-    """Test storing embeddings with proper schema handling."""
-    with patch("backend.data.db.get_database_schema") as mock_schema:
-        mock_schema.return_value = "platform"
-
-        with patch("prisma.get_client") as mock_get_client:
-            mock_client = AsyncMock()
-            mock_get_client.return_value = mock_client
-
-            result = await embeddings.store_content_embedding(
-                content_type=ContentType.STORE_AGENT,
-                content_id="test-id",
-                embedding=[0.1] * EMBEDDING_DIM,
-                searchable_text="test text",
-                metadata={"test": "data"},
-                user_id=None,
-            )
-
-            # Verify the query was called
-            assert mock_client.execute_raw.called
-
-            # Get the SQL query that was executed
-            call_args = mock_client.execute_raw.call_args
-            sql_query = call_args[0][0]
-
-            # Verify schema prefix is in the query
-            assert '"platform"."UnifiedContentEmbedding"' in sql_query
-
-            # Verify result
-            assert result is True
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_get_content_embedding_with_schema():
-    """Test retrieving embeddings with proper schema handling."""
-    with patch("backend.data.db.get_database_schema") as mock_schema:
-        mock_schema.return_value = "platform"
-
-        with patch("prisma.get_client") as mock_get_client:
-            mock_client = AsyncMock()
-            mock_client.query_raw.return_value = [
-                {
-                    "contentType": "STORE_AGENT",
-                    "contentId": "test-id",
-                    "userId": None,
-                    "embedding": "[0.1, 0.2]",
-                    "searchableText": "test",
-                    "metadata": {},
-                    "createdAt": "2024-01-01",
-                    "updatedAt": "2024-01-01",
-                }
-            ]
-            mock_get_client.return_value = mock_client
-
-            result = await embeddings.get_content_embedding(
-                ContentType.STORE_AGENT,
-                "test-id",
-                user_id=None,
-            )
-
-            # Verify the query was called
-            assert mock_client.query_raw.called
-
-            # Get the SQL query that was executed
-            call_args = mock_client.query_raw.call_args
-            sql_query = call_args[0][0]
-
-            # Verify schema prefix is in the query
-            assert '"platform"."UnifiedContentEmbedding"' in sql_query
-
-            # Verify result
-            assert result is not None
-            assert result["contentId"] == "test-id"
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_delete_content_embedding_with_schema():
-    """Test deleting embeddings with proper schema handling."""
-    with patch("backend.data.db.get_database_schema") as mock_schema:
-        mock_schema.return_value = "platform"
-
-        with patch("prisma.get_client") as mock_get_client:
-            mock_client = AsyncMock()
-            mock_get_client.return_value = mock_client
-
-            result = await embeddings.delete_content_embedding(
-                ContentType.STORE_AGENT,
-                "test-id",
-            )
-
-            # Verify the query was called
-            assert mock_client.execute_raw.called
-
-            # Get the SQL query that was executed
-            call_args = mock_client.execute_raw.call_args
-            sql_query = call_args[0][0]
-
-            # Verify schema prefix is in the query
-            assert '"platform"."UnifiedContentEmbedding"' in sql_query
-
-            # Verify result
-            assert result is True
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_get_embedding_stats_with_schema():
-    """Test embedding statistics with proper schema handling via content handlers."""
-    # Mock handler to return stats
-    mock_handler = MagicMock()
-    mock_handler.get_stats = AsyncMock(
-        return_value={
-            "total": 100,
-            "with_embeddings": 80,
-            "without_embeddings": 20,
-        }
-    )
-
-    with patch(
-        "backend.api.features.store.embeddings.CONTENT_HANDLERS",
-        {ContentType.STORE_AGENT: mock_handler},
-    ):
-        result = await embeddings.get_embedding_stats()
-
-        # Verify handler was called
-        mock_handler.get_stats.assert_called_once()
-
-        # Verify new result structure
-        assert "by_type" in result
-        assert "totals" in result
-        assert result["totals"]["total"] == 100
-        assert result["totals"]["with_embeddings"] == 80
-        assert result["totals"]["without_embeddings"] == 20
-        assert result["totals"]["coverage_percent"] == 80.0
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_backfill_missing_embeddings_with_schema():
-    """Test backfilling embeddings via content handlers."""
-    from backend.api.features.store.content_handlers import ContentItem
-
-    # Create mock content item
-    mock_item = ContentItem(
-        content_id="version-1",
-        content_type=ContentType.STORE_AGENT,
-        searchable_text="Test Agent Test description",
-        metadata={"name": "Test Agent"},
-    )
-
-    # Mock handler
-    mock_handler = MagicMock()
-    mock_handler.get_missing_items = AsyncMock(return_value=[mock_item])
-
-    with patch(
-        "backend.api.features.store.embeddings.CONTENT_HANDLERS",
-        {ContentType.STORE_AGENT: mock_handler},
-    ):
-        with patch(
-            "backend.api.features.store.embeddings.generate_embedding",
-            return_value=[0.1] * EMBEDDING_DIM,
-        ):
-            with patch(
-                "backend.api.features.store.embeddings.store_content_embedding",
-                return_value=True,
-            ):
-                result = await embeddings.backfill_missing_embeddings(batch_size=10)
-
-                # Verify handler was called
-                mock_handler.get_missing_items.assert_called_once_with(10)
-
-                # Verify results
-                assert result["processed"] == 1
-                assert result["success"] == 1
-                assert result["failed"] == 0
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_ensure_content_embedding_with_schema():
-    """Test ensuring embeddings exist with proper schema handling."""
-    with patch("backend.data.db.get_database_schema") as mock_schema:
-        mock_schema.return_value = "platform"
-
-        with patch(
-            "backend.api.features.store.embeddings.get_content_embedding"
-        ) as mock_get:
-            # Simulate no existing embedding
-            mock_get.return_value = None
-
-            with patch(
-                "backend.api.features.store.embeddings.generate_embedding"
-            ) as mock_generate:
-                mock_generate.return_value = [0.1] * EMBEDDING_DIM
-
-                with patch(
-                    "backend.api.features.store.embeddings.store_content_embedding"
-                ) as mock_store:
-                    mock_store.return_value = True
-
-                    result = await embeddings.ensure_content_embedding(
-                        content_type=ContentType.STORE_AGENT,
-                        content_id="test-id",
-                        searchable_text="test text",
-                        metadata={"test": "data"},
-                        user_id=None,
-                        force=False,
-                    )
-
-                    # Verify the flow
-                    assert mock_get.called
-                    assert mock_generate.called
-                    assert mock_store.called
-                    assert result is True
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_backward_compatibility_store_embedding():
-    """Test backward compatibility wrapper for store_embedding."""
-    with patch(
-        "backend.api.features.store.embeddings.store_content_embedding"
-    ) as mock_store:
-        mock_store.return_value = True
-
-        result = await embeddings.store_embedding(
-            version_id="test-version-id",
-            embedding=[0.1] * EMBEDDING_DIM,
-            tx=None,
-        )
-
-        # Verify it calls the new function with correct parameters
-        assert mock_store.called
-        call_args = mock_store.call_args
-
-        assert call_args[1]["content_type"] == ContentType.STORE_AGENT
-        assert call_args[1]["content_id"] == "test-version-id"
-        assert call_args[1]["user_id"] is None
-        assert result is True
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_backward_compatibility_get_embedding():
-    """Test backward compatibility wrapper for get_embedding."""
-    with patch(
-        "backend.api.features.store.embeddings.get_content_embedding"
-    ) as mock_get:
-        mock_get.return_value = {
-            "contentType": "STORE_AGENT",
-            "contentId": "test-version-id",
-            "embedding": "[0.1, 0.2]",
-            "createdAt": "2024-01-01",
-            "updatedAt": "2024-01-01",
-        }
-
-        result = await embeddings.get_embedding("test-version-id")
-
-        # Verify it calls the new function
-        assert mock_get.called
-
-        # Verify it transforms to old format
-        assert result is not None
-        assert result["storeListingVersionId"] == "test-version-id"
-        assert "embedding" in result
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_schema_handling_error_cases():
-    """Test error handling in schema-aware operations."""
-    with patch("backend.data.db.get_database_schema") as mock_schema:
-        mock_schema.return_value = "platform"
-
-        with patch("prisma.get_client") as mock_get_client:
-            mock_client = AsyncMock()
-            mock_client.execute_raw.side_effect = Exception("Database error")
-            mock_get_client.return_value = mock_client
-
-            result = await embeddings.store_content_embedding(
-                content_type=ContentType.STORE_AGENT,
-                content_id="test-id",
-                embedding=[0.1] * EMBEDDING_DIM,
-                searchable_text="test",
-                metadata=None,
-                user_id=None,
-            )
-
-            # Should return False on error, not raise
-            assert result is False
-
-
-if __name__ == "__main__":
-    pytest.main([__file__, "-v", "-s"])

autogpt_platform/backend/backend/api/features/store/embeddings_test.py

@@ -1,407 +0,0 @@
-from unittest.mock import AsyncMock, MagicMock, patch
-
-import prisma
-import pytest
-from prisma import Prisma
-from prisma.enums import ContentType
-
-from backend.api.features.store import embeddings
-
-
-@pytest.fixture(autouse=True)
-async def setup_prisma():
-    """Setup Prisma client for tests."""
-    try:
-        Prisma()
-    except prisma.errors.ClientAlreadyRegisteredError:
-        pass
-    yield
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_build_searchable_text():
-    """Test searchable text building from listing fields."""
-    result = embeddings.build_searchable_text(
-        name="AI Assistant",
-        description="A helpful AI assistant for productivity",
-        sub_heading="Boost your productivity",
-        categories=["AI", "Productivity"],
-    )
-
-    expected = "AI Assistant Boost your productivity A helpful AI assistant for productivity AI Productivity"
-    assert result == expected
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_build_searchable_text_empty_fields():
-    """Test searchable text building with empty fields."""
-    result = embeddings.build_searchable_text(
-        name="", description="Test description", sub_heading="", categories=[]
-    )
-
-    assert result == "Test description"
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_generate_embedding_success():
-    """Test successful embedding generation."""
-    # Mock OpenAI response
-    mock_client = MagicMock()
-    mock_response = MagicMock()
-    mock_response.data = [MagicMock()]
-    mock_response.data[0].embedding = [0.1, 0.2, 0.3] * 512  # 1536 dimensions
-
-    # Use AsyncMock for async embeddings.create method
-    mock_client.embeddings.create = AsyncMock(return_value=mock_response)
-
-    # Patch at the point of use in embeddings.py
-    with patch(
-        "backend.api.features.store.embeddings.get_openai_client"
-    ) as mock_get_client:
-        mock_get_client.return_value = mock_client
-
-        result = await embeddings.generate_embedding("test text")
-
-        assert result is not None
-        assert len(result) == embeddings.EMBEDDING_DIM
-        assert result[0] == 0.1
-
-        mock_client.embeddings.create.assert_called_once_with(
-            model="text-embedding-3-small", input="test text"
-        )
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_generate_embedding_no_api_key():
-    """Test embedding generation without API key."""
-    # Patch at the point of use in embeddings.py
-    with patch(
-        "backend.api.features.store.embeddings.get_openai_client"
-    ) as mock_get_client:
-        mock_get_client.return_value = None
-
-        result = await embeddings.generate_embedding("test text")
-
-        assert result is None
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_generate_embedding_api_error():
-    """Test embedding generation with API error."""
-    mock_client = MagicMock()
-    mock_client.embeddings.create = AsyncMock(side_effect=Exception("API Error"))
-
-    # Patch at the point of use in embeddings.py
-    with patch(
-        "backend.api.features.store.embeddings.get_openai_client"
-    ) as mock_get_client:
-        mock_get_client.return_value = mock_client
-
-        result = await embeddings.generate_embedding("test text")
-
-        assert result is None
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_generate_embedding_text_truncation():
-    """Test that long text is properly truncated using tiktoken."""
-    from tiktoken import encoding_for_model
-
-    mock_client = MagicMock()
-    mock_response = MagicMock()
-    mock_response.data = [MagicMock()]
-    mock_response.data[0].embedding = [0.1] * embeddings.EMBEDDING_DIM
-
-    # Use AsyncMock for async embeddings.create method
-    mock_client.embeddings.create = AsyncMock(return_value=mock_response)
-
-    # Patch at the point of use in embeddings.py
-    with patch(
-        "backend.api.features.store.embeddings.get_openai_client"
-    ) as mock_get_client:
-        mock_get_client.return_value = mock_client
-
-        # Create text that will exceed 8191 tokens
-        # Use varied characters to ensure token-heavy text: each word is ~1 token
-        words = [f"word{i}" for i in range(10000)]
-        long_text = " ".join(words)  # ~10000 tokens
-
-        await embeddings.generate_embedding(long_text)
-
-        # Verify text was truncated to 8191 tokens
-        call_args = mock_client.embeddings.create.call_args
-        truncated_text = call_args.kwargs["input"]
-
-        # Count actual tokens in truncated text
-        enc = encoding_for_model("text-embedding-3-small")
-        actual_tokens = len(enc.encode(truncated_text))
-
-        # Should be at or just under 8191 tokens
-        assert actual_tokens <= 8191
-        # Should be close to the limit (not over-truncated)
-        assert actual_tokens >= 8100
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_store_embedding_success(mocker):
-    """Test successful embedding storage."""
-    mock_client = mocker.AsyncMock()
-    mock_client.execute_raw = mocker.AsyncMock()
-
-    embedding = [0.1, 0.2, 0.3]
-
-    result = await embeddings.store_embedding(
-        version_id="test-version-id", embedding=embedding, tx=mock_client
-    )
-
-    assert result is True
-    # execute_raw is called twice: once for SET search_path, once for INSERT
-    assert mock_client.execute_raw.call_count == 2
-
-    # First call: SET search_path
-    first_call_args = mock_client.execute_raw.call_args_list[0][0]
-    assert "SET search_path" in first_call_args[0]
-
-    # Second call: INSERT query with the actual data
-    second_call_args = mock_client.execute_raw.call_args_list[1][0]
-    assert "test-version-id" in second_call_args
-    assert "[0.1,0.2,0.3]" in second_call_args
-    assert None in second_call_args  # userId should be None for store agents
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_store_embedding_database_error(mocker):
-    """Test embedding storage with database error."""
-    mock_client = mocker.AsyncMock()
-    mock_client.execute_raw.side_effect = Exception("Database error")
-
-    embedding = [0.1, 0.2, 0.3]
-
-    result = await embeddings.store_embedding(
-        version_id="test-version-id", embedding=embedding, tx=mock_client
-    )
-
-    assert result is False
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_get_embedding_success():
-    """Test successful embedding retrieval."""
-    mock_result = [
-        {
-            "contentType": "STORE_AGENT",
-            "contentId": "test-version-id",
-            "userId": None,
-            "embedding": "[0.1,0.2,0.3]",
-            "searchableText": "Test text",
-            "metadata": {},
-            "createdAt": "2024-01-01T00:00:00Z",
-            "updatedAt": "2024-01-01T00:00:00Z",
-        }
-    ]
-
-    with patch(
-        "backend.api.features.store.embeddings.query_raw_with_schema",
-        return_value=mock_result,
-    ):
-        result = await embeddings.get_embedding("test-version-id")
-
-        assert result is not None
-        assert result["storeListingVersionId"] == "test-version-id"
-        assert result["embedding"] == "[0.1,0.2,0.3]"
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_get_embedding_not_found():
-    """Test embedding retrieval when not found."""
-    with patch(
-        "backend.api.features.store.embeddings.query_raw_with_schema",
-        return_value=[],
-    ):
-        result = await embeddings.get_embedding("test-version-id")
-
-        assert result is None
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@patch("backend.api.features.store.embeddings.generate_embedding")
-@patch("backend.api.features.store.embeddings.store_embedding")
-@patch("backend.api.features.store.embeddings.get_embedding")
-async def test_ensure_embedding_already_exists(mock_get, mock_store, mock_generate):
-    """Test ensure_embedding when embedding already exists."""
-    mock_get.return_value = {"embedding": "[0.1,0.2,0.3]"}
-
-    result = await embeddings.ensure_embedding(
-        version_id="test-id",
-        name="Test",
-        description="Test description",
-        sub_heading="Test heading",
-        categories=["test"],
-    )
-
-    assert result is True
-    mock_generate.assert_not_called()
-    mock_store.assert_not_called()
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@patch("backend.api.features.store.embeddings.generate_embedding")
-@patch("backend.api.features.store.embeddings.store_content_embedding")
-@patch("backend.api.features.store.embeddings.get_embedding")
-async def test_ensure_embedding_create_new(mock_get, mock_store, mock_generate):
-    """Test ensure_embedding creating new embedding."""
-    mock_get.return_value = None
-    mock_generate.return_value = [0.1, 0.2, 0.3]
-    mock_store.return_value = True
-
-    result = await embeddings.ensure_embedding(
-        version_id="test-id",
-        name="Test",
-        description="Test description",
-        sub_heading="Test heading",
-        categories=["test"],
-    )
-
-    assert result is True
-    mock_generate.assert_called_once_with("Test Test heading Test description test")
-    mock_store.assert_called_once_with(
-        content_type=ContentType.STORE_AGENT,
-        content_id="test-id",
-        embedding=[0.1, 0.2, 0.3],
-        searchable_text="Test Test heading Test description test",
-        metadata={"name": "Test", "subHeading": "Test heading", "categories": ["test"]},
-        user_id=None,
-        tx=None,
-    )
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@patch("backend.api.features.store.embeddings.generate_embedding")
-@patch("backend.api.features.store.embeddings.get_embedding")
-async def test_ensure_embedding_generation_fails(mock_get, mock_generate):
-    """Test ensure_embedding when generation fails."""
-    mock_get.return_value = None
-    mock_generate.return_value = None
-
-    result = await embeddings.ensure_embedding(
-        version_id="test-id",
-        name="Test",
-        description="Test description",
-        sub_heading="Test heading",
-        categories=["test"],
-    )
-
-    assert result is False
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_get_embedding_stats():
-    """Test embedding statistics retrieval."""
-    # Mock handler stats for each content type
-    mock_handler = MagicMock()
-    mock_handler.get_stats = AsyncMock(
-        return_value={
-            "total": 100,
-            "with_embeddings": 75,
-            "without_embeddings": 25,
-        }
-    )
-
-    # Patch the CONTENT_HANDLERS where it's used (in embeddings module)
-    with patch(
-        "backend.api.features.store.embeddings.CONTENT_HANDLERS",
-        {ContentType.STORE_AGENT: mock_handler},
-    ):
-        result = await embeddings.get_embedding_stats()
-
-        assert "by_type" in result
-        assert "totals" in result
-        assert result["totals"]["total"] == 100
-        assert result["totals"]["with_embeddings"] == 75
-        assert result["totals"]["without_embeddings"] == 25
-        assert result["totals"]["coverage_percent"] == 75.0
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@patch("backend.api.features.store.embeddings.store_content_embedding")
-async def test_backfill_missing_embeddings_success(mock_store):
-    """Test backfill with successful embedding generation."""
-    # Mock ContentItem from handlers
-    from backend.api.features.store.content_handlers import ContentItem
-
-    mock_items = [
-        ContentItem(
-            content_id="version-1",
-            content_type=ContentType.STORE_AGENT,
-            searchable_text="Agent 1 Description 1",
-            metadata={"name": "Agent 1"},
-        ),
-        ContentItem(
-            content_id="version-2",
-            content_type=ContentType.STORE_AGENT,
-            searchable_text="Agent 2 Description 2",
-            metadata={"name": "Agent 2"},
-        ),
-    ]
-
-    # Mock handler to return missing items
-    mock_handler = MagicMock()
-    mock_handler.get_missing_items = AsyncMock(return_value=mock_items)
-
-    # Mock store_content_embedding to succeed for first, fail for second
-    mock_store.side_effect = [True, False]
-
-    with patch(
-        "backend.api.features.store.embeddings.CONTENT_HANDLERS",
-        {ContentType.STORE_AGENT: mock_handler},
-    ):
-        with patch(
-            "backend.api.features.store.embeddings.generate_embedding",
-            return_value=[0.1] * embeddings.EMBEDDING_DIM,
-        ):
-            result = await embeddings.backfill_missing_embeddings(batch_size=5)
-
-            assert result["processed"] == 2
-            assert result["success"] == 1
-            assert result["failed"] == 1
-            assert mock_store.call_count == 2
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_backfill_missing_embeddings_no_missing():
-    """Test backfill when no embeddings are missing."""
-    # Mock handler to return no missing items
-    mock_handler = MagicMock()
-    mock_handler.get_missing_items = AsyncMock(return_value=[])
-
-    with patch(
-        "backend.api.features.store.embeddings.CONTENT_HANDLERS",
-        {ContentType.STORE_AGENT: mock_handler},
-    ):
-        result = await embeddings.backfill_missing_embeddings(batch_size=5)
-
-        assert result["processed"] == 0
-        assert result["success"] == 0
-        assert result["failed"] == 0
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_embedding_to_vector_string():
-    """Test embedding to PostgreSQL vector string conversion."""
-    embedding = [0.1, 0.2, 0.3, -0.4]
-    result = embeddings.embedding_to_vector_string(embedding)
-    assert result == "[0.1,0.2,0.3,-0.4]"
-
-
-@pytest.mark.asyncio(loop_scope="session")
-async def test_embed_query():
-    """Test embed_query function (alias for generate_embedding)."""
-    with patch(
-        "backend.api.features.store.embeddings.generate_embedding"
-    ) as mock_generate:
-        mock_generate.return_value = [0.1, 0.2, 0.3]
-
-        result = await embeddings.embed_query("test query")
-
-        assert result == [0.1, 0.2, 0.3]
-        mock_generate.assert_called_once_with("test query")

autogpt_platform/backend/backend/api/features/store/hybrid_search.py

@@ -1,343 +1,69 @@
 """
-Unified Hybrid Search
+Hybrid Search for Store Agents
 
 Combines semantic (embedding) search with lexical (tsvector) search
-for improved relevance across all content types (agents, blocks, docs).
+for improved relevance in marketplace agent discovery.
 """
 
 import logging
 from dataclasses import dataclass
+from datetime import datetime
 from typing import Any, Literal
 
-from prisma.enums import ContentType
+import prisma
 
 from backend.api.features.store.embeddings import (
-    EMBEDDING_DIM,
     embed_query,
     embedding_to_vector_string,
 )
-from backend.data.db import query_raw_with_schema
 
 logger = logging.getLogger(__name__)
 
 
 @dataclass
-class UnifiedSearchWeights:
-    """Weights for unified search (no popularity signal)."""
+class HybridSearchWeights:
+    """Weights for combining search signals."""
 
-    semantic: float = 0.40  # Embedding cosine similarity
-    lexical: float = 0.40  # tsvector ts_rank_cd score
-    category: float = 0.10  # Category match boost (for types that have categories)
-    recency: float = 0.10  # Newer content ranked higher
-
-    def __post_init__(self):
-        """Validate weights are non-negative and sum to approximately 1.0."""
-        total = self.semantic + self.lexical + self.category + self.recency
-
-        if any(
-            w < 0 for w in [self.semantic, self.lexical, self.category, self.recency]
-        ):
-            raise ValueError("All weights must be non-negative")
-
-        if not (0.99 <= total <= 1.01):
-            raise ValueError(f"Weights must sum to ~1.0, got {total:.3f}")
+    semantic: float = 0.35  # Embedding cosine similarity
+    lexical: float = 0.35  # tsvector ts_rank_cd score
+    category: float = 0.20  # Category match boost
+    recency: float = 0.10  # Newer agents ranked higher
 
 
-# Default weights for unified search
-DEFAULT_UNIFIED_WEIGHTS = UnifiedSearchWeights()
+DEFAULT_WEIGHTS = HybridSearchWeights()
 
-# Minimum relevance score thresholds
-DEFAULT_MIN_SCORE = 0.15  # For unified search (more permissive)
-DEFAULT_STORE_AGENT_MIN_SCORE = 0.20  # For store agent search (original threshold)
-
-
-async def unified_hybrid_search(
-    query: str,
-    content_types: list[ContentType] | None = None,
-    category: str | None = None,
-    page: int = 1,
-    page_size: int = 20,
-    weights: UnifiedSearchWeights | None = None,
-    min_score: float | None = None,
-    user_id: str | None = None,
-) -> tuple[list[dict[str, Any]], int]:
-    """
-    Unified hybrid search across all content types.
-
-    Searches UnifiedContentEmbedding using both semantic (vector) and lexical (tsvector) signals.
-
-    Args:
-        query: Search query string
-        content_types: List of content types to search. Defaults to all public types.
-        category: Filter by category (for content types that support it)
-        page: Page number (1-indexed)
-        page_size: Results per page
-        weights: Custom weights for search signals
-        min_score: Minimum relevance score threshold (0-1)
-        user_id: User ID for searching private content (library agents)
-
-    Returns:
-        Tuple of (results list, total count)
-    """
-    # Validate inputs
-    query = query.strip()
-    if not query:
-        return [], 0
-
-    if page < 1:
-        page = 1
-    if page_size < 1:
-        page_size = 1
-    if page_size > 100:
-        page_size = 100
-
-    if content_types is None:
-        content_types = [
-            ContentType.STORE_AGENT,
-            ContentType.BLOCK,
-            ContentType.DOCUMENTATION,
-        ]
-
-    if weights is None:
-        weights = DEFAULT_UNIFIED_WEIGHTS
-    if min_score is None:
-        min_score = DEFAULT_MIN_SCORE
-
-    offset = (page - 1) * page_size
-
-    # Generate query embedding
-    query_embedding = await embed_query(query)
-
-    # Graceful degradation if embedding unavailable
-    if query_embedding is None or not query_embedding:
-        logger.warning(
-            "Failed to generate query embedding - falling back to lexical-only search. "
-            "Check that openai_internal_api_key is configured and OpenAI API is accessible."
-        )
-        query_embedding = [0.0] * EMBEDDING_DIM
-        # Redistribute semantic weight to lexical
-        total_non_semantic = weights.lexical + weights.category + weights.recency
-        if total_non_semantic > 0:
-            factor = 1.0 / total_non_semantic
-            weights = UnifiedSearchWeights(
-                semantic=0.0,
-                lexical=weights.lexical * factor,
-                category=weights.category * factor,
-                recency=weights.recency * factor,
-            )
-        else:
-            weights = UnifiedSearchWeights(
-                semantic=0.0, lexical=1.0, category=0.0, recency=0.0
-            )
-
-    # Build parameters
-    params: list[Any] = []
-    param_idx = 1
-
-    # Query for lexical search
-    params.append(query)
-    query_param = f"${param_idx}"
-    param_idx += 1
-
-    # Query lowercase for category matching
-    params.append(query.lower())
-    query_lower_param = f"${param_idx}"
-    param_idx += 1
-
-    # Embedding
-    embedding_str = embedding_to_vector_string(query_embedding)
-    params.append(embedding_str)
-    embedding_param = f"${param_idx}"
-    param_idx += 1
-
-    # Content types
-    content_type_values = [ct.value for ct in content_types]
-    params.append(content_type_values)
-    content_types_param = f"${param_idx}"
-    param_idx += 1
-
-    # User ID filter (for private content)
-    user_filter = ""
-    if user_id is not None:
-        params.append(user_id)
-        user_filter = f'AND (uce."userId" = ${param_idx} OR uce."userId" IS NULL)'
-        param_idx += 1
-    else:
-        user_filter = 'AND uce."userId" IS NULL'
-
-    # Weights
-    params.append(weights.semantic)
-    w_semantic = f"${param_idx}"
-    param_idx += 1
-
-    params.append(weights.lexical)
-    w_lexical = f"${param_idx}"
-    param_idx += 1
-
-    params.append(weights.category)
-    w_category = f"${param_idx}"
-    param_idx += 1
-
-    params.append(weights.recency)
-    w_recency = f"${param_idx}"
-    param_idx += 1
-
-    # Min score
-    params.append(min_score)
-    min_score_param = f"${param_idx}"
-    param_idx += 1
-
-    # Pagination
-    params.append(page_size)
-    limit_param = f"${param_idx}"
-    param_idx += 1
-
-    params.append(offset)
-    offset_param = f"${param_idx}"
-    param_idx += 1
-
-    # Unified search query on UnifiedContentEmbedding
-    sql_query = f"""
-        WITH candidates AS (
-            -- Lexical matches (uses GIN index on search column)
-            SELECT uce.id, uce."contentType", uce."contentId"
-            FROM {{schema_prefix}}"UnifiedContentEmbedding" uce
-            WHERE uce."contentType" = ANY({content_types_param}::{{schema_prefix}}"ContentType"[])
-            {user_filter}
-            AND uce.search @@ plainto_tsquery('english', {query_param})
-
-            UNION
-
-            -- Semantic matches (uses HNSW index on embedding)
-            (
-                SELECT uce.id, uce."contentType", uce."contentId"
-                FROM {{schema_prefix}}"UnifiedContentEmbedding" uce
-                WHERE uce."contentType" = ANY({content_types_param}::{{schema_prefix}}"ContentType"[])
-                {user_filter}
-                ORDER BY uce.embedding <=> {embedding_param}::vector
-                LIMIT 200
-            )
-        ),
-        search_scores AS (
-            SELECT
-                uce."contentType" as content_type,
-                uce."contentId" as content_id,
-                uce."searchableText" as searchable_text,
-                uce.metadata,
-                uce."updatedAt" as updated_at,
-                -- Semantic score: cosine similarity (1 - distance)
-                COALESCE(1 - (uce.embedding <=> {embedding_param}::vector), 0) as semantic_score,
-                -- Lexical score: ts_rank_cd
-                COALESCE(ts_rank_cd(uce.search, plainto_tsquery('english', {query_param})), 0) as lexical_raw,
-                -- Category match from metadata
-                CASE
-                    WHEN uce.metadata ? 'categories' AND EXISTS (
-                        SELECT 1 FROM jsonb_array_elements_text(uce.metadata->'categories') cat
-                        WHERE LOWER(cat) LIKE '%' || {query_lower_param} || '%'
-                    )
-                    THEN 1.0
-                    ELSE 0.0
-                END as category_score,
-                -- Recency score: linear decay over 90 days
-                GREATEST(0, 1 - EXTRACT(EPOCH FROM (NOW() - uce."updatedAt")) / (90 * 24 * 3600)) as recency_score
-            FROM candidates c
-            INNER JOIN {{schema_prefix}}"UnifiedContentEmbedding" uce ON c.id = uce.id
-        ),
-        max_lexical AS (
-            SELECT GREATEST(MAX(lexical_raw), 0.001) as max_val FROM search_scores
-        ),
-        normalized AS (
-            SELECT
-                ss.*,
-                ss.lexical_raw / ml.max_val as lexical_score
-            FROM search_scores ss
-            CROSS JOIN max_lexical ml
-        ),
-        scored AS (
-            SELECT
-                content_type,
-                content_id,
-                searchable_text,
-                metadata,
-                updated_at,
-                semantic_score,
-                lexical_score,
-                category_score,
-                recency_score,
-                (
-                    {w_semantic} * semantic_score +
-                    {w_lexical} * lexical_score +
-                    {w_category} * category_score +
-                    {w_recency} * recency_score
-                ) as combined_score
-            FROM normalized
-        ),
-        filtered AS (
-            SELECT
-                *,
-                COUNT(*) OVER () as total_count
-            FROM scored
-            WHERE combined_score >= {min_score_param}
-        )
-        SELECT * FROM filtered
-        ORDER BY combined_score DESC
-        LIMIT {limit_param} OFFSET {offset_param}
-    """
-
-    results = await query_raw_with_schema(
-        sql_query, *params, set_public_search_path=True
-    )
-
-    total = results[0]["total_count"] if results else 0
-
-    # Clean up results
-    for result in results:
-        result.pop("total_count", None)
-
-    logger.info(f"Unified hybrid search: {len(results)} results, {total} total")
-
-    return results, total
-
-
-# ============================================================================
-# Store Agent specific search (with full metadata)
-# ============================================================================
+# Minimum relevance score threshold - agents below this are filtered out
+# With weights (0.35 semantic + 0.35 lexical + 0.20 category + 0.10 recency):
+# - 0.20 means at least ~50% semantic match OR strong lexical match required
+# - Ensures only genuinely relevant results are returned
+# - Recency alone (0.10 max) won't pass the threshold
+DEFAULT_MIN_SCORE = 0.20
 
 
 @dataclass
-class StoreAgentSearchWeights:
-    """Weights for store agent search including popularity."""
+class HybridSearchResult:
+    """A single search result with score breakdown."""
 
-    semantic: float = 0.30
-    lexical: float = 0.30
-    category: float = 0.20
-    recency: float = 0.10
-    popularity: float = 0.10
+    slug: str
+    agent_name: str
+    agent_image: str
+    creator_username: str
+    creator_avatar: str
+    sub_heading: str
+    description: str
+    runs: int
+    rating: float
+    categories: list[str]
+    featured: bool
+    is_available: bool
+    updated_at: datetime
 
-    def __post_init__(self):
-        total = (
-            self.semantic
-            + self.lexical
-            + self.category
-            + self.recency
-            + self.popularity
-        )
-        if any(
-            w < 0
-            for w in [
-                self.semantic,
-                self.lexical,
-                self.category,
-                self.recency,
-                self.popularity,
-            ]
-        ):
-            raise ValueError("All weights must be non-negative")
-        if not (0.99 <= total <= 1.01):
-            raise ValueError(f"Weights must sum to ~1.0, got {total:.3f}")
-
-
-DEFAULT_STORE_AGENT_WEIGHTS = StoreAgentSearchWeights()
+    # Score breakdown (for debugging/tuning)
+    combined_score: float
+    semantic_score: float = 0.0
+    lexical_score: float = 0.0
+    category_score: float = 0.0
+    recency_score: float = 0.0
 
 
 async def hybrid_search(
@@ -350,265 +76,351 @@ async def hybrid_search(
     ) = None,
     page: int = 1,
     page_size: int = 20,
-    weights: StoreAgentSearchWeights | None = None,
+    weights: HybridSearchWeights | None = None,
     min_score: float | None = None,
 ) -> tuple[list[dict[str, Any]], int]:
     """
-    Hybrid search for store agents with full metadata.
+    Perform hybrid search combining semantic and lexical signals.
 
-    Uses UnifiedContentEmbedding for search, joins to StoreAgent for metadata.
+    Args:
+        query: Search query string
+        featured: Filter for featured agents only
+        creators: Filter by creator usernames
+        category: Filter by category
+        sorted_by: Sort order (relevance uses hybrid scoring)
+        page: Page number (1-indexed)
+        page_size: Results per page
+        weights: Custom weights for search signals
+        min_score: Minimum relevance score threshold (0-1). Results below
+                   this score are filtered out. Defaults to DEFAULT_MIN_SCORE.
+
+    Returns:
+        Tuple of (results list, total count). Returns empty list if no
+        results meet the minimum relevance threshold.
     """
-    query = query.strip()
-    if not query:
-        return [], 0
-
-    if page < 1:
-        page = 1
-    if page_size < 1:
-        page_size = 1
-    if page_size > 100:
-        page_size = 100
-
     if weights is None:
-        weights = DEFAULT_STORE_AGENT_WEIGHTS
+        weights = DEFAULT_WEIGHTS
     if min_score is None:
-        min_score = (
-            DEFAULT_STORE_AGENT_MIN_SCORE  # Use original threshold for store agents
-        )
+        min_score = DEFAULT_MIN_SCORE
 
     offset = (page - 1) * page_size
+    client = prisma.get_client()
 
     # Generate query embedding
     query_embedding = await embed_query(query)
 
-    # Graceful degradation
-    if query_embedding is None or not query_embedding:
-        logger.warning(
-            "Failed to generate query embedding - falling back to lexical-only search."
-        )
-        query_embedding = [0.0] * EMBEDDING_DIM
-        total_non_semantic = (
-            weights.lexical + weights.category + weights.recency + weights.popularity
-        )
-        if total_non_semantic > 0:
-            factor = 1.0 / total_non_semantic
-            weights = StoreAgentSearchWeights(
-                semantic=0.0,
-                lexical=weights.lexical * factor,
-                category=weights.category * factor,
-                recency=weights.recency * factor,
-                popularity=weights.popularity * factor,
-            )
-        else:
-            weights = StoreAgentSearchWeights(
-                semantic=0.0, lexical=1.0, category=0.0, recency=0.0, popularity=0.0
-            )
-
-    # Build parameters
+    # Build WHERE clause conditions
+    where_parts: list[str] = ["sa.is_available = true"]
     params: list[Any] = []
-    param_idx = 1
+    param_index = 1
 
+    # Add search query for lexical matching
     params.append(query)
-    query_param = f"${param_idx}"
-    param_idx += 1
-
-    params.append(query.lower())
-    query_lower_param = f"${param_idx}"
-    param_idx += 1
-
-    embedding_str = embedding_to_vector_string(query_embedding)
-    params.append(embedding_str)
-    embedding_param = f"${param_idx}"
-    param_idx += 1
-
-    # Build WHERE clause for StoreAgent filters
-    where_parts = ["sa.is_available = true"]
+    query_param = f"${param_index}"
+    param_index += 1
 
     if featured:
         where_parts.append("sa.featured = true")
 
     if creators:
+        where_parts.append(f"sa.creator_username = ANY(${param_index})")
         params.append(creators)
-        where_parts.append(f"sa.creator_username = ANY(${param_idx})")
-        param_idx += 1
+        param_index += 1
 
     if category:
+        where_parts.append(f"${param_index} = ANY(sa.categories)")
         params.append(category)
-        where_parts.append(f"${param_idx} = ANY(sa.categories)")
-        param_idx += 1
+        param_index += 1
 
     where_clause = " AND ".join(where_parts)
 
-    # Weights
-    params.append(weights.semantic)
-    w_semantic = f"${param_idx}"
-    param_idx += 1
+    # Determine if we can use hybrid search (have query embedding)
+    use_hybrid = query_embedding is not None
 
-    params.append(weights.lexical)
-    w_lexical = f"${param_idx}"
-    param_idx += 1
+    if use_hybrid:
+        # Add embedding parameter
+        embedding_str = embedding_to_vector_string(query_embedding)
+        params.append(embedding_str)
+        embedding_param = f"${param_index}"
+        param_index += 1
 
-    params.append(weights.category)
-    w_category = f"${param_idx}"
-    param_idx += 1
+        # Build hybrid search query with weighted scoring
+        # The semantic score is (1 - cosine_distance), normalized to [0,1]
+        # The lexical score is ts_rank_cd, normalized by max value
+        # Set search_path to include public for vector type visibility
+        sql_query = f"""
+            SET LOCAL search_path TO platform, public;
+            WITH search_scores AS (
+                SELECT
+                    sa.*,
+                    -- Semantic score: cosine similarity (1 - distance)
+                    COALESCE(1 - (sle.embedding <=> {embedding_param}::vector), 0) as semantic_score,
+                    -- Lexical score: ts_rank_cd normalized
+                    COALESCE(ts_rank_cd(sa.search, plainto_tsquery('english', {query_param})), 0) as lexical_raw,
+                    -- Category match: 1 if query term appears in categories, else 0
+                    CASE
+                        WHEN EXISTS (
+                            SELECT 1 FROM unnest(sa.categories) cat
+                            WHERE LOWER(cat) LIKE '%' || LOWER({query_param}) || '%'
+                        ) THEN 1.0
+                        ELSE 0.0
+                    END as category_score,
+                    -- Recency score: exponential decay over 90 days
+                    EXP(-EXTRACT(EPOCH FROM (NOW() - sa.updated_at)) / (90 * 24 * 3600)) as recency_score
+                FROM platform."StoreAgent" sa
+                LEFT JOIN platform."StoreListing" sl ON sa.slug = sl.slug
+                LEFT JOIN platform."StoreListingVersion" slv ON sl."activeVersionId" = slv.id
+                LEFT JOIN platform."StoreListingEmbedding" sle ON slv.id = sle."storeListingVersionId"
+                WHERE {where_clause}
+                AND (
+                    sa.search @@ plainto_tsquery('english', {query_param})
+                    OR sle.embedding IS NOT NULL
+                )
+            ),
+            normalized AS (
+                SELECT
+                    *,
+                    -- Normalize lexical score by max in result set
+                    CASE
+                        WHEN MAX(lexical_raw) OVER () > 0
+                        THEN lexical_raw / MAX(lexical_raw) OVER ()
+                        ELSE 0
+                    END as lexical_score
+                FROM search_scores
+            ),
+            scored AS (
+                SELECT
+                    slug,
+                    agent_name,
+                    agent_image,
+                    creator_username,
+                    creator_avatar,
+                    sub_heading,
+                    description,
+                    runs,
+                    rating,
+                    categories,
+                    featured,
+                    is_available,
+                    updated_at,
+                    semantic_score,
+                    lexical_score,
+                    category_score,
+                    recency_score,
+                    (
+                        {weights.semantic} * semantic_score +
+                        {weights.lexical} * lexical_score +
+                        {weights.category} * category_score +
+                        {weights.recency} * recency_score
+                    ) as combined_score
+                FROM normalized
+            )
+            SELECT * FROM scored
+            WHERE combined_score >= {min_score}
+            ORDER BY combined_score DESC
+            LIMIT ${param_index} OFFSET ${param_index + 1}
+        """
 
-    params.append(weights.recency)
-    w_recency = f"${param_idx}"
-    param_idx += 1
+        # Add pagination params
+        params.extend([page_size, offset])
 
-    params.append(weights.popularity)
-    w_popularity = f"${param_idx}"
-    param_idx += 1
+        # Count query - must also filter by min_score
+        count_query = f"""
+            SET LOCAL search_path TO platform, public;
+            WITH search_scores AS (
+                SELECT
+                    sa.slug,
+                    COALESCE(1 - (sle.embedding <=> {embedding_param}::vector), 0) as semantic_score,
+                    COALESCE(ts_rank_cd(sa.search, plainto_tsquery('english', {query_param})), 0) as lexical_raw,
+                    CASE
+                        WHEN EXISTS (
+                            SELECT 1 FROM unnest(sa.categories) cat
+                            WHERE LOWER(cat) LIKE '%' || LOWER({query_param}) || '%'
+                        ) THEN 1.0
+                        ELSE 0.0
+                    END as category_score,
+                    EXP(-EXTRACT(EPOCH FROM (NOW() - sa.updated_at)) / (90 * 24 * 3600)) as recency_score
+                FROM platform."StoreAgent" sa
+                LEFT JOIN platform."StoreListing" sl ON sa.slug = sl.slug
+                LEFT JOIN platform."StoreListingVersion" slv ON sl."activeVersionId" = slv.id
+                LEFT JOIN platform."StoreListingEmbedding" sle ON slv.id = sle."storeListingVersionId"
+                WHERE {where_clause}
+                AND (
+                    sa.search @@ plainto_tsquery('english', {query_param})
+                    OR sle.embedding IS NOT NULL
+                )
+            ),
+            normalized AS (
+                SELECT
+                    slug,
+                    semantic_score,
+                    category_score,
+                    recency_score,
+                    CASE
+                        WHEN MAX(lexical_raw) OVER () > 0
+                        THEN lexical_raw / MAX(lexical_raw) OVER ()
+                        ELSE 0
+                    END as lexical_score
+                FROM search_scores
+            ),
+            scored AS (
+                SELECT
+                    slug,
+                    (
+                        {weights.semantic} * semantic_score +
+                        {weights.lexical} * lexical_score +
+                        {weights.category} * category_score +
+                        {weights.recency} * recency_score
+                    ) as combined_score
+                FROM normalized
+            )
+            SELECT COUNT(*) as count FROM scored
+            WHERE combined_score >= {min_score}
+        """
 
-    params.append(min_score)
-    min_score_param = f"${param_idx}"
-    param_idx += 1
+    else:
+        # Fallback to lexical-only search (existing behavior)
+        # Note: For lexical-only, we still require tsvector match but don't
+        # apply min_score since ts_rank_cd isn't normalized to [0,1]
+        logger.warning("Falling back to lexical-only search (no query embedding)")
 
-    params.append(page_size)
-    limit_param = f"${param_idx}"
-    param_idx += 1
+        sql_query = f"""
+            WITH lexical_scores AS (
+                SELECT
+                    slug,
+                    agent_name,
+                    agent_image,
+                    creator_username,
+                    creator_avatar,
+                    sub_heading,
+                    description,
+                    runs,
+                    rating,
+                    categories,
+                    featured,
+                    is_available,
+                    updated_at,
+                    0.0 as semantic_score,
+                    ts_rank_cd(search, plainto_tsquery('english', {query_param})) as lexical_raw,
+                    CASE
+                        WHEN EXISTS (
+                            SELECT 1 FROM unnest(categories) cat
+                            WHERE LOWER(cat) LIKE '%' || LOWER({query_param}) || '%'
+                        ) THEN 1.0
+                        ELSE 0.0
+                    END as category_score,
+                    EXP(-EXTRACT(EPOCH FROM (NOW() - updated_at)) / (90 * 24 * 3600)) as recency_score
+                FROM platform."StoreAgent" sa
+                WHERE {where_clause}
+                AND search @@ plainto_tsquery('english', {query_param})
+            ),
+            normalized AS (
+                SELECT
+                    *,
+                    CASE
+                        WHEN MAX(lexical_raw) OVER () > 0
+                        THEN lexical_raw / MAX(lexical_raw) OVER ()
+                        ELSE 0
+                    END as lexical_score
+                FROM lexical_scores
+            ),
+            scored AS (
+                SELECT
+                    slug,
+                    agent_name,
+                    agent_image,
+                    creator_username,
+                    creator_avatar,
+                    sub_heading,
+                    description,
+                    runs,
+                    rating,
+                    categories,
+                    featured,
+                    is_available,
+                    updated_at,
+                    semantic_score,
+                    lexical_score,
+                    category_score,
+                    recency_score,
+                    (
+                        {weights.lexical} * lexical_score +
+                        {weights.category} * category_score +
+                        {weights.recency} * recency_score
+                    ) as combined_score
+                FROM normalized
+            )
+            SELECT * FROM scored
+            WHERE combined_score >= {min_score}
+            ORDER BY combined_score DESC
+            LIMIT ${param_index} OFFSET ${param_index + 1}
+        """
 
-    params.append(offset)
-    offset_param = f"${param_idx}"
-    param_idx += 1
+        params.extend([page_size, offset])
 
-    # Query using UnifiedContentEmbedding for search, StoreAgent for metadata
-    sql_query = f"""
-        WITH candidates AS (
-            -- Lexical matches via UnifiedContentEmbedding.search
-            SELECT uce."contentId" as "storeListingVersionId"
-            FROM {{schema_prefix}}"UnifiedContentEmbedding" uce
-            INNER JOIN {{schema_prefix}}"StoreAgent" sa
-                ON uce."contentId" = sa."storeListingVersionId"
-            WHERE uce."contentType" = 'STORE_AGENT'::{{schema_prefix}}"ContentType"
-            AND uce."userId" IS NULL
-            AND uce.search @@ plainto_tsquery('english', {query_param})
-            AND {where_clause}
+        count_query = f"""
+            WITH lexical_scores AS (
+                SELECT
+                    slug,
+                    ts_rank_cd(search, plainto_tsquery('english', {query_param})) as lexical_raw,
+                    CASE
+                        WHEN EXISTS (
+                            SELECT 1 FROM unnest(categories) cat
+                            WHERE LOWER(cat) LIKE '%' || LOWER({query_param}) || '%'
+                        ) THEN 1.0
+                        ELSE 0.0
+                    END as category_score,
+                    EXP(-EXTRACT(EPOCH FROM (NOW() - updated_at)) / (90 * 24 * 3600)) as recency_score
+                FROM platform."StoreAgent" sa
+                WHERE {where_clause}
+                AND search @@ plainto_tsquery('english', {query_param})
+            ),
+            normalized AS (
+                SELECT
+                    slug,
+                    category_score,
+                    recency_score,
+                    CASE
+                        WHEN MAX(lexical_raw) OVER () > 0
+                        THEN lexical_raw / MAX(lexical_raw) OVER ()
+                        ELSE 0
+                    END as lexical_score
+                FROM lexical_scores
+            ),
+            scored AS (
+                SELECT
+                    slug,
+                    (
+                        {weights.lexical} * lexical_score +
+                        {weights.category} * category_score +
+                        {weights.recency} * recency_score
+                    ) as combined_score
+                FROM normalized
+            )
+            SELECT COUNT(*) as count FROM scored
+            WHERE combined_score >= {min_score}
+        """
 
-            UNION
+    try:
+        # Execute search query
+        # Dynamic SQL is safe here - all user inputs are parameterized ($1, $2, etc.)
+        results = await client.query_raw(sql_query, *params)  # type: ignore[arg-type]
 
-            -- Semantic matches via UnifiedContentEmbedding.embedding
-            SELECT uce."contentId" as "storeListingVersionId"
-            FROM (
-                SELECT uce."contentId", uce.embedding
-                FROM {{schema_prefix}}"UnifiedContentEmbedding" uce
-                INNER JOIN {{schema_prefix}}"StoreAgent" sa
-                    ON uce."contentId" = sa."storeListingVersionId"
-                WHERE uce."contentType" = 'STORE_AGENT'::{{schema_prefix}}"ContentType"
-                AND uce."userId" IS NULL
-                AND {where_clause}
-                ORDER BY uce.embedding <=> {embedding_param}::vector
-                LIMIT 200
-            ) uce
-        ),
-        search_scores AS (
-            SELECT
-                sa.slug,
-                sa.agent_name,
-                sa.agent_image,
-                sa.creator_username,
-                sa.creator_avatar,
-                sa.sub_heading,
-                sa.description,
-                sa.runs,
-                sa.rating,
-                sa.categories,
-                sa.featured,
-                sa.is_available,
-                sa.updated_at,
-                -- Semantic score
-                COALESCE(1 - (uce.embedding <=> {embedding_param}::vector), 0) as semantic_score,
-                -- Lexical score (raw, will normalize)
-                COALESCE(ts_rank_cd(uce.search, plainto_tsquery('english', {query_param})), 0) as lexical_raw,
-                -- Category match
-                CASE
-                    WHEN EXISTS (
-                        SELECT 1 FROM unnest(sa.categories) cat
-                        WHERE LOWER(cat) LIKE '%' || {query_lower_param} || '%'
-                    )
-                    THEN 1.0
-                    ELSE 0.0
-                END as category_score,
-                -- Recency
-                GREATEST(0, 1 - EXTRACT(EPOCH FROM (NOW() - sa.updated_at)) / (90 * 24 * 3600)) as recency_score,
-                -- Popularity (raw)
-                sa.runs as popularity_raw
-            FROM candidates c
-            INNER JOIN {{schema_prefix}}"StoreAgent" sa
-                ON c."storeListingVersionId" = sa."storeListingVersionId"
-            INNER JOIN {{schema_prefix}}"UnifiedContentEmbedding" uce
-                ON sa."storeListingVersionId" = uce."contentId"
-                AND uce."contentType" = 'STORE_AGENT'::{{schema_prefix}}"ContentType"
-        ),
-        max_vals AS (
-            SELECT
-                GREATEST(MAX(lexical_raw), 0.001) as max_lexical,
-                GREATEST(MAX(popularity_raw), 1) as max_popularity
-            FROM search_scores
-        ),
-        normalized AS (
-            SELECT
-                ss.*,
-                ss.lexical_raw / mv.max_lexical as lexical_score,
-                CASE
-                    WHEN ss.popularity_raw > 0
-                    THEN LN(1 + ss.popularity_raw) / LN(1 + mv.max_popularity)
-                    ELSE 0
-                END as popularity_score
-            FROM search_scores ss
-            CROSS JOIN max_vals mv
-        ),
-        scored AS (
-            SELECT
-                slug,
-                agent_name,
-                agent_image,
-                creator_username,
-                creator_avatar,
-                sub_heading,
-                description,
-                runs,
-                rating,
-                categories,
-                featured,
-                is_available,
-                updated_at,
-                semantic_score,
-                lexical_score,
-                category_score,
-                recency_score,
-                popularity_score,
-                (
-                    {w_semantic} * semantic_score +
-                    {w_lexical} * lexical_score +
-                    {w_category} * category_score +
-                    {w_recency} * recency_score +
-                    {w_popularity} * popularity_score
-                ) as combined_score
-            FROM normalized
-        ),
-        filtered AS (
-            SELECT *, COUNT(*) OVER () as total_count
-            FROM scored
-            WHERE combined_score >= {min_score_param}
+        # Execute count query (without pagination params)
+        count_params = params[:-2]  # Remove LIMIT and OFFSET params
+        count_result = await client.query_raw(count_query, *count_params)  # type: ignore[arg-type]
+        total = count_result[0]["count"] if count_result else 0
+
+        logger.info(
+            f"Hybrid search for '{query}': {len(results)} results, {total} total "
+            f"(hybrid={use_hybrid})"
         )
-        SELECT * FROM filtered
-        ORDER BY combined_score DESC
-        LIMIT {limit_param} OFFSET {offset_param}
-    """
 
-    results = await query_raw_with_schema(
-        sql_query, *params, set_public_search_path=True
-    )
+        return results, total
 
-    total = results[0]["total_count"] if results else 0
-
-    for result in results:
-        result.pop("total_count", None)
-
-    logger.info(f"Hybrid search (store agents): {len(results)} results, {total} total")
-
-    return results, total
+    except Exception as e:
+        logger.error(f"Hybrid search failed: {e}")
+        raise
 
 
 async def hybrid_search_simple(
@@ -616,10 +428,13 @@ async def hybrid_search_simple(
     page: int = 1,
     page_size: int = 20,
 ) -> tuple[list[dict[str, Any]], int]:
-    """Simplified hybrid search for store agents."""
-    return await hybrid_search(query=query, page=page, page_size=page_size)
+    """
+    Simplified hybrid search for common use cases.
 
-
-# Backward compatibility alias - HybridSearchWeights maps to StoreAgentSearchWeights
-# for existing code that expects the popularity parameter
-HybridSearchWeights = StoreAgentSearchWeights
+    Uses default weights and no filters.
+    """
+    return await hybrid_search(
+        query=query,
+        page=page,
+        page_size=page_size,
+    )

autogpt_platform/backend/backend/api/features/store/hybrid_search_test.py

@@ -1,667 +0,0 @@
-"""
-Integration tests for hybrid search with schema handling.
-
-These tests verify that hybrid search works correctly across different database schemas.
-"""
-
-from unittest.mock import patch
-
-import pytest
-from prisma.enums import ContentType
-
-from backend.api.features.store import embeddings
-from backend.api.features.store.hybrid_search import (
-    HybridSearchWeights,
-    UnifiedSearchWeights,
-    hybrid_search,
-    unified_hybrid_search,
-)
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_hybrid_search_with_schema_handling():
-    """Test that hybrid search correctly handles database schema prefixes."""
-    # Test with a mock query to ensure schema handling works
-    query = "test agent"
-
-    with patch(
-        "backend.api.features.store.hybrid_search.query_raw_with_schema"
-    ) as mock_query:
-        # Mock the query result
-        mock_query.return_value = [
-            {
-                "slug": "test/agent",
-                "agent_name": "Test Agent",
-                "agent_image": "test.png",
-                "creator_username": "test",
-                "creator_avatar": "avatar.png",
-                "sub_heading": "Test sub-heading",
-                "description": "Test description",
-                "runs": 10,
-                "rating": 4.5,
-                "categories": ["test"],
-                "featured": False,
-                "is_available": True,
-                "updated_at": "2024-01-01T00:00:00Z",
-                "combined_score": 0.8,
-                "semantic_score": 0.7,
-                "lexical_score": 0.6,
-                "category_score": 0.5,
-                "recency_score": 0.4,
-                "total_count": 1,
-            }
-        ]
-
-        with patch(
-            "backend.api.features.store.hybrid_search.embed_query"
-        ) as mock_embed:
-            mock_embed.return_value = [0.1] * embeddings.EMBEDDING_DIM  # Mock embedding
-
-            results, total = await hybrid_search(
-                query=query,
-                page=1,
-                page_size=20,
-            )
-
-            # Verify the query was called
-            assert mock_query.called
-            # Verify the SQL template uses schema_prefix placeholder
-            call_args = mock_query.call_args
-            sql_template = call_args[0][0]
-            assert "{schema_prefix}" in sql_template
-
-            # Verify results
-            assert len(results) == 1
-            assert total == 1
-            assert results[0]["slug"] == "test/agent"
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_hybrid_search_with_public_schema():
-    """Test hybrid search when using public schema (no prefix needed)."""
-    with patch("backend.data.db.get_database_schema") as mock_schema:
-        mock_schema.return_value = "public"
-
-        with patch(
-            "backend.api.features.store.hybrid_search.query_raw_with_schema"
-        ) as mock_query:
-            mock_query.return_value = []
-
-            with patch(
-                "backend.api.features.store.hybrid_search.embed_query"
-            ) as mock_embed:
-                mock_embed.return_value = [0.1] * embeddings.EMBEDDING_DIM
-
-                results, total = await hybrid_search(
-                    query="test",
-                    page=1,
-                    page_size=20,
-                )
-
-                # Verify the mock was set up correctly
-                assert mock_schema.return_value == "public"
-
-                # Results should work even with empty results
-                assert results == []
-                assert total == 0
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_hybrid_search_with_custom_schema():
-    """Test hybrid search when using custom schema (e.g., 'platform')."""
-    with patch("backend.data.db.get_database_schema") as mock_schema:
-        mock_schema.return_value = "platform"
-
-        with patch(
-            "backend.api.features.store.hybrid_search.query_raw_with_schema"
-        ) as mock_query:
-            mock_query.return_value = []
-
-            with patch(
-                "backend.api.features.store.hybrid_search.embed_query"
-            ) as mock_embed:
-                mock_embed.return_value = [0.1] * embeddings.EMBEDDING_DIM
-
-                results, total = await hybrid_search(
-                    query="test",
-                    page=1,
-                    page_size=20,
-                )
-
-                # Verify the mock was set up correctly
-                assert mock_schema.return_value == "platform"
-
-                assert results == []
-                assert total == 0
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_hybrid_search_without_embeddings():
-    """Test hybrid search gracefully degrades when embeddings are unavailable."""
-    # Mock database to return some results
-    mock_results = [
-        {
-            "slug": "test-agent",
-            "agent_name": "Test Agent",
-            "agent_image": "test.png",
-            "creator_username": "creator",
-            "creator_avatar": "avatar.png",
-            "sub_heading": "Test heading",
-            "description": "Test description",
-            "runs": 100,
-            "rating": 4.5,
-            "categories": ["AI"],
-            "featured": False,
-            "is_available": True,
-            "updated_at": "2025-01-01T00:00:00Z",
-            "semantic_score": 0.0,  # Zero because no embedding
-            "lexical_score": 0.5,
-            "category_score": 0.0,
-            "recency_score": 0.1,
-            "popularity_score": 0.2,
-            "combined_score": 0.3,
-            "total_count": 1,
-        }
-    ]
-
-    with patch("backend.api.features.store.hybrid_search.embed_query") as mock_embed:
-        with patch(
-            "backend.api.features.store.hybrid_search.query_raw_with_schema"
-        ) as mock_query:
-            # Simulate embedding failure
-            mock_embed.return_value = None
-            mock_query.return_value = mock_results
-
-            # Should NOT raise - graceful degradation
-            results, total = await hybrid_search(
-                query="test",
-                page=1,
-                page_size=20,
-            )
-
-            # Verify it returns results even without embeddings
-            assert len(results) == 1
-            assert results[0]["slug"] == "test-agent"
-            assert total == 1
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_hybrid_search_with_filters():
-    """Test hybrid search with various filters."""
-    with patch(
-        "backend.api.features.store.hybrid_search.query_raw_with_schema"
-    ) as mock_query:
-        mock_query.return_value = []
-
-        with patch(
-            "backend.api.features.store.hybrid_search.embed_query"
-        ) as mock_embed:
-            mock_embed.return_value = [0.1] * embeddings.EMBEDDING_DIM
-
-            # Test with featured filter
-            results, total = await hybrid_search(
-                query="test",
-                featured=True,
-                creators=["user1", "user2"],
-                category="productivity",
-                page=1,
-                page_size=10,
-            )
-
-            # Verify filters were applied in the query
-            call_args = mock_query.call_args
-            params = call_args[0][1:]  # Skip SQL template
-
-            # Should have query, query_lower, creators array, category
-            assert len(params) >= 4
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_hybrid_search_weights():
-    """Test hybrid search with custom weights."""
-    custom_weights = HybridSearchWeights(
-        semantic=0.5,
-        lexical=0.3,
-        category=0.1,
-        recency=0.1,
-        popularity=0.0,
-    )
-
-    with patch(
-        "backend.api.features.store.hybrid_search.query_raw_with_schema"
-    ) as mock_query:
-        mock_query.return_value = []
-
-        with patch(
-            "backend.api.features.store.hybrid_search.embed_query"
-        ) as mock_embed:
-            mock_embed.return_value = [0.1] * embeddings.EMBEDDING_DIM
-
-            results, total = await hybrid_search(
-                query="test",
-                weights=custom_weights,
-                page=1,
-                page_size=20,
-            )
-
-            # Verify custom weights were used in the query
-            call_args = mock_query.call_args
-            sql_template = call_args[0][0]
-            params = call_args[0][1:]  # Get all parameters passed
-
-            # Check that SQL uses parameterized weights (not f-string interpolation)
-            assert "$" in sql_template  # Verify parameterization is used
-
-            # Check that custom weights are in the params
-            assert 0.5 in params  # semantic weight
-            assert 0.3 in params  # lexical weight
-            assert 0.1 in params  # category and recency weights
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_hybrid_search_min_score_filtering():
-    """Test hybrid search minimum score threshold."""
-    with patch(
-        "backend.api.features.store.hybrid_search.query_raw_with_schema"
-    ) as mock_query:
-        # Return results with varying scores
-        mock_query.return_value = [
-            {
-                "slug": "high-score/agent",
-                "agent_name": "High Score Agent",
-                "combined_score": 0.8,
-                "total_count": 1,
-                # ... other fields
-            }
-        ]
-
-        with patch(
-            "backend.api.features.store.hybrid_search.embed_query"
-        ) as mock_embed:
-            mock_embed.return_value = [0.1] * embeddings.EMBEDDING_DIM
-
-            # Test with custom min_score
-            results, total = await hybrid_search(
-                query="test",
-                min_score=0.5,  # High threshold
-                page=1,
-                page_size=20,
-            )
-
-            # Verify min_score was applied in query
-            call_args = mock_query.call_args
-            sql_template = call_args[0][0]
-            params = call_args[0][1:]  # Get all parameters
-
-            # Check that SQL uses parameterized min_score
-            assert "combined_score >=" in sql_template
-            assert "$" in sql_template  # Verify parameterization
-
-            # Check that custom min_score is in the params
-            assert 0.5 in params
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_hybrid_search_pagination():
-    """Test hybrid search pagination."""
-    with patch(
-        "backend.api.features.store.hybrid_search.query_raw_with_schema"
-    ) as mock_query:
-        mock_query.return_value = []
-
-        with patch(
-            "backend.api.features.store.hybrid_search.embed_query"
-        ) as mock_embed:
-            mock_embed.return_value = [0.1] * embeddings.EMBEDDING_DIM
-
-            # Test page 2 with page_size 10
-            results, total = await hybrid_search(
-                query="test",
-                page=2,
-                page_size=10,
-            )
-
-            # Verify pagination parameters
-            call_args = mock_query.call_args
-            params = call_args[0]
-
-            # Last two params should be LIMIT and OFFSET
-            limit = params[-2]
-            offset = params[-1]
-
-            assert limit == 10  # page_size
-            assert offset == 10  # (page - 1) * page_size = (2 - 1) * 10
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_hybrid_search_error_handling():
-    """Test hybrid search error handling."""
-    with patch(
-        "backend.api.features.store.hybrid_search.query_raw_with_schema"
-    ) as mock_query:
-        # Simulate database error
-        mock_query.side_effect = Exception("Database connection error")
-
-        with patch(
-            "backend.api.features.store.hybrid_search.embed_query"
-        ) as mock_embed:
-            mock_embed.return_value = [0.1] * embeddings.EMBEDDING_DIM
-
-            # Should raise exception
-            with pytest.raises(Exception) as exc_info:
-                await hybrid_search(
-                    query="test",
-                    page=1,
-                    page_size=20,
-                )
-
-            assert "Database connection error" in str(exc_info.value)
-
-
-# =============================================================================
-# Unified Hybrid Search Tests
-# =============================================================================
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_unified_hybrid_search_basic():
-    """Test basic unified hybrid search across all content types."""
-    mock_results = [
-        {
-            "content_type": "STORE_AGENT",
-            "content_id": "agent-1",
-            "searchable_text": "Test Agent Description",
-            "metadata": {"name": "Test Agent"},
-            "updated_at": "2025-01-01T00:00:00Z",
-            "semantic_score": 0.7,
-            "lexical_score": 0.8,
-            "category_score": 0.5,
-            "recency_score": 0.3,
-            "combined_score": 0.6,
-            "total_count": 2,
-        },
-        {
-            "content_type": "BLOCK",
-            "content_id": "block-1",
-            "searchable_text": "Test Block Description",
-            "metadata": {"name": "Test Block"},
-            "updated_at": "2025-01-01T00:00:00Z",
-            "semantic_score": 0.6,
-            "lexical_score": 0.7,
-            "category_score": 0.4,
-            "recency_score": 0.2,
-            "combined_score": 0.5,
-            "total_count": 2,
-        },
-    ]
-
-    with patch(
-        "backend.api.features.store.hybrid_search.query_raw_with_schema"
-    ) as mock_query:
-        with patch(
-            "backend.api.features.store.hybrid_search.embed_query"
-        ) as mock_embed:
-            mock_query.return_value = mock_results
-            mock_embed.return_value = [0.1] * embeddings.EMBEDDING_DIM
-
-            results, total = await unified_hybrid_search(
-                query="test",
-                page=1,
-                page_size=20,
-            )
-
-            assert len(results) == 2
-            assert total == 2
-            assert results[0]["content_type"] == "STORE_AGENT"
-            assert results[1]["content_type"] == "BLOCK"
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_unified_hybrid_search_filter_by_content_type():
-    """Test unified search filtering by specific content types."""
-    mock_results = [
-        {
-            "content_type": "BLOCK",
-            "content_id": "block-1",
-            "searchable_text": "Test Block",
-            "metadata": {},
-            "updated_at": "2025-01-01T00:00:00Z",
-            "semantic_score": 0.7,
-            "lexical_score": 0.8,
-            "category_score": 0.0,
-            "recency_score": 0.3,
-            "combined_score": 0.5,
-            "total_count": 1,
-        },
-    ]
-
-    with patch(
-        "backend.api.features.store.hybrid_search.query_raw_with_schema"
-    ) as mock_query:
-        with patch(
-            "backend.api.features.store.hybrid_search.embed_query"
-        ) as mock_embed:
-            mock_query.return_value = mock_results
-            mock_embed.return_value = [0.1] * embeddings.EMBEDDING_DIM
-
-            results, total = await unified_hybrid_search(
-                query="test",
-                content_types=[ContentType.BLOCK],
-                page=1,
-                page_size=20,
-            )
-
-            # Verify content_types parameter was passed correctly
-            call_args = mock_query.call_args
-            params = call_args[0][1:]
-            # The content types should be in the params as a list
-            assert ["BLOCK"] in params
-
-            assert len(results) == 1
-            assert total == 1
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_unified_hybrid_search_with_user_id():
-    """Test unified search with user_id for private content."""
-    mock_results = [
-        {
-            "content_type": "STORE_AGENT",
-            "content_id": "agent-1",
-            "searchable_text": "My Private Agent",
-            "metadata": {},
-            "updated_at": "2025-01-01T00:00:00Z",
-            "semantic_score": 0.7,
-            "lexical_score": 0.8,
-            "category_score": 0.0,
-            "recency_score": 0.3,
-            "combined_score": 0.6,
-            "total_count": 1,
-        },
-    ]
-
-    with patch(
-        "backend.api.features.store.hybrid_search.query_raw_with_schema"
-    ) as mock_query:
-        with patch(
-            "backend.api.features.store.hybrid_search.embed_query"
-        ) as mock_embed:
-            mock_query.return_value = mock_results
-            mock_embed.return_value = [0.1] * embeddings.EMBEDDING_DIM
-
-            results, total = await unified_hybrid_search(
-                query="test",
-                user_id="user-123",
-                page=1,
-                page_size=20,
-            )
-
-            # Verify SQL contains user_id filter
-            call_args = mock_query.call_args
-            sql_template = call_args[0][0]
-            params = call_args[0][1:]
-
-            assert 'uce."userId"' in sql_template
-            assert "user-123" in params
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_unified_hybrid_search_custom_weights():
-    """Test unified search with custom weights."""
-    custom_weights = UnifiedSearchWeights(
-        semantic=0.6,
-        lexical=0.2,
-        category=0.1,
-        recency=0.1,
-    )
-
-    with patch(
-        "backend.api.features.store.hybrid_search.query_raw_with_schema"
-    ) as mock_query:
-        with patch(
-            "backend.api.features.store.hybrid_search.embed_query"
-        ) as mock_embed:
-            mock_query.return_value = []
-            mock_embed.return_value = [0.1] * embeddings.EMBEDDING_DIM
-
-            results, total = await unified_hybrid_search(
-                query="test",
-                weights=custom_weights,
-                page=1,
-                page_size=20,
-            )
-
-            # Verify custom weights are in parameters
-            call_args = mock_query.call_args
-            params = call_args[0][1:]
-
-            assert 0.6 in params  # semantic weight
-            assert 0.2 in params  # lexical weight
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_unified_hybrid_search_graceful_degradation():
-    """Test unified search gracefully degrades when embeddings unavailable."""
-    mock_results = [
-        {
-            "content_type": "DOCUMENTATION",
-            "content_id": "doc-1",
-            "searchable_text": "API Documentation",
-            "metadata": {},
-            "updated_at": "2025-01-01T00:00:00Z",
-            "semantic_score": 0.0,  # Zero because no embedding
-            "lexical_score": 0.8,
-            "category_score": 0.0,
-            "recency_score": 0.2,
-            "combined_score": 0.5,
-            "total_count": 1,
-        },
-    ]
-
-    with patch(
-        "backend.api.features.store.hybrid_search.query_raw_with_schema"
-    ) as mock_query:
-        with patch(
-            "backend.api.features.store.hybrid_search.embed_query"
-        ) as mock_embed:
-            mock_query.return_value = mock_results
-            mock_embed.return_value = None  # Embedding failure
-
-            # Should NOT raise - graceful degradation
-            results, total = await unified_hybrid_search(
-                query="test",
-                page=1,
-                page_size=20,
-            )
-
-            assert len(results) == 1
-            assert total == 1
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_unified_hybrid_search_empty_query():
-    """Test unified search with empty query returns empty results."""
-    results, total = await unified_hybrid_search(
-        query="",
-        page=1,
-        page_size=20,
-    )
-
-    assert results == []
-    assert total == 0
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_unified_hybrid_search_pagination():
-    """Test unified search pagination."""
-    with patch(
-        "backend.api.features.store.hybrid_search.query_raw_with_schema"
-    ) as mock_query:
-        with patch(
-            "backend.api.features.store.hybrid_search.embed_query"
-        ) as mock_embed:
-            mock_query.return_value = []
-            mock_embed.return_value = [0.1] * embeddings.EMBEDDING_DIM
-
-            results, total = await unified_hybrid_search(
-                query="test",
-                page=3,
-                page_size=15,
-            )
-
-            # Verify pagination parameters (last two params are LIMIT and OFFSET)
-            call_args = mock_query.call_args
-            params = call_args[0]
-
-            limit = params[-2]
-            offset = params[-1]
-
-            assert limit == 15  # page_size
-            assert offset == 30  # (page - 1) * page_size = (3 - 1) * 15
-
-
-@pytest.mark.asyncio(loop_scope="session")
-@pytest.mark.integration
-async def test_unified_hybrid_search_schema_prefix():
-    """Test unified search uses schema_prefix placeholder."""
-    with patch(
-        "backend.api.features.store.hybrid_search.query_raw_with_schema"
-    ) as mock_query:
-        with patch(
-            "backend.api.features.store.hybrid_search.embed_query"
-        ) as mock_embed:
-            mock_query.return_value = []
-            mock_embed.return_value = [0.1] * embeddings.EMBEDDING_DIM
-
-            await unified_hybrid_search(
-                query="test",
-                page=1,
-                page_size=20,
-            )
-
-            call_args = mock_query.call_args
-            sql_template = call_args[0][0]
-
-            # Verify schema_prefix placeholder is used for table references
-            assert "{schema_prefix}" in sql_template
-            assert '"UnifiedContentEmbedding"' in sql_template
-
-
-if __name__ == "__main__":
-    pytest.main([__file__, "-v", "-s"])

autogpt_platform/backend/backend/api/features/store/model.py

@@ -110,7 +110,6 @@ class Profile(pydantic.BaseModel):
 
 
 class StoreSubmission(pydantic.BaseModel):
-    listing_id: str
     agent_id: str
     agent_version: int
     name: str
@@ -165,12 +164,8 @@ class StoreListingsWithVersionsResponse(pydantic.BaseModel):
 
 
 class StoreSubmissionRequest(pydantic.BaseModel):
-    agent_id: str = pydantic.Field(
-        ..., min_length=1, description="Agent ID cannot be empty"
-    )
-    agent_version: int = pydantic.Field(
-        ..., gt=0, description="Agent version must be greater than 0"
-    )
+    agent_id: str
+    agent_version: int
     slug: str
     name: str
     sub_heading: str
@@ -221,23 +216,3 @@ class ReviewSubmissionRequest(pydantic.BaseModel):
     is_approved: bool
     comments: str  # External comments visible to creator
     internal_comments: str | None = None  # Private admin notes
-
-
-class UnifiedSearchResult(pydantic.BaseModel):
-    """A single result from unified hybrid search across all content types."""
-
-    content_type: str  # STORE_AGENT, BLOCK, DOCUMENTATION
-    content_id: str
-    searchable_text: str
-    metadata: dict | None = None
-    updated_at: datetime.datetime | None = None
-    combined_score: float | None = None
-    semantic_score: float | None = None
-    lexical_score: float | None = None
-
-
-class UnifiedSearchResponse(pydantic.BaseModel):
-    """Response model for unified search across all content types."""
-
-    results: list[UnifiedSearchResult]
-    pagination: Pagination

autogpt_platform/backend/backend/api/features/store/model_test.py

@@ -138,7 +138,6 @@ def test_creator_details():
 
 def test_store_submission():
     submission = store_model.StoreSubmission(
-        listing_id="listing123",
         agent_id="agent123",
         agent_version=1,
         sub_heading="Test subheading",
@@ -160,7 +159,6 @@ def test_store_submissions_response():
     response = store_model.StoreSubmissionsResponse(
         submissions=[
             store_model.StoreSubmission(
-                listing_id="listing123",
                 agent_id="agent123",
                 agent_version=1,
                 sub_heading="Test subheading",

autogpt_platform/backend/backend/api/features/store/routes.py

@@ -7,15 +7,12 @@ from typing import Literal
 import autogpt_libs.auth
 import fastapi
 import fastapi.responses
-import prisma.enums
 
 import backend.data.graph
 import backend.util.json
-from backend.util.models import Pagination
 
 from . import cache as store_cache
 from . import db as store_db
-from . import hybrid_search as store_hybrid_search
 from . import image_gen as store_image_gen
 from . import media as store_media
 from . import model as store_model
@@ -149,102 +146,6 @@ async def get_agents(
     return agents
 
 
-##############################################
-############### Search Endpoints #############
-##############################################
-
-
-@router.get(
-    "/search",
-    summary="Unified search across all content types",
-    tags=["store", "public"],
-    response_model=store_model.UnifiedSearchResponse,
-)
-async def unified_search(
-    query: str,
-    content_types: list[str] | None = fastapi.Query(
-        default=None,
-        description="Content types to search: STORE_AGENT, BLOCK, DOCUMENTATION. If not specified, searches all.",
-    ),
-    page: int = 1,
-    page_size: int = 20,
-    user_id: str | None = fastapi.Security(
-        autogpt_libs.auth.get_optional_user_id, use_cache=False
-    ),
-):
-    """
-    Search across all content types (store agents, blocks, documentation) using hybrid search.
-
-    Combines semantic (embedding-based) and lexical (text-based) search for best results.
-
-    Args:
-        query: The search query string
-        content_types: Optional list of content types to filter by (STORE_AGENT, BLOCK, DOCUMENTATION)
-        page: Page number for pagination (default 1)
-        page_size: Number of results per page (default 20)
-        user_id: Optional authenticated user ID (for user-scoped content in future)
-
-    Returns:
-        UnifiedSearchResponse: Paginated list of search results with relevance scores
-    """
-    if page < 1:
-        raise fastapi.HTTPException(
-            status_code=422, detail="Page must be greater than 0"
-        )
-
-    if page_size < 1:
-        raise fastapi.HTTPException(
-            status_code=422, detail="Page size must be greater than 0"
-        )
-
-    # Convert string content types to enum
-    content_type_enums: list[prisma.enums.ContentType] | None = None
-    if content_types:
-        try:
-            content_type_enums = [prisma.enums.ContentType(ct) for ct in content_types]
-        except ValueError as e:
-            raise fastapi.HTTPException(
-                status_code=422,
-                detail=f"Invalid content type. Valid values: STORE_AGENT, BLOCK, DOCUMENTATION. Error: {e}",
-            )
-
-    # Perform unified hybrid search
-    results, total = await store_hybrid_search.unified_hybrid_search(
-        query=query,
-        content_types=content_type_enums,
-        user_id=user_id,
-        page=page,
-        page_size=page_size,
-    )
-
-    # Convert results to response model
-    search_results = [
-        store_model.UnifiedSearchResult(
-            content_type=r["content_type"],
-            content_id=r["content_id"],
-            searchable_text=r.get("searchable_text", ""),
-            metadata=r.get("metadata"),
-            updated_at=r.get("updated_at"),
-            combined_score=r.get("combined_score"),
-            semantic_score=r.get("semantic_score"),
-            lexical_score=r.get("lexical_score"),
-        )
-        for r in results
-    ]
-
-    total_pages = (total + page_size - 1) // page_size if total > 0 else 0
-
-    return store_model.UnifiedSearchResponse(
-        results=search_results,
-        pagination=Pagination(
-            total_items=total,
-            total_pages=total_pages,
-            current_page=page,
-            page_size=page_size,
-        ),
-    )
-
-
 @router.get(
     "/agents/{username}/{agent_name}",
     summary="Get specific agent",

autogpt_platform/backend/backend/api/features/store/routes_test.py

@@ -521,7 +521,6 @@ def test_get_submissions_success(
     mocked_value = store_model.StoreSubmissionsResponse(
         submissions=[
             store_model.StoreSubmission(
-                listing_id="test-listing-id",
                 name="Test Agent",
                 description="Test agent description",
                 image_urls=["test.jpg"],

autogpt_platform/backend/backend/api/features/store/semantic_search_test.py

@@ -1,272 +0,0 @@
-"""Tests for the semantic_search function."""
-
-import pytest
-from prisma.enums import ContentType
-
-from backend.api.features.store.embeddings import EMBEDDING_DIM, semantic_search
-
-
-@pytest.mark.asyncio
-async def test_search_blocks_only(mocker):
-    """Test searching only BLOCK content type."""
-    # Mock embed_query to return a test embedding
-    mock_embedding = [0.1] * EMBEDDING_DIM
-    mocker.patch(
-        "backend.api.features.store.embeddings.embed_query",
-        return_value=mock_embedding,
-    )
-
-    # Mock query_raw_with_schema to return test results
-    mock_results = [
-        {
-            "content_id": "block-123",
-            "content_type": "BLOCK",
-            "searchable_text": "Calculator Block - Performs arithmetic operations",
-            "metadata": {"name": "Calculator", "categories": ["Math"]},
-            "similarity": 0.85,
-        }
-    ]
-    mocker.patch(
-        "backend.api.features.store.embeddings.query_raw_with_schema",
-        return_value=mock_results,
-    )
-
-    results = await semantic_search(
-        query="calculate numbers",
-        content_types=[ContentType.BLOCK],
-    )
-
-    assert len(results) == 1
-    assert results[0]["content_type"] == "BLOCK"
-    assert results[0]["content_id"] == "block-123"
-    assert results[0]["similarity"] == 0.85
-
-
-@pytest.mark.asyncio
-async def test_search_multiple_content_types(mocker):
-    """Test searching multiple content types simultaneously."""
-    mock_embedding = [0.1] * EMBEDDING_DIM
-    mocker.patch(
-        "backend.api.features.store.embeddings.embed_query",
-        return_value=mock_embedding,
-    )
-
-    mock_results = [
-        {
-            "content_id": "block-123",
-            "content_type": "BLOCK",
-            "searchable_text": "Calculator Block",
-            "metadata": {},
-            "similarity": 0.85,
-        },
-        {
-            "content_id": "doc-456",
-            "content_type": "DOCUMENTATION",
-            "searchable_text": "How to use Calculator",
-            "metadata": {},
-            "similarity": 0.75,
-        },
-    ]
-    mocker.patch(
-        "backend.api.features.store.embeddings.query_raw_with_schema",
-        return_value=mock_results,
-    )
-
-    results = await semantic_search(
-        query="calculator",
-        content_types=[ContentType.BLOCK, ContentType.DOCUMENTATION],
-    )
-
-    assert len(results) == 2
-    assert results[0]["content_type"] == "BLOCK"
-    assert results[1]["content_type"] == "DOCUMENTATION"
-
-
-@pytest.mark.asyncio
-async def test_search_with_min_similarity_threshold(mocker):
-    """Test that results below min_similarity are filtered out."""
-    mock_embedding = [0.1] * EMBEDDING_DIM
-    mocker.patch(
-        "backend.api.features.store.embeddings.embed_query",
-        return_value=mock_embedding,
-    )
-
-    # Only return results above 0.7 similarity
-    mock_results = [
-        {
-            "content_id": "block-123",
-            "content_type": "BLOCK",
-            "searchable_text": "Calculator Block",
-            "metadata": {},
-            "similarity": 0.85,
-        }
-    ]
-    mocker.patch(
-        "backend.api.features.store.embeddings.query_raw_with_schema",
-        return_value=mock_results,
-    )
-
-    results = await semantic_search(
-        query="calculate",
-        content_types=[ContentType.BLOCK],
-        min_similarity=0.7,
-    )
-
-    assert len(results) == 1
-    assert results[0]["similarity"] >= 0.7
-
-
-@pytest.mark.asyncio
-async def test_search_fallback_to_lexical(mocker):
-    """Test fallback to lexical search when embeddings fail."""
-    # Mock embed_query to return None (embeddings unavailable)
-    mocker.patch(
-        "backend.api.features.store.embeddings.embed_query",
-        return_value=None,
-    )
-
-    mock_lexical_results = [
-        {
-            "content_id": "block-123",
-            "content_type": "BLOCK",
-            "searchable_text": "Calculator Block performs calculations",
-            "metadata": {},
-            "similarity": 0.0,
-        }
-    ]
-    mocker.patch(
-        "backend.api.features.store.embeddings.query_raw_with_schema",
-        return_value=mock_lexical_results,
-    )
-
-    results = await semantic_search(
-        query="calculator",
-        content_types=[ContentType.BLOCK],
-    )
-
-    assert len(results) == 1
-    assert results[0]["similarity"] == 0.0  # Lexical search returns 0 similarity
-
-
-@pytest.mark.asyncio
-async def test_search_empty_query():
-    """Test that empty query returns no results."""
-    results = await semantic_search(query="")
-    assert results == []
-
-    results = await semantic_search(query="   ")
-    assert results == []
-
-
-@pytest.mark.asyncio
-async def test_search_with_user_id_filter(mocker):
-    """Test searching with user_id filter for private content."""
-    mock_embedding = [0.1] * EMBEDDING_DIM
-    mocker.patch(
-        "backend.api.features.store.embeddings.embed_query",
-        return_value=mock_embedding,
-    )
-
-    mock_results = [
-        {
-            "content_id": "agent-789",
-            "content_type": "LIBRARY_AGENT",
-            "searchable_text": "My Custom Agent",
-            "metadata": {},
-            "similarity": 0.9,
-        }
-    ]
-    mocker.patch(
-        "backend.api.features.store.embeddings.query_raw_with_schema",
-        return_value=mock_results,
-    )
-
-    results = await semantic_search(
-        query="custom agent",
-        content_types=[ContentType.LIBRARY_AGENT],
-        user_id="user-123",
-    )
-
-    assert len(results) == 1
-    assert results[0]["content_type"] == "LIBRARY_AGENT"
-
-
-@pytest.mark.asyncio
-async def test_search_limit_parameter(mocker):
-    """Test that limit parameter correctly limits results."""
-    mock_embedding = [0.1] * EMBEDDING_DIM
-    mocker.patch(
-        "backend.api.features.store.embeddings.embed_query",
-        return_value=mock_embedding,
-    )
-
-    # Return 5 results
-    mock_results = [
-        {
-            "content_id": f"block-{i}",
-            "content_type": "BLOCK",
-            "searchable_text": f"Block {i}",
-            "metadata": {},
-            "similarity": 0.8,
-        }
-        for i in range(5)
-    ]
-    mocker.patch(
-        "backend.api.features.store.embeddings.query_raw_with_schema",
-        return_value=mock_results,
-    )
-
-    results = await semantic_search(
-        query="block",
-        content_types=[ContentType.BLOCK],
-        limit=5,
-    )
-
-    assert len(results) == 5
-
-
-@pytest.mark.asyncio
-async def test_search_default_content_types(mocker):
-    """Test that default content_types includes BLOCK, STORE_AGENT, and DOCUMENTATION."""
-    mock_embedding = [0.1] * EMBEDDING_DIM
-    mocker.patch(
-        "backend.api.features.store.embeddings.embed_query",
-        return_value=mock_embedding,
-    )
-
-    mock_query_raw = mocker.patch(
-        "backend.api.features.store.embeddings.query_raw_with_schema",
-        return_value=[],
-    )
-
-    await semantic_search(query="test")
-
-    # Check that the SQL query includes all three default content types
-    call_args = mock_query_raw.call_args
-    assert "BLOCK" in str(call_args)
-    assert "STORE_AGENT" in str(call_args)
-    assert "DOCUMENTATION" in str(call_args)
-
-
-@pytest.mark.asyncio
-async def test_search_handles_database_error(mocker):
-    """Test that database errors are handled gracefully."""
-    mock_embedding = [0.1] * EMBEDDING_DIM
-    mocker.patch(
-        "backend.api.features.store.embeddings.embed_query",
-        return_value=mock_embedding,
-    )
-
-    # Simulate database error
-    mocker.patch(
-        "backend.api.features.store.embeddings.query_raw_with_schema",
-        side_effect=Exception("Database connection failed"),
-    )
-
-    results = await semantic_search(
-        query="test",
-        content_types=[ContentType.BLOCK],
-    )
-
-    # Should return empty list on error
-    assert results == []

autogpt_platform/backend/backend/api/features/v1.py

@@ -64,6 +64,7 @@ from backend.data.onboarding import (
     complete_re_run_agent,
     get_recommended_agents,
     get_user_onboarding,
+    increment_runs,
     onboarding_enabled,
     reset_user_onboarding,
     update_user_onboarding,
@@ -693,13 +694,13 @@ class DeleteGraphResponse(TypedDict):
 async def list_graphs(
     user_id: Annotated[str, Security(get_user_id)],
 ) -> Sequence[graph_db.GraphMeta]:
-    graphs, _ = await graph_db.list_graphs_paginated(
+    paginated_result = await graph_db.list_graphs_paginated(
         user_id=user_id,
         page=1,
         page_size=250,
         filter_by="active",
     )
-    return graphs
+    return paginated_result.graphs
 
 
 @v1_router.get(
@@ -974,6 +975,7 @@ async def execute_graph(
         # Record successful graph execution
         record_graph_execution(graph_id=graph_id, status="success", user_id=user_id)
         record_graph_operation(operation="execute", status="success")
+        await increment_runs(user_id)
         await complete_re_run_agent(user_id, graph_id)
         if source == "library":
             await complete_onboarding_step(

autogpt_platform/backend/backend/blocks/airtable/_webhook.py

@@ -6,9 +6,6 @@ import hashlib
 import hmac
 import logging
 from enum import Enum
-from typing import cast
-
-from prisma.types import Serializable
 
 from backend.sdk import (
     BaseWebhooksManager,
@@ -87,9 +84,7 @@ class AirtableWebhookManager(BaseWebhooksManager):
         # update webhook config
         await update_webhook(
             webhook.id,
-            config=cast(
-                dict[str, Serializable], {"base_id": base_id, "cursor": response.cursor}
-            ),
+            config={"base_id": base_id, "cursor": response.cursor},
         )
 
         event_type = "notification"

autogpt_platform/backend/backend/blocks/helpers/review.py

@@ -1,184 +0,0 @@
-"""
-Shared helpers for Human-In-The-Loop (HITL) review functionality.
-Used by both the dedicated HumanInTheLoopBlock and blocks that require human review.
-"""
-
-import logging
-from typing import Any, Optional
-
-from prisma.enums import ReviewStatus
-from pydantic import BaseModel
-
-from backend.data.execution import ExecutionContext, ExecutionStatus
-from backend.data.human_review import ReviewResult
-from backend.executor.manager import async_update_node_execution_status
-from backend.util.clients import get_database_manager_async_client
-
-logger = logging.getLogger(__name__)
-
-
-class ReviewDecision(BaseModel):
-    """Result of a review decision."""
-
-    should_proceed: bool
-    message: str
-    review_result: ReviewResult
-
-
-class HITLReviewHelper:
-    """Helper class for Human-In-The-Loop review operations."""
-
-    @staticmethod
-    async def get_or_create_human_review(**kwargs) -> Optional[ReviewResult]:
-        """Create or retrieve a human review from the database."""
-        return await get_database_manager_async_client().get_or_create_human_review(
-            **kwargs
-        )
-
-    @staticmethod
-    async def update_node_execution_status(**kwargs) -> None:
-        """Update the execution status of a node."""
-        await async_update_node_execution_status(
-            db_client=get_database_manager_async_client(), **kwargs
-        )
-
-    @staticmethod
-    async def update_review_processed_status(
-        node_exec_id: str, processed: bool
-    ) -> None:
-        """Update the processed status of a review."""
-        return await get_database_manager_async_client().update_review_processed_status(
-            node_exec_id, processed
-        )
-
-    @staticmethod
-    async def _handle_review_request(
-        input_data: Any,
-        user_id: str,
-        node_exec_id: str,
-        graph_exec_id: str,
-        graph_id: str,
-        graph_version: int,
-        execution_context: ExecutionContext,
-        block_name: str = "Block",
-        editable: bool = False,
-    ) -> Optional[ReviewResult]:
-        """
-        Handle a review request for a block that requires human review.
-
-        Args:
-            input_data: The input data to be reviewed
-            user_id: ID of the user requesting the review
-            node_exec_id: ID of the node execution
-            graph_exec_id: ID of the graph execution
-            graph_id: ID of the graph
-            graph_version: Version of the graph
-            execution_context: Current execution context
-            block_name: Name of the block requesting review
-            editable: Whether the reviewer can edit the data
-
-        Returns:
-            ReviewResult if review is complete, None if waiting for human input
-
-        Raises:
-            Exception: If review creation or status update fails
-        """
-        # Skip review if safe mode is disabled - return auto-approved result
-        if not execution_context.safe_mode:
-            logger.info(
-                f"Block {block_name} skipping review for node {node_exec_id} - safe mode disabled"
-            )
-            return ReviewResult(
-                data=input_data,
-                status=ReviewStatus.APPROVED,
-                message="Auto-approved (safe mode disabled)",
-                processed=True,
-                node_exec_id=node_exec_id,
-            )
-
-        result = await HITLReviewHelper.get_or_create_human_review(
-            user_id=user_id,
-            node_exec_id=node_exec_id,
-            graph_exec_id=graph_exec_id,
-            graph_id=graph_id,
-            graph_version=graph_version,
-            input_data=input_data,
-            message=f"Review required for {block_name} execution",
-            editable=editable,
-        )
-
-        if result is None:
-            logger.info(
-                f"Block {block_name} pausing execution for node {node_exec_id} - awaiting human review"
-            )
-            await HITLReviewHelper.update_node_execution_status(
-                exec_id=node_exec_id,
-                status=ExecutionStatus.REVIEW,
-            )
-            return None  # Signal that execution should pause
-
-        # Mark review as processed if not already done
-        if not result.processed:
-            await HITLReviewHelper.update_review_processed_status(
-                node_exec_id=node_exec_id, processed=True
-            )
-
-        return result
-
-    @staticmethod
-    async def handle_review_decision(
-        input_data: Any,
-        user_id: str,
-        node_exec_id: str,
-        graph_exec_id: str,
-        graph_id: str,
-        graph_version: int,
-        execution_context: ExecutionContext,
-        block_name: str = "Block",
-        editable: bool = False,
-    ) -> Optional[ReviewDecision]:
-        """
-        Handle a review request and return the decision in a single call.
-
-        Args:
-            input_data: The input data to be reviewed
-            user_id: ID of the user requesting the review
-            node_exec_id: ID of the node execution
-            graph_exec_id: ID of the graph execution
-            graph_id: ID of the graph
-            graph_version: Version of the graph
-            execution_context: Current execution context
-            block_name: Name of the block requesting review
-            editable: Whether the reviewer can edit the data
-
-        Returns:
-            ReviewDecision if review is complete (approved/rejected),
-            None if execution should pause (awaiting review)
-        """
-        review_result = await HITLReviewHelper._handle_review_request(
-            input_data=input_data,
-            user_id=user_id,
-            node_exec_id=node_exec_id,
-            graph_exec_id=graph_exec_id,
-            graph_id=graph_id,
-            graph_version=graph_version,
-            execution_context=execution_context,
-            block_name=block_name,
-            editable=editable,
-        )
-
-        if review_result is None:
-            # Still awaiting review - return None to pause execution
-            return None
-
-        # Review is complete, determine outcome
-        should_proceed = review_result.status == ReviewStatus.APPROVED
-        message = review_result.message or (
-            "Execution approved by reviewer"
-            if should_proceed
-            else "Execution rejected by reviewer"
-        )
-
-        return ReviewDecision(
-            should_proceed=should_proceed, message=message, review_result=review_result
-        )

autogpt_platform/backend/backend/blocks/human_in_the_loop.py

@@ -3,7 +3,6 @@ from typing import Any
 
 from prisma.enums import ReviewStatus
 
-from backend.blocks.helpers.review import HITLReviewHelper
 from backend.data.block import (
     Block,
     BlockCategory,
@@ -12,9 +11,11 @@ from backend.data.block import (
     BlockSchemaOutput,
     BlockType,
 )
-from backend.data.execution import ExecutionContext
+from backend.data.execution import ExecutionContext, ExecutionStatus
 from backend.data.human_review import ReviewResult
 from backend.data.model import SchemaField
+from backend.executor.manager import async_update_node_execution_status
+from backend.util.clients import get_database_manager_async_client
 
 logger = logging.getLogger(__name__)
 
@@ -71,26 +72,32 @@ class HumanInTheLoopBlock(Block):
                 ("approved_data", {"name": "John Doe", "age": 30}),
             ],
             test_mock={
-                "handle_review_decision": lambda **kwargs: type(
-                    "ReviewDecision",
-                    (),
-                    {
-                        "should_proceed": True,
-                        "message": "Test approval message",
-                        "review_result": ReviewResult(
-                            data={"name": "John Doe", "age": 30},
-                            status=ReviewStatus.APPROVED,
-                            message="",
-                            processed=False,
-                            node_exec_id="test-node-exec-id",
-                        ),
-                    },
-                )(),
+                "get_or_create_human_review": lambda *_args, **_kwargs: ReviewResult(
+                    data={"name": "John Doe", "age": 30},
+                    status=ReviewStatus.APPROVED,
+                    message="",
+                    processed=False,
+                    node_exec_id="test-node-exec-id",
+                ),
+                "update_node_execution_status": lambda *_args, **_kwargs: None,
+                "update_review_processed_status": lambda *_args, **_kwargs: None,
             },
         )
 
-    async def handle_review_decision(self, **kwargs):
-        return await HITLReviewHelper.handle_review_decision(**kwargs)
+    async def get_or_create_human_review(self, **kwargs):
+        return await get_database_manager_async_client().get_or_create_human_review(
+            **kwargs
+        )
+
+    async def update_node_execution_status(self, **kwargs):
+        return await async_update_node_execution_status(
+            db_client=get_database_manager_async_client(), **kwargs
+        )
+
+    async def update_review_processed_status(self, node_exec_id: str, processed: bool):
+        return await get_database_manager_async_client().update_review_processed_status(
+            node_exec_id, processed
+        )
 
     async def run(
         self,
@@ -102,7 +109,7 @@ class HumanInTheLoopBlock(Block):
         graph_id: str,
         graph_version: int,
         execution_context: ExecutionContext,
-        **_kwargs,
+        **kwargs,
     ) -> BlockOutput:
         if not execution_context.safe_mode:
             logger.info(
@@ -112,28 +119,48 @@ class HumanInTheLoopBlock(Block):
             yield "review_message", "Auto-approved (safe mode disabled)"
             return
 
-        decision = await self.handle_review_decision(
-            input_data=input_data.data,
-            user_id=user_id,
-            node_exec_id=node_exec_id,
-            graph_exec_id=graph_exec_id,
-            graph_id=graph_id,
-            graph_version=graph_version,
-            execution_context=execution_context,
-            block_name=self.name,
-            editable=input_data.editable,
-        )
+        try:
+            result = await self.get_or_create_human_review(
+                user_id=user_id,
+                node_exec_id=node_exec_id,
+                graph_exec_id=graph_exec_id,
+                graph_id=graph_id,
+                graph_version=graph_version,
+                input_data=input_data.data,
+                message=input_data.name,
+                editable=input_data.editable,
+            )
+        except Exception as e:
+            logger.error(f"Error in HITL block for node {node_exec_id}: {str(e)}")
+            raise
 
-        if decision is None:
-            return
+        if result is None:
+            logger.info(
+                f"HITL block pausing execution for node {node_exec_id} - awaiting human review"
+            )
+            try:
+                await self.update_node_execution_status(
+                    exec_id=node_exec_id,
+                    status=ExecutionStatus.REVIEW,
+                )
+                return
+            except Exception as e:
+                logger.error(
+                    f"Failed to update node status for HITL block {node_exec_id}: {str(e)}"
+                )
+                raise
 
-        status = decision.review_result.status
-        if status == ReviewStatus.APPROVED:
-            yield "approved_data", decision.review_result.data
-        elif status == ReviewStatus.REJECTED:
-            yield "rejected_data", decision.review_result.data
-        else:
-            raise RuntimeError(f"Unexpected review status: {status}")
+        if not result.processed:
+            await self.update_review_processed_status(
+                node_exec_id=node_exec_id, processed=True
+            )
 
-        if decision.message:
-            yield "review_message", decision.message
+            if result.status == ReviewStatus.APPROVED:
+                yield "approved_data", result.data
+                if result.message:
+                    yield "review_message", result.message
+
+            elif result.status == ReviewStatus.REJECTED:
+                yield "rejected_data", result.data
+                if result.message:
+                    yield "review_message", result.message

autogpt_platform/backend/backend/blocks/search.py

@@ -18,7 +18,6 @@ from backend.data.model import (
     SchemaField,
 )
 from backend.integrations.providers import ProviderName
-from backend.util.request import DEFAULT_USER_AGENT
 
 
 class GetWikipediaSummaryBlock(Block, GetRequest):
@@ -40,27 +39,17 @@ class GetWikipediaSummaryBlock(Block, GetRequest):
             output_schema=GetWikipediaSummaryBlock.Output,
             test_input={"topic": "Artificial Intelligence"},
             test_output=("summary", "summary content"),
-            test_mock={
-                "get_request": lambda url, headers, json: {"extract": "summary content"}
-            },
+            test_mock={"get_request": lambda url, json: {"extract": "summary content"}},
         )
 
     async def run(self, input_data: Input, **kwargs) -> BlockOutput:
         topic = input_data.topic
-        # URL-encode the topic to handle spaces and special characters
-        encoded_topic = quote(topic, safe="")
-        url = f"https://en.wikipedia.org/api/rest_v1/page/summary/{encoded_topic}"
-
-        # Set headers per Wikimedia robot policy (https://w.wiki/4wJS)
-        # - User-Agent: Required, must identify the bot
-        # - Accept-Encoding: gzip recommended to reduce bandwidth
-        headers = {
-            "User-Agent": DEFAULT_USER_AGENT,
-            "Accept-Encoding": "gzip, deflate",
-        }
+        url = f"https://en.wikipedia.org/api/rest_v1/page/summary/{topic}"
 
+        # Note: User-Agent is now automatically set by the request library
+        # to comply with Wikimedia's robot policy (https://w.wiki/4wJS)
         try:
-            response = await self.get_request(url, headers=headers, json=True)
+            response = await self.get_request(url, json=True)
             if "extract" not in response:
                 raise ValueError(f"Unable to parse Wikipedia response: {response}")
             yield "summary", response["extract"]

autogpt_platform/backend/backend/blocks/smart_decision_maker.py

@@ -391,12 +391,8 @@ class SmartDecisionMakerBlock(Block):
         """
         block = sink_node.block
 
-        # Use custom name from node metadata if set, otherwise fall back to block.name
-        custom_name = sink_node.metadata.get("customized_name")
-        tool_name = custom_name if custom_name else block.name
-
         tool_function: dict[str, Any] = {
-            "name": SmartDecisionMakerBlock.cleanup(tool_name),
+            "name": SmartDecisionMakerBlock.cleanup(block.name),
             "description": block.description,
         }
         sink_block_input_schema = block.input_schema
@@ -493,24 +489,14 @@ class SmartDecisionMakerBlock(Block):
                 f"Sink graph metadata not found: {graph_id} {graph_version}"
             )
 
-        # Use custom name from node metadata if set, otherwise fall back to graph name
-        custom_name = sink_node.metadata.get("customized_name")
-        tool_name = custom_name if custom_name else sink_graph_meta.name
-
         tool_function: dict[str, Any] = {
-            "name": SmartDecisionMakerBlock.cleanup(tool_name),
+            "name": SmartDecisionMakerBlock.cleanup(sink_graph_meta.name),
             "description": sink_graph_meta.description,
         }
 
         properties = {}
-        field_mapping = {}
 
         for link in links:
-            field_name = link.sink_name
-
-            clean_field_name = SmartDecisionMakerBlock.cleanup(field_name)
-            field_mapping[clean_field_name] = field_name
-
             sink_block_input_schema = sink_node.input_default["input_schema"]
             sink_block_properties = sink_block_input_schema.get("properties", {}).get(
                 link.sink_name, {}
@@ -520,7 +506,7 @@ class SmartDecisionMakerBlock(Block):
                 if "description" in sink_block_properties
                 else f"The {link.sink_name} of the tool"
             )
-            properties[clean_field_name] = {
+            properties[link.sink_name] = {
                 "type": "string",
                 "description": description,
                 "default": json.dumps(sink_block_properties.get("default", None)),
@@ -533,7 +519,7 @@ class SmartDecisionMakerBlock(Block):
             "strict": True,
         }
 
-        tool_function["_field_mapping"] = field_mapping
+        # Store node info for later use in output processing
         tool_function["_sink_node_id"] = sink_node.id
 
         return {"type": "function", "function": tool_function}
@@ -989,28 +975,10 @@ class SmartDecisionMakerBlock(Block):
         graph_version: int,
         execution_context: ExecutionContext,
         execution_processor: "ExecutionProcessor",
-        nodes_to_skip: set[str] | None = None,
         **kwargs,
     ) -> BlockOutput:
 
         tool_functions = await self._create_tool_node_signatures(node_id)
-        original_tool_count = len(tool_functions)
-
-        # Filter out tools for nodes that should be skipped (e.g., missing optional credentials)
-        if nodes_to_skip:
-            tool_functions = [
-                tf
-                for tf in tool_functions
-                if tf.get("function", {}).get("_sink_node_id") not in nodes_to_skip
-            ]
-
-            # Only raise error if we had tools but they were all filtered out
-            if original_tool_count > 0 and not tool_functions:
-                raise ValueError(
-                    "No available tools to execute - all downstream nodes are unavailable "
-                    "(possibly due to missing optional credentials)"
-                )
-
         yield "tool_functions", json.dumps(tool_functions)
 
         conversation_history = input_data.conversation_history or []
@@ -1161,9 +1129,8 @@ class SmartDecisionMakerBlock(Block):
                 original_field_name = field_mapping.get(clean_arg_name, clean_arg_name)
                 arg_value = tool_args.get(clean_arg_name)
 
-                # Use original_field_name directly (not sanitized) to match link sink_name
-                # The field_mapping already translates from LLM's cleaned names to original names
-                emit_key = f"tools_^_{sink_node_id}_~_{original_field_name}"
+                sanitized_arg_name = self.cleanup(original_field_name)
+                emit_key = f"tools_^_{sink_node_id}_~_{sanitized_arg_name}"
 
                 logger.debug(
                     "[SmartDecisionMakerBlock|geid:%s|neid:%s] emit %s",

autogpt_platform/backend/backend/blocks/test/test_smart_decision_maker.py

@@ -1057,153 +1057,3 @@ async def test_smart_decision_maker_traditional_mode_default():
         )  # Should yield individual tool parameters
         assert "tools_^_test-sink-node-id_~_max_keyword_difficulty" in outputs
         assert "conversations" in outputs
-
-
-@pytest.mark.asyncio
-async def test_smart_decision_maker_uses_customized_name_for_blocks():
-    """Test that SmartDecisionMakerBlock uses customized_name from node metadata for tool names."""
-    from unittest.mock import MagicMock
-
-    from backend.blocks.basic import StoreValueBlock
-    from backend.blocks.smart_decision_maker import SmartDecisionMakerBlock
-    from backend.data.graph import Link, Node
-
-    # Create a mock node with customized_name in metadata
-    mock_node = MagicMock(spec=Node)
-    mock_node.id = "test-node-id"
-    mock_node.block_id = StoreValueBlock().id
-    mock_node.metadata = {"customized_name": "My Custom Tool Name"}
-    mock_node.block = StoreValueBlock()
-
-    # Create a mock link
-    mock_link = MagicMock(spec=Link)
-    mock_link.sink_name = "input"
-
-    # Call the function directly
-    result = await SmartDecisionMakerBlock._create_block_function_signature(
-        mock_node, [mock_link]
-    )
-
-    # Verify the tool name uses the customized name (cleaned up)
-    assert result["type"] == "function"
-    assert result["function"]["name"] == "my_custom_tool_name"  # Cleaned version
-    assert result["function"]["_sink_node_id"] == "test-node-id"
-
-
-@pytest.mark.asyncio
-async def test_smart_decision_maker_falls_back_to_block_name():
-    """Test that SmartDecisionMakerBlock falls back to block.name when no customized_name."""
-    from unittest.mock import MagicMock
-
-    from backend.blocks.basic import StoreValueBlock
-    from backend.blocks.smart_decision_maker import SmartDecisionMakerBlock
-    from backend.data.graph import Link, Node
-
-    # Create a mock node without customized_name
-    mock_node = MagicMock(spec=Node)
-    mock_node.id = "test-node-id"
-    mock_node.block_id = StoreValueBlock().id
-    mock_node.metadata = {}  # No customized_name
-    mock_node.block = StoreValueBlock()
-
-    # Create a mock link
-    mock_link = MagicMock(spec=Link)
-    mock_link.sink_name = "input"
-
-    # Call the function directly
-    result = await SmartDecisionMakerBlock._create_block_function_signature(
-        mock_node, [mock_link]
-    )
-
-    # Verify the tool name uses the block's default name
-    assert result["type"] == "function"
-    assert result["function"]["name"] == "storevalueblock"  # Default block name cleaned
-    assert result["function"]["_sink_node_id"] == "test-node-id"
-
-
-@pytest.mark.asyncio
-async def test_smart_decision_maker_uses_customized_name_for_agents():
-    """Test that SmartDecisionMakerBlock uses customized_name from metadata for agent nodes."""
-    from unittest.mock import AsyncMock, MagicMock, patch
-
-    from backend.blocks.smart_decision_maker import SmartDecisionMakerBlock
-    from backend.data.graph import Link, Node
-
-    # Create a mock node with customized_name in metadata
-    mock_node = MagicMock(spec=Node)
-    mock_node.id = "test-agent-node-id"
-    mock_node.metadata = {"customized_name": "My Custom Agent"}
-    mock_node.input_default = {
-        "graph_id": "test-graph-id",
-        "graph_version": 1,
-        "input_schema": {"properties": {"test_input": {"description": "Test input"}}},
-    }
-
-    # Create a mock link
-    mock_link = MagicMock(spec=Link)
-    mock_link.sink_name = "test_input"
-
-    # Mock the database client
-    mock_graph_meta = MagicMock()
-    mock_graph_meta.name = "Original Agent Name"
-    mock_graph_meta.description = "Agent description"
-
-    mock_db_client = AsyncMock()
-    mock_db_client.get_graph_metadata.return_value = mock_graph_meta
-
-    with patch(
-        "backend.blocks.smart_decision_maker.get_database_manager_async_client",
-        return_value=mock_db_client,
-    ):
-        result = await SmartDecisionMakerBlock._create_agent_function_signature(
-            mock_node, [mock_link]
-        )
-
-    # Verify the tool name uses the customized name (cleaned up)
-    assert result["type"] == "function"
-    assert result["function"]["name"] == "my_custom_agent"  # Cleaned version
-    assert result["function"]["_sink_node_id"] == "test-agent-node-id"
-
-
-@pytest.mark.asyncio
-async def test_smart_decision_maker_agent_falls_back_to_graph_name():
-    """Test that agent node falls back to graph name when no customized_name."""
-    from unittest.mock import AsyncMock, MagicMock, patch
-
-    from backend.blocks.smart_decision_maker import SmartDecisionMakerBlock
-    from backend.data.graph import Link, Node
-
-    # Create a mock node without customized_name
-    mock_node = MagicMock(spec=Node)
-    mock_node.id = "test-agent-node-id"
-    mock_node.metadata = {}  # No customized_name
-    mock_node.input_default = {
-        "graph_id": "test-graph-id",
-        "graph_version": 1,
-        "input_schema": {"properties": {"test_input": {"description": "Test input"}}},
-    }
-
-    # Create a mock link
-    mock_link = MagicMock(spec=Link)
-    mock_link.sink_name = "test_input"
-
-    # Mock the database client
-    mock_graph_meta = MagicMock()
-    mock_graph_meta.name = "Original Agent Name"
-    mock_graph_meta.description = "Agent description"
-
-    mock_db_client = AsyncMock()
-    mock_db_client.get_graph_metadata.return_value = mock_graph_meta
-
-    with patch(
-        "backend.blocks.smart_decision_maker.get_database_manager_async_client",
-        return_value=mock_db_client,
-    ):
-        result = await SmartDecisionMakerBlock._create_agent_function_signature(
-            mock_node, [mock_link]
-        )
-
-    # Verify the tool name uses the graph's default name
-    assert result["type"] == "function"
-    assert result["function"]["name"] == "original_agent_name"  # Graph name cleaned
-    assert result["function"]["_sink_node_id"] == "test-agent-node-id"

autogpt_platform/backend/backend/blocks/test/test_smart_decision_maker_dict.py

@@ -15,7 +15,6 @@ async def test_smart_decision_maker_handles_dynamic_dict_fields():
     mock_node.block = CreateDictionaryBlock()
     mock_node.block_id = CreateDictionaryBlock().id
     mock_node.input_default = {}
-    mock_node.metadata = {}
 
     # Create mock links with dynamic dictionary fields
     mock_links = [
@@ -78,7 +77,6 @@ async def test_smart_decision_maker_handles_dynamic_list_fields():
     mock_node.block = AddToListBlock()
     mock_node.block_id = AddToListBlock().id
     mock_node.input_default = {}
-    mock_node.metadata = {}
 
     # Create mock links with dynamic list fields
     mock_links = [

autogpt_platform/backend/backend/blocks/test/test_smart_decision_maker_dynamic_fields.py

@@ -44,7 +44,6 @@ async def test_create_block_function_signature_with_dict_fields():
     mock_node.block = CreateDictionaryBlock()
     mock_node.block_id = CreateDictionaryBlock().id
     mock_node.input_default = {}
-    mock_node.metadata = {}
 
     # Create mock links with dynamic dictionary fields (source sanitized, sink original)
     mock_links = [
@@ -107,7 +106,6 @@ async def test_create_block_function_signature_with_list_fields():
     mock_node.block = AddToListBlock()
     mock_node.block_id = AddToListBlock().id
     mock_node.input_default = {}
-    mock_node.metadata = {}
 
     # Create mock links with dynamic list fields
     mock_links = [
@@ -161,7 +159,6 @@ async def test_create_block_function_signature_with_object_fields():
     mock_node.block = MatchTextPatternBlock()
     mock_node.block_id = MatchTextPatternBlock().id
     mock_node.input_default = {}
-    mock_node.metadata = {}
 
     # Create mock links with dynamic object fields
     mock_links = [
@@ -211,13 +208,11 @@ async def test_create_tool_node_signatures():
         mock_dict_node.block = CreateDictionaryBlock()
         mock_dict_node.block_id = CreateDictionaryBlock().id
         mock_dict_node.input_default = {}
-        mock_dict_node.metadata = {}
 
         mock_list_node = Mock()
         mock_list_node.block = AddToListBlock()
         mock_list_node.block_id = AddToListBlock().id
         mock_list_node.input_default = {}
-        mock_list_node.metadata = {}
 
         # Mock links with dynamic fields
         dict_link1 = Mock(
@@ -428,7 +423,6 @@ async def test_mixed_regular_and_dynamic_fields():
     mock_node.block.name = "TestBlock"
     mock_node.block.description = "A test block"
     mock_node.block.input_schema = Mock()
-    mock_node.metadata = {}
 
     # Mock the get_field_schema to return a proper schema for regular fields
     def get_field_schema(field_name):

autogpt_platform/backend/backend/blocks/wordpress/__init__.py

@@ -1,3 +1,3 @@
-from .blog import WordPressCreatePostBlock, WordPressGetAllPostsBlock
+from .blog import WordPressCreatePostBlock
 
-__all__ = ["WordPressCreatePostBlock", "WordPressGetAllPostsBlock"]
+__all__ = ["WordPressCreatePostBlock"]

autogpt_platform/backend/backend/blocks/wordpress/_api.py

@@ -161,7 +161,7 @@ async def oauth_exchange_code_for_tokens(
         grant_type="authorization_code",
     ).model_dump(exclude_none=True)
 
-    response = await Requests(raise_for_status=False).post(
+    response = await Requests().post(
         f"{WORDPRESS_BASE_URL}oauth2/token",
         headers=headers,
         data=data,
@@ -205,7 +205,7 @@ async def oauth_refresh_tokens(
         grant_type="refresh_token",
     ).model_dump(exclude_none=True)
 
-    response = await Requests(raise_for_status=False).post(
+    response = await Requests().post(
         f"{WORDPRESS_BASE_URL}oauth2/token",
         headers=headers,
         data=data,
@@ -252,7 +252,7 @@ async def validate_token(
         "token": token,
     }
 
-    response = await Requests(raise_for_status=False).get(
+    response = await Requests().get(
         f"{WORDPRESS_BASE_URL}oauth2/token-info",
         params=params,
     )
@@ -296,7 +296,7 @@ async def make_api_request(
 
     url = f"{WORDPRESS_BASE_URL.rstrip('/')}{endpoint}"
 
-    request_method = getattr(Requests(raise_for_status=False), method.lower())
+    request_method = getattr(Requests(), method.lower())
     response = await request_method(
         url,
         headers=headers,
@@ -476,7 +476,6 @@ async def create_post(
         data["tags"] = ",".join(str(t) for t in data["tags"])
 
     # Make the API request
-    site = normalize_site(site)
     endpoint = f"/rest/v1.1/sites/{site}/posts/new"
 
     headers = {
@@ -484,7 +483,7 @@ async def create_post(
         "Content-Type": "application/x-www-form-urlencoded",
     }
 
-    response = await Requests(raise_for_status=False).post(
+    response = await Requests().post(
         f"{WORDPRESS_BASE_URL.rstrip('/')}{endpoint}",
         headers=headers,
         data=data,
@@ -500,132 +499,3 @@ async def create_post(
     )
     error_message = error_data.get("message", response.text)
     raise ValueError(f"Failed to create post: {response.status} - {error_message}")
-
-
-class Post(BaseModel):
-    """Response model for individual posts in a posts list response.
-
-    This is a simplified version compared to PostResponse, as the list endpoint
-    returns less detailed information than the create/get single post endpoints.
-    """
-
-    ID: int
-    site_ID: int
-    author: PostAuthor
-    date: datetime
-    modified: datetime
-    title: str
-    URL: str
-    short_URL: str
-    content: str | None = None
-    excerpt: str | None = None
-    slug: str
-    guid: str
-    status: str
-    sticky: bool
-    password: str | None = ""
-    parent: Union[Dict[str, Any], bool, None] = None
-    type: str
-    discussion: Dict[str, Union[str, bool, int]] | None = None
-    likes_enabled: bool | None = None
-    sharing_enabled: bool | None = None
-    like_count: int | None = None
-    i_like: bool | None = None
-    is_reblogged: bool | None = None
-    is_following: bool | None = None
-    global_ID: str | None = None
-    featured_image: str | None = None
-    post_thumbnail: Dict[str, Any] | None = None
-    format: str | None = None
-    geo: Union[Dict[str, Any], bool, None] = None
-    menu_order: int | None = None
-    page_template: str | None = None
-    publicize_URLs: List[str] | None = None
-    terms: Dict[str, Dict[str, Any]] | None = None
-    tags: Dict[str, Dict[str, Any]] | None = None
-    categories: Dict[str, Dict[str, Any]] | None = None
-    attachments: Dict[str, Dict[str, Any]] | None = None
-    attachment_count: int | None = None
-    metadata: List[Dict[str, Any]] | None = None
-    meta: Dict[str, Any] | None = None
-    capabilities: Dict[str, bool] | None = None
-    revisions: List[int] | None = None
-    other_URLs: Dict[str, Any] | None = None
-
-
-class PostsResponse(BaseModel):
-    """Response model for WordPress posts list."""
-
-    found: int
-    posts: List[Post]
-    meta: Dict[str, Any]
-
-
-def normalize_site(site: str) -> str:
-    """
-    Normalize a site identifier by stripping protocol and trailing slashes.
-
-    Args:
-        site: Site URL, domain, or ID (e.g., "https://myblog.wordpress.com/", "myblog.wordpress.com", "123456789")
-
-    Returns:
-        Normalized site identifier (domain or ID only)
-    """
-    site = site.strip()
-    if site.startswith("https://"):
-        site = site[8:]
-    elif site.startswith("http://"):
-        site = site[7:]
-    return site.rstrip("/")
-
-
-async def get_posts(
-    credentials: Credentials,
-    site: str,
-    status: PostStatus | None = None,
-    number: int = 100,
-    offset: int = 0,
-) -> PostsResponse:
-    """
-    Get posts from a WordPress site.
-
-    Args:
-        credentials: OAuth credentials
-        site: Site ID or domain (e.g., "myblog.wordpress.com" or "123456789")
-        status: Filter by post status using PostStatus enum, or None for all
-        number: Number of posts to retrieve (max 100)
-        offset: Number of posts to skip (for pagination)
-
-    Returns:
-        PostsResponse with the list of posts
-    """
-    site = normalize_site(site)
-    endpoint = f"/rest/v1.1/sites/{site}/posts"
-
-    headers = {
-        "Authorization": credentials.auth_header(),
-    }
-
-    params: Dict[str, Any] = {
-        "number": max(1, min(number, 100)),  # 1–100 posts per request
-        "offset": offset,
-    }
-
-    if status:
-        params["status"] = status.value
-    response = await Requests(raise_for_status=False).get(
-        f"{WORDPRESS_BASE_URL.rstrip('/')}{endpoint}",
-        headers=headers,
-        params=params,
-    )
-
-    if response.ok:
-        return PostsResponse.model_validate(response.json())
-
-    error_data = (
-        response.json()
-        if response.headers.get("content-type", "").startswith("application/json")
-        else {}
-    )
-    error_message = error_data.get("message", response.text)
-    raise ValueError(f"Failed to get posts: {response.status} - {error_message}")

autogpt_platform/backend/backend/blocks/wordpress/blog.py

@@ -9,15 +9,7 @@ from backend.sdk import (
     SchemaField,
 )
 
-from ._api import (
-    CreatePostRequest,
-    Post,
-    PostResponse,
-    PostsResponse,
-    PostStatus,
-    create_post,
-    get_posts,
-)
+from ._api import CreatePostRequest, PostResponse, PostStatus, create_post
 from ._config import wordpress
 
 
@@ -57,15 +49,8 @@ class WordPressCreatePostBlock(Block):
         media_urls: list[str] = SchemaField(
             description="URLs of images to sideload and attach to the post", default=[]
         )
-        publish_as_draft: bool = SchemaField(
-            description="If True, publishes the post as a draft. If False, publishes it publicly.",
-            default=False,
-        )
 
     class Output(BlockSchemaOutput):
-        site: str = SchemaField(
-            description="The site ID or domain (pass-through for chaining with other blocks)"
-        )
         post_id: int = SchemaField(description="The ID of the created post")
         post_url: str = SchemaField(description="The full URL of the created post")
         short_url: str = SchemaField(description="The shortened wp.me URL")
@@ -93,9 +78,7 @@ class WordPressCreatePostBlock(Block):
             tags=input_data.tags,
             featured_image=input_data.featured_image,
             media_urls=input_data.media_urls,
-            status=(
-                PostStatus.DRAFT if input_data.publish_as_draft else PostStatus.PUBLISH
-            ),
+            status=PostStatus.PUBLISH,
         )
 
         post_response: PostResponse = await create_post(
@@ -104,69 +87,7 @@ class WordPressCreatePostBlock(Block):
             post_data=post_request,
         )
 
-        yield "site", input_data.site
         yield "post_id", post_response.ID
         yield "post_url", post_response.URL
         yield "short_url", post_response.short_URL
         yield "post_data", post_response.model_dump()
-
-
-class WordPressGetAllPostsBlock(Block):
-    """
-    Fetches all posts from a WordPress.com site or Jetpack-enabled site.
-    Supports filtering by status and pagination.
-    """
-
-    class Input(BlockSchemaInput):
-        credentials: CredentialsMetaInput = wordpress.credentials_field()
-        site: str = SchemaField(
-            description="Site ID or domain (e.g., 'myblog.wordpress.com' or '123456789')"
-        )
-        status: PostStatus | None = SchemaField(
-            description="Filter by post status, or None for all",
-            default=None,
-        )
-        number: int = SchemaField(
-            description="Number of posts to retrieve (max 100 per request)", default=20
-        )
-        offset: int = SchemaField(
-            description="Number of posts to skip (for pagination)", default=0
-        )
-
-    class Output(BlockSchemaOutput):
-        site: str = SchemaField(
-            description="The site ID or domain (pass-through for chaining with other blocks)"
-        )
-        found: int = SchemaField(description="Total number of posts found")
-        posts: list[Post] = SchemaField(
-            description="List of post objects with their details"
-        )
-        post: Post = SchemaField(
-            description="Individual post object (yielded for each post)"
-        )
-
-    def __init__(self):
-        super().__init__(
-            id="97728fa7-7f6f-4789-ba0c-f2c114119536",
-            description="Fetch all posts from WordPress.com or Jetpack sites",
-            categories={BlockCategory.SOCIAL},
-            input_schema=self.Input,
-            output_schema=self.Output,
-        )
-
-    async def run(
-        self, input_data: Input, *, credentials: Credentials, **kwargs
-    ) -> BlockOutput:
-        posts_response: PostsResponse = await get_posts(
-            credentials=credentials,
-            site=input_data.site,
-            status=input_data.status,
-            number=input_data.number,
-            offset=input_data.offset,
-        )
-
-        yield "site", input_data.site
-        yield "found", posts_response.found
-        yield "posts", posts_response.posts
-        for post in posts_response.posts:
-            yield "post", post

autogpt_platform/backend/backend/data/block.py

@@ -50,8 +50,6 @@ from .model import (
 logger = logging.getLogger(__name__)
 
 if TYPE_CHECKING:
-    from backend.data.execution import ExecutionContext
-
     from .graph import Link
 
 app_config = Config()
@@ -474,7 +472,6 @@ class Block(ABC, Generic[BlockSchemaInputType, BlockSchemaOutputType]):
         self.block_type = block_type
         self.webhook_config = webhook_config
         self.execution_stats: NodeExecutionStats = NodeExecutionStats()
-        self.requires_human_review: bool = False
 
         if self.webhook_config:
             if isinstance(self.webhook_config, BlockWebhookConfig):
@@ -617,77 +614,7 @@ class Block(ABC, Generic[BlockSchemaInputType, BlockSchemaOutputType]):
                     block_id=self.id,
                 ) from ex
 
-    async def is_block_exec_need_review(
-        self,
-        input_data: BlockInput,
-        *,
-        user_id: str,
-        node_exec_id: str,
-        graph_exec_id: str,
-        graph_id: str,
-        graph_version: int,
-        execution_context: "ExecutionContext",
-        **kwargs,
-    ) -> tuple[bool, BlockInput]:
-        """
-        Check if this block execution needs human review and handle the review process.
-
-        Returns:
-            Tuple of (should_pause, input_data_to_use)
-            - should_pause: True if execution should be paused for review
-            - input_data_to_use: The input data to use (may be modified by reviewer)
-        """
-        # Skip review if not required or safe mode is disabled
-        if not self.requires_human_review or not execution_context.safe_mode:
-            return False, input_data
-
-        from backend.blocks.helpers.review import HITLReviewHelper
-
-        # Handle the review request and get decision
-        decision = await HITLReviewHelper.handle_review_decision(
-            input_data=input_data,
-            user_id=user_id,
-            node_exec_id=node_exec_id,
-            graph_exec_id=graph_exec_id,
-            graph_id=graph_id,
-            graph_version=graph_version,
-            execution_context=execution_context,
-            block_name=self.name,
-            editable=True,
-        )
-
-        if decision is None:
-            # We're awaiting review - pause execution
-            return True, input_data
-
-        if not decision.should_proceed:
-            # Review was rejected, raise an error to stop execution
-            raise BlockExecutionError(
-                message=f"Block execution rejected by reviewer: {decision.message}",
-                block_name=self.name,
-                block_id=self.id,
-            )
-
-        # Review was approved - use the potentially modified data
-        # ReviewResult.data must be a dict for block inputs
-        reviewed_data = decision.review_result.data
-        if not isinstance(reviewed_data, dict):
-            raise BlockExecutionError(
-                message=f"Review data must be a dict for block input, got {type(reviewed_data).__name__}",
-                block_name=self.name,
-                block_id=self.id,
-            )
-        return False, reviewed_data
-
     async def _execute(self, input_data: BlockInput, **kwargs) -> BlockOutput:
-        # Check for review requirement and get potentially modified input data
-        should_pause, input_data = await self.is_block_exec_need_review(
-            input_data, **kwargs
-        )
-        if should_pause:
-            return
-
-        # Validate the input data (original or reviewer-modified) once
         if error := self.input_schema.validate_data(input_data):
             raise BlockInputError(
                 message=f"Unable to execute block with invalid input data: {error}",
@@ -695,7 +622,6 @@ class Block(ABC, Generic[BlockSchemaInputType, BlockSchemaOutputType]):
                 block_id=self.id,
             )
 
-        # Use the validated input data
         async for output_name, output_data in self.run(
             self.input_schema(**{k: v for k, v in input_data.items() if v is not None}),
             **kwargs,