diff --git a/.branchlet.json b/.branchlet.json index cc13ff9f74..d02cd60e20 100644 --- a/.branchlet.json +++ b/.branchlet.json @@ -29,8 +29,7 @@ "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" + "cd autogpt_platform/frontend && pnpm install" ], "terminalCommand": "code .", "deleteBranchWithWorktree": false diff --git a/.github/copilot-instructions.md b/.github/copilot-instructions.md index 870e6b4b0a..3c72eaae18 100644 --- a/.github/copilot-instructions.md +++ b/.github/copilot-instructions.md @@ -160,7 +160,7 @@ pnpm storybook # Start component development server **Backend Entry Points:** -- `backend/backend/server/server.py` - FastAPI application setup +- `backend/backend/api/rest_api.py` - FastAPI application setup - `backend/backend/data/` - Database models and user management - `backend/blocks/` - Agent execution blocks and logic @@ -219,7 +219,7 @@ Agents are built using a visual block-based system where each block performs a s ### API Development -1. Update routes in `/backend/backend/server/routers/` +1. Update routes in `/backend/backend/api/features/` 2. Add/update Pydantic models in same directory 3. Write tests alongside route files 4. For `data/*.py` changes, validate user ID checks @@ -285,7 +285,7 @@ Agents are built using a visual block-based system where each block performs a s ### Security Guidelines -**Cache Protection Middleware** (`/backend/backend/server/middleware/security.py`): +**Cache Protection Middleware** (`/backend/backend/api/middleware/security.py`): - Default: Disables caching for ALL endpoints with `Cache-Control: no-store, no-cache, must-revalidate, private` - Uses allow list approach for cacheable paths (static assets, health checks, public pages) diff --git a/.gitignore b/.gitignore index dfce8ba810..1a2291b516 100644 --- a/.gitignore +++ b/.gitignore @@ -178,4 +178,5 @@ autogpt_platform/backend/settings.py *.ign.* .test-contents .claude/settings.local.json +CLAUDE.local.md /autogpt_platform/backend/logs diff --git a/AGENTS.md b/AGENTS.md index cd176f8a2d..202c4c6e02 100644 --- a/AGENTS.md +++ b/AGENTS.md @@ -16,7 +16,6 @@ See `docs/content/platform/getting-started.md` for setup instructions. - Format Python code with `poetry run format`. - Format frontend code using `pnpm format`. - ## Frontend guidelines: See `/frontend/CONTRIBUTING.md` for complete patterns. Quick reference: @@ -33,14 +32,17 @@ See `/frontend/CONTRIBUTING.md` for complete patterns. Quick reference: 4. **Styling**: Tailwind CSS only, use design tokens, Phosphor Icons only 5. **Testing**: Add Storybook stories for new components, Playwright for E2E 6. **Code conventions**: Function declarations (not arrow functions) for components/handlers + - Component props should be `interface Props { ... }` (not exported) unless the interface needs to be used outside the component - Separate render logic from business logic (component.tsx + useComponent.ts + helpers.ts) - Colocate state when possible and avoid creating large components, use sub-components ( local `/components` folder next to the parent component ) when sensible - Avoid large hooks, abstract logic into `helpers.ts` files when sensible - Use function declarations for components, arrow functions only for callbacks - No barrel files or `index.ts` re-exports -- Do not use `useCallback` or `useMemo` unless strictly needed - Avoid comments at all times unless the code is very complex +- Do not use `useCallback` or `useMemo` unless asked to optimise a given function +- Do not type hook returns, let Typescript infer as much as possible +- Never type with `any`, if not types available use `unknown` ## Testing @@ -49,22 +51,8 @@ See `/frontend/CONTRIBUTING.md` for complete patterns. Quick reference: Always run the relevant linters and tests before committing. Use conventional commit messages for all commits (e.g. `feat(backend): add API`). - Types: - - feat - - fix - - refactor - - ci - - dx (developer experience) - Scopes: - - platform - - platform/library - - platform/marketplace - - backend - - backend/executor - - frontend - - frontend/library - - frontend/marketplace - - blocks +Types: - feat - fix - refactor - ci - dx (developer experience) +Scopes: - platform - platform/library - platform/marketplace - backend - backend/executor - frontend - frontend/library - frontend/marketplace - blocks ## Pull requests diff --git a/README.md b/README.md index 3572fe318b..349d8818ef 100644 --- a/README.md +++ b/README.md @@ -54,7 +54,7 @@ Before proceeding with the installation, ensure your system meets the following ### Updated Setup Instructions: We've moved to a fully maintained and regularly updated documentation site. -👉 [Follow the official self-hosting guide here](https://docs.agpt.co/platform/getting-started/) +👉 [Follow the official self-hosting guide here](https://agpt.co/docs/platform/getting-started/getting-started) This tutorial assumes you have Docker, VSCode, git and npm installed. diff --git a/autogpt_platform/CLAUDE.md b/autogpt_platform/CLAUDE.md index 9690178587..62adbdaefa 100644 --- a/autogpt_platform/CLAUDE.md +++ b/autogpt_platform/CLAUDE.md @@ -6,152 +6,30 @@ This file provides guidance to Claude Code (claude.ai/code) when working with co AutoGPT Platform is a monorepo containing: -- **Backend** (`/backend`): Python FastAPI server with async support -- **Frontend** (`/frontend`): Next.js React application -- **Shared Libraries** (`/autogpt_libs`): Common Python utilities +- **Backend** (`backend`): Python FastAPI server with async support +- **Frontend** (`frontend`): Next.js React application +- **Shared Libraries** (`autogpt_libs`): Common Python utilities -## Essential Commands +## Component Documentation -### Backend Development +- **Backend**: See @backend/CLAUDE.md for backend-specific commands, architecture, and development tasks +- **Frontend**: See @frontend/CLAUDE.md for frontend-specific commands, architecture, and development patterns -```bash -# Install dependencies -cd backend && poetry install - -# Run database migrations -poetry run prisma migrate dev - -# Start all services (database, redis, rabbitmq, clamav) -docker compose up -d - -# Run the backend server -poetry run serve - -# Run tests -poetry run test - -# Run specific test -poetry run pytest path/to/test_file.py::test_function_name - -# Run block tests (tests that validate all blocks work correctly) -poetry run pytest backend/blocks/test/test_block.py -xvs - -# Run tests for a specific block (e.g., GetCurrentTimeBlock) -poetry run pytest 'backend/blocks/test/test_block.py::test_available_blocks[GetCurrentTimeBlock]' -xvs - -# Lint and format -# prefer format if you want to just "fix" it and only get the errors that can't be autofixed -poetry run format # Black + isort -poetry run lint # ruff -``` - -More details can be found in TESTING.md - -#### Creating/Updating Snapshots - -When you first write a test or when the expected output changes: - -```bash -poetry run pytest path/to/test.py --snapshot-update -``` - -⚠️ **Important**: Always review snapshot changes before committing! Use `git diff` to verify the changes are expected. - -### Frontend Development - -```bash -# Install dependencies -cd frontend && pnpm i - -# Generate API client from OpenAPI spec -pnpm generate:api - -# Start development server -pnpm dev - -# Run E2E tests -pnpm test - -# Run Storybook for component development -pnpm storybook - -# Build production -pnpm build - -# Format and lint -pnpm format - -# Type checking -pnpm types -``` - -**📖 Complete Guide**: See `/frontend/CONTRIBUTING.md` and `/frontend/.cursorrules` for comprehensive frontend patterns. - -**Key Frontend Conventions:** - -- Separate render logic from data/behavior in components -- Use generated API hooks from `@/app/api/__generated__/endpoints/` -- Use function declarations (not arrow functions) for components/handlers -- Use design system components from `src/components/` (atoms, molecules, organisms) -- Only use Phosphor Icons -- Never use `src/components/__legacy__/*` or deprecated `BackendAPI` - -## Architecture Overview - -### Backend Architecture - -- **API Layer**: FastAPI with REST and WebSocket endpoints -- **Database**: PostgreSQL with Prisma ORM, includes pgvector for embeddings -- **Queue System**: RabbitMQ for async task processing -- **Execution Engine**: Separate executor service processes agent workflows -- **Authentication**: JWT-based with Supabase integration -- **Security**: Cache protection middleware prevents sensitive data caching in browsers/proxies - -### Frontend Architecture - -- **Framework**: Next.js 15 App Router (client-first approach) -- **Data Fetching**: Type-safe generated API hooks via Orval + React Query -- **State Management**: React Query for server state, co-located UI state in components/hooks -- **Component Structure**: Separate render logic (`.tsx`) from business logic (`use*.ts` hooks) -- **Workflow Builder**: Visual graph editor using @xyflow/react -- **UI Components**: shadcn/ui (Radix UI primitives) with Tailwind CSS styling -- **Icons**: Phosphor Icons only -- **Feature Flags**: LaunchDarkly integration -- **Error Handling**: ErrorCard for render errors, toast for mutations, Sentry for exceptions -- **Testing**: Playwright for E2E, Storybook for component development - -### Key Concepts +## Key Concepts 1. **Agent Graphs**: Workflow definitions stored as JSON, executed by the backend -2. **Blocks**: Reusable components in `/backend/blocks/` that perform specific tasks +2. **Blocks**: Reusable components in `backend/backend/blocks/` that perform specific tasks 3. **Integrations**: OAuth and API connections stored per user 4. **Store**: Marketplace for sharing agent templates 5. **Virus Scanning**: ClamAV integration for file upload security -### Testing Approach - -- Backend uses pytest with snapshot testing for API responses -- Test files are colocated with source files (`*_test.py`) -- Frontend uses Playwright for E2E tests -- Component testing via Storybook - -### Database Schema - -Key models (defined in `/backend/schema.prisma`): - -- `User`: Authentication and profile data -- `AgentGraph`: Workflow definitions with version control -- `AgentGraphExecution`: Execution history and results -- `AgentNode`: Individual nodes in a workflow -- `StoreListing`: Marketplace listings for sharing agents - ### Environment Configuration #### Configuration Files -- **Backend**: `/backend/.env.default` (defaults) → `/backend/.env` (user overrides) -- **Frontend**: `/frontend/.env.default` (defaults) → `/frontend/.env` (user overrides) -- **Platform**: `/.env.default` (Supabase/shared defaults) → `/.env` (user overrides) +- **Backend**: `backend/.env.default` (defaults) → `backend/.env` (user overrides) +- **Frontend**: `frontend/.env.default` (defaults) → `frontend/.env` (user overrides) +- **Platform**: `.env.default` (Supabase/shared defaults) → `.env` (user overrides) #### Docker Environment Loading Order @@ -167,127 +45,12 @@ Key models (defined in `/backend/schema.prisma`): - Backend/Frontend services use YAML anchors for consistent configuration - Supabase services (`db/docker/docker-compose.yml`) follow the same pattern -### Common Development Tasks - -**Adding a new block:** - -Follow the comprehensive [Block SDK Guide](../../../docs/content/platform/block-sdk-guide.md) which covers: - -- Provider configuration with `ProviderBuilder` -- Block schema definition -- Authentication (API keys, OAuth, webhooks) -- Testing and validation -- File organization - -Quick steps: - -1. Create new file in `/backend/backend/blocks/` -2. Configure provider using `ProviderBuilder` in `_config.py` -3. Inherit from `Block` base class -4. Define input/output schemas using `BlockSchema` -5. Implement async `run` method -6. Generate unique block ID using `uuid.uuid4()` -7. Test with `poetry run pytest backend/blocks/test/test_block.py` - -Note: when making many new blocks analyze the interfaces for each of these blocks and picture if they would go well together in a graph based editor or would they struggle to connect productively? -ex: do the inputs and outputs tie well together? - -If you get any pushback or hit complex block conditions check the new_blocks guide in the docs. - -**Handling files in blocks with `store_media_file()`:** - -When blocks need to work with files (images, videos, documents), use `store_media_file()` from `backend.util.file`. The `return_format` parameter determines what you get back: - -| Format | Use When | Returns | -|--------|----------|---------| -| `"for_local_processing"` | Processing with local tools (ffmpeg, MoviePy, PIL) | Local file path (e.g., `"image.png"`) | -| `"for_external_api"` | Sending content to external APIs (Replicate, OpenAI) | Data URI (e.g., `"data:image/png;base64,..."`) | -| `"for_block_output"` | Returning output from your block | Smart: `workspace://` in CoPilot, data URI in graphs | - -**Examples:** -```python -# INPUT: Need to process file locally with ffmpeg -local_path = await store_media_file( - file=input_data.video, - execution_context=execution_context, - return_format="for_local_processing", -) -# local_path = "video.mp4" - use with Path/ffmpeg/etc - -# INPUT: Need to send to external API like Replicate -image_b64 = await store_media_file( - file=input_data.image, - execution_context=execution_context, - return_format="for_external_api", -) -# image_b64 = "data:image/png;base64,iVBORw0..." - send to API - -# OUTPUT: Returning result from block -result_url = await store_media_file( - file=generated_image_url, - execution_context=execution_context, - return_format="for_block_output", -) -yield "image_url", result_url -# In CoPilot: result_url = "workspace://abc123" -# In graphs: result_url = "data:image/png;base64,..." -``` - -**Key points:** -- `for_block_output` is the ONLY format that auto-adapts to execution context -- Always use `for_block_output` for block outputs unless you have a specific reason not to -- Never hardcode workspace checks - let `for_block_output` handle it - -**Modifying the API:** - -1. Update route in `/backend/backend/server/routers/` -2. Add/update Pydantic models in same directory -3. Write tests alongside the route file -4. Run `poetry run test` to verify - -### Frontend guidelines: - -See `/frontend/CONTRIBUTING.md` for complete patterns. Quick reference: - -1. **Pages**: Create in `src/app/(platform)/feature-name/page.tsx` - - Add `usePageName.ts` hook for logic - - Put sub-components in local `components/` folder -2. **Components**: Structure as `ComponentName/ComponentName.tsx` + `useComponentName.ts` + `helpers.ts` - - Use design system components from `src/components/` (atoms, molecules, organisms) - - Never use `src/components/__legacy__/*` -3. **Data fetching**: Use generated API hooks from `@/app/api/__generated__/endpoints/` - - Regenerate with `pnpm generate:api` - - Pattern: `use{Method}{Version}{OperationName}` -4. **Styling**: Tailwind CSS only, use design tokens, Phosphor Icons only -5. **Testing**: Add Storybook stories for new components, Playwright for E2E -6. **Code conventions**: Function declarations (not arrow functions) for components/handlers -- Component props should be `interface Props { ... }` (not exported) unless the interface needs to be used outside the component -- Separate render logic from business logic (component.tsx + useComponent.ts + helpers.ts) -- Colocate state when possible and avoid creating large components, use sub-components ( local `/components` folder next to the parent component ) when sensible -- Avoid large hooks, abstract logic into `helpers.ts` files when sensible -- Use function declarations for components, arrow functions only for callbacks -- No barrel files or `index.ts` re-exports -- Do not use `useCallback` or `useMemo` unless strictly needed -- Avoid comments at all times unless the code is very complex - -### Security Implementation - -**Cache Protection Middleware:** - -- Located in `/backend/backend/server/middleware/security.py` -- Default behavior: Disables caching for ALL endpoints with `Cache-Control: no-store, no-cache, must-revalidate, private` -- Uses an allow list approach - only explicitly permitted paths can be cached -- Cacheable paths include: static assets (`/static/*`, `/_next/static/*`), health checks, public store pages, documentation -- Prevents sensitive data (auth tokens, API keys, user data) from being cached by browsers/proxies -- To allow caching for a new endpoint, add it to `CACHEABLE_PATHS` in the middleware -- Applied to both main API server and external API applications - ### Creating Pull Requests -- Create the PR aginst the `dev` branch of the repository. -- Ensure the branch name is descriptive (e.g., `feature/add-new-block`)/ -- Use conventional commit messages (see below)/ -- Fill out the .github/PULL_REQUEST_TEMPLATE.md template as the PR description/ +- Create the PR against the `dev` branch of the repository. +- Ensure the branch name is descriptive (e.g., `feature/add-new-block`) +- Use conventional commit messages (see below) +- Fill out the .github/PULL_REQUEST_TEMPLATE.md template as the PR description - Run the github pre-commit hooks to ensure code quality. ### Reviewing/Revising Pull Requests diff --git a/autogpt_platform/backend/CLAUDE.md b/autogpt_platform/backend/CLAUDE.md new file mode 100644 index 0000000000..53d52bb4d3 --- /dev/null +++ b/autogpt_platform/backend/CLAUDE.md @@ -0,0 +1,170 @@ +# CLAUDE.md - Backend + +This file provides guidance to Claude Code when working with the backend. + +## Essential Commands + +To run something with Python package dependencies you MUST use `poetry run ...`. + +```bash +# Install dependencies +poetry install + +# Run database migrations +poetry run prisma migrate dev + +# Start all services (database, redis, rabbitmq, clamav) +docker compose up -d + +# Run the backend as a whole +poetry run app + +# Run tests +poetry run test + +# Run specific test +poetry run pytest path/to/test_file.py::test_function_name + +# Run block tests (tests that validate all blocks work correctly) +poetry run pytest backend/blocks/test/test_block.py -xvs + +# Run tests for a specific block (e.g., GetCurrentTimeBlock) +poetry run pytest 'backend/blocks/test/test_block.py::test_available_blocks[GetCurrentTimeBlock]' -xvs + +# Lint and format +# prefer format if you want to just "fix" it and only get the errors that can't be autofixed +poetry run format # Black + isort +poetry run lint # ruff +``` + +More details can be found in @TESTING.md + +### Creating/Updating Snapshots + +When you first write a test or when the expected output changes: + +```bash +poetry run pytest path/to/test.py --snapshot-update +``` + +⚠️ **Important**: Always review snapshot changes before committing! Use `git diff` to verify the changes are expected. + +## Architecture + +- **API Layer**: FastAPI with REST and WebSocket endpoints +- **Database**: PostgreSQL with Prisma ORM, includes pgvector for embeddings +- **Queue System**: RabbitMQ for async task processing +- **Execution Engine**: Separate executor service processes agent workflows +- **Authentication**: JWT-based with Supabase integration +- **Security**: Cache protection middleware prevents sensitive data caching in browsers/proxies + +## Testing Approach + +- Uses pytest with snapshot testing for API responses +- Test files are colocated with source files (`*_test.py`) + +## Database Schema + +Key models (defined in `schema.prisma`): + +- `User`: Authentication and profile data +- `AgentGraph`: Workflow definitions with version control +- `AgentGraphExecution`: Execution history and results +- `AgentNode`: Individual nodes in a workflow +- `StoreListing`: Marketplace listings for sharing agents + +## Environment Configuration + +- **Backend**: `.env.default` (defaults) → `.env` (user overrides) + +## Common Development Tasks + +### Adding a new block + +Follow the comprehensive [Block SDK Guide](@../../docs/content/platform/block-sdk-guide.md) which covers: + +- Provider configuration with `ProviderBuilder` +- Block schema definition +- Authentication (API keys, OAuth, webhooks) +- Testing and validation +- File organization + +Quick steps: + +1. Create new file in `backend/blocks/` +2. Configure provider using `ProviderBuilder` in `_config.py` +3. Inherit from `Block` base class +4. Define input/output schemas using `BlockSchema` +5. Implement async `run` method +6. Generate unique block ID using `uuid.uuid4()` +7. Test with `poetry run pytest backend/blocks/test/test_block.py` + +Note: when making many new blocks analyze the interfaces for each of these blocks and picture if they would go well together in a graph-based editor or would they struggle to connect productively? +ex: do the inputs and outputs tie well together? + +If you get any pushback or hit complex block conditions check the new_blocks guide in the docs. + +#### Handling files in blocks with `store_media_file()` + +When blocks need to work with files (images, videos, documents), use `store_media_file()` from `backend.util.file`. The `return_format` parameter determines what you get back: + +| Format | Use When | Returns | +|--------|----------|---------| +| `"for_local_processing"` | Processing with local tools (ffmpeg, MoviePy, PIL) | Local file path (e.g., `"image.png"`) | +| `"for_external_api"` | Sending content to external APIs (Replicate, OpenAI) | Data URI (e.g., `"data:image/png;base64,..."`) | +| `"for_block_output"` | Returning output from your block | Smart: `workspace://` in CoPilot, data URI in graphs | + +**Examples:** + +```python +# INPUT: Need to process file locally with ffmpeg +local_path = await store_media_file( + file=input_data.video, + execution_context=execution_context, + return_format="for_local_processing", +) +# local_path = "video.mp4" - use with Path/ffmpeg/etc + +# INPUT: Need to send to external API like Replicate +image_b64 = await store_media_file( + file=input_data.image, + execution_context=execution_context, + return_format="for_external_api", +) +# image_b64 = "data:image/png;base64,iVBORw0..." - send to API + +# OUTPUT: Returning result from block +result_url = await store_media_file( + file=generated_image_url, + execution_context=execution_context, + return_format="for_block_output", +) +yield "image_url", result_url +# In CoPilot: result_url = "workspace://abc123" +# In graphs: result_url = "data:image/png;base64,..." +``` + +**Key points:** + +- `for_block_output` is the ONLY format that auto-adapts to execution context +- Always use `for_block_output` for block outputs unless you have a specific reason not to +- Never hardcode workspace checks - let `for_block_output` handle it + +### Modifying the API + +1. Update route in `backend/api/features/` +2. Add/update Pydantic models in same directory +3. Write tests alongside the route file +4. Run `poetry run test` to verify + +## Security Implementation + +### Cache Protection Middleware + +- Located in `backend/api/middleware/security.py` +- Default behavior: Disables caching for ALL endpoints with `Cache-Control: no-store, no-cache, must-revalidate, private` +- Uses an allow list approach - only explicitly permitted paths can be cached +- Cacheable paths include: static assets (`static/*`, `_next/static/*`), health checks, public store pages, documentation +- Prevents sensitive data (auth tokens, API keys, user data) from being cached by browsers/proxies +- To allow caching for a new endpoint, add it to `CACHEABLE_PATHS` in the middleware +- Applied to both main API server and external API applications diff --git a/autogpt_platform/backend/TESTING.md b/autogpt_platform/backend/TESTING.md index a3a5db68ef..2e09144485 100644 --- a/autogpt_platform/backend/TESTING.md +++ b/autogpt_platform/backend/TESTING.md @@ -138,7 +138,7 @@ If the test doesn't need the `user_id` specifically, mocking is not necessary as #### Using Global Auth Fixtures -Two global auth fixtures are provided by `backend/server/conftest.py`: +Two global auth fixtures are provided by `backend/api/conftest.py`: - `mock_jwt_user` - Regular user with `test_user_id` ("test-user-id") - `mock_jwt_admin` - Admin user with `admin_user_id` ("admin-user-id") diff --git a/autogpt_platform/backend/backend/api/features/builder/routes.py b/autogpt_platform/backend/backend/api/features/builder/routes.py index 1e78a1d599..5db1b5bfaa 100644 --- a/autogpt_platform/backend/backend/api/features/builder/routes.py +++ b/autogpt_platform/backend/backend/api/features/builder/routes.py @@ -19,7 +19,7 @@ router = fastapi.APIRouter( ) -# Taken from backend/server/v2/store/db.py +# Taken from backend/api/features/store/db.py def sanitize_query(query: str | None) -> str | None: if query is None: return query diff --git a/autogpt_platform/backend/backend/api/features/chat/service.py b/autogpt_platform/backend/backend/api/features/chat/service.py index 20216162b5..f1f3156713 100644 --- a/autogpt_platform/backend/backend/api/features/chat/service.py +++ b/autogpt_platform/backend/backend/api/features/chat/service.py @@ -3,7 +3,8 @@ import logging import time from asyncio import CancelledError from collections.abc import AsyncGenerator -from typing import Any +from dataclasses import dataclass +from typing import Any, cast import openai import orjson @@ -15,7 +16,14 @@ from openai import ( PermissionDeniedError, RateLimitError, ) -from openai.types.chat import ChatCompletionChunk, ChatCompletionToolParam +from openai.types.chat import ( + ChatCompletionAssistantMessageParam, + ChatCompletionChunk, + ChatCompletionMessageParam, + ChatCompletionStreamOptionsParam, + ChatCompletionSystemMessageParam, + ChatCompletionToolParam, +) from backend.data.redis_client import get_redis_async from backend.data.understanding import ( @@ -23,6 +31,7 @@ from backend.data.understanding import ( get_business_understanding, ) from backend.util.exceptions import NotFoundError +from backend.util.prompt import estimate_token_count from backend.util.settings import Settings from . import db as chat_db @@ -794,6 +803,201 @@ def _is_region_blocked_error(error: Exception) -> bool: return "not available in your region" in str(error).lower() +# Context window management constants +TOKEN_THRESHOLD = 120_000 +KEEP_RECENT_MESSAGES = 15 + + +@dataclass +class ContextWindowResult: + """Result of context window management.""" + + messages: list[dict[str, Any]] + token_count: int + was_compacted: bool + error: str | None = None + + +def _messages_to_dicts(messages: list) -> list[dict[str, Any]]: + """Convert message objects to dicts, filtering None values. + + Handles both TypedDict (dict-like) and other message formats. + """ + result = [] + for msg in messages: + if msg is None: + continue + if isinstance(msg, dict): + msg_dict = {k: v for k, v in msg.items() if v is not None} + else: + msg_dict = dict(msg) + result.append(msg_dict) + return result + + +async def _manage_context_window( + messages: list, + model: str, + api_key: str | None = None, + base_url: str | None = None, +) -> ContextWindowResult: + """ + Manage context window by summarizing old messages if token count exceeds threshold. + + This function handles context compaction for LLM calls by: + 1. Counting tokens in the message list + 2. If over threshold, summarizing old messages while keeping recent ones + 3. Ensuring tool_call/tool_response pairs stay intact + 4. Progressively reducing message count if still over limit + + Args: + messages: List of messages in OpenAI format (with system prompt if present) + model: Model name for token counting + api_key: API key for summarization calls + base_url: Base URL for summarization calls + + Returns: + ContextWindowResult with compacted messages and metadata + """ + if not messages: + return ContextWindowResult([], 0, False, "No messages to compact") + + messages_dict = _messages_to_dicts(messages) + + # Normalize model name for token counting (tiktoken only supports OpenAI models) + token_count_model = model.split("/")[-1] if "/" in model else model + if "claude" in token_count_model.lower() or not any( + known in token_count_model.lower() + for known in ["gpt", "o1", "chatgpt", "text-"] + ): + token_count_model = "gpt-4o" + + try: + token_count = estimate_token_count(messages_dict, model=token_count_model) + except Exception as e: + logger.warning(f"Token counting failed: {e}. Using gpt-4o approximation.") + token_count_model = "gpt-4o" + token_count = estimate_token_count(messages_dict, model=token_count_model) + + if token_count <= TOKEN_THRESHOLD: + return ContextWindowResult(messages, token_count, False) + + has_system_prompt = messages[0].get("role") == "system" + slice_start = max(0, len(messages_dict) - KEEP_RECENT_MESSAGES) + recent_messages = _ensure_tool_pairs_intact( + messages_dict[-KEEP_RECENT_MESSAGES:], messages_dict, slice_start + ) + + # Determine old messages to summarize (explicit bounds to avoid slice edge cases) + system_msg = messages[0] if has_system_prompt else None + if has_system_prompt: + old_messages_dict = ( + messages_dict[1:-KEEP_RECENT_MESSAGES] + if len(messages_dict) > KEEP_RECENT_MESSAGES + 1 + else [] + ) + else: + old_messages_dict = ( + messages_dict[:-KEEP_RECENT_MESSAGES] + if len(messages_dict) > KEEP_RECENT_MESSAGES + else [] + ) + + # Try to summarize old messages, fall back to truncation on failure + summary_msg = None + if old_messages_dict: + try: + summary_text = await _summarize_messages( + old_messages_dict, model=model, api_key=api_key, base_url=base_url + ) + summary_msg = ChatCompletionAssistantMessageParam( + role="assistant", + content=f"[Previous conversation summary — for context only]: {summary_text}", + ) + base = [system_msg, summary_msg] if has_system_prompt else [summary_msg] + messages = base + recent_messages + logger.info( + f"Context summarized: {token_count} tokens, " + f"summarized {len(old_messages_dict)} msgs, kept {KEEP_RECENT_MESSAGES}" + ) + except Exception as e: + logger.warning(f"Summarization failed, falling back to truncation: {e}") + messages = ( + [system_msg] + recent_messages if has_system_prompt else recent_messages + ) + else: + logger.warning( + f"Token count {token_count} exceeds threshold but no old messages to summarize" + ) + + new_token_count = estimate_token_count( + _messages_to_dicts(messages), model=token_count_model + ) + + # Progressive truncation if still over limit + if new_token_count > TOKEN_THRESHOLD: + logger.warning( + f"Still over limit: {new_token_count} tokens. Reducing messages." + ) + base_msgs = ( + recent_messages + if old_messages_dict + else (messages_dict[1:] if has_system_prompt else messages_dict) + ) + + def build_messages(recent: list) -> list: + """Build message list with optional system prompt and summary.""" + prefix = [] + if has_system_prompt and system_msg: + prefix.append(system_msg) + if summary_msg: + prefix.append(summary_msg) + return prefix + recent + + for keep_count in [12, 10, 8, 5, 3, 2, 1, 0]: + if keep_count == 0: + messages = build_messages([]) + if not messages: + continue + elif len(base_msgs) < keep_count: + continue + else: + reduced = _ensure_tool_pairs_intact( + base_msgs[-keep_count:], + base_msgs, + max(0, len(base_msgs) - keep_count), + ) + messages = build_messages(reduced) + + new_token_count = estimate_token_count( + _messages_to_dicts(messages), model=token_count_model + ) + if new_token_count <= TOKEN_THRESHOLD: + logger.info( + f"Reduced to {keep_count} messages, {new_token_count} tokens" + ) + break + else: + logger.error( + f"Cannot reduce below threshold. Final: {new_token_count} tokens" + ) + if has_system_prompt and len(messages) > 1: + messages = messages[1:] + logger.critical("Dropped system prompt as last resort") + return ContextWindowResult( + messages, new_token_count, True, "System prompt dropped" + ) + # No system prompt to drop - return error so callers don't proceed with oversized context + return ContextWindowResult( + messages, + new_token_count, + True, + "Unable to reduce context below token limit", + ) + + return ContextWindowResult(messages, new_token_count, True) + + async def _summarize_messages( messages: list, model: str, @@ -1022,11 +1226,8 @@ async def _stream_chat_chunks( logger.info("Starting pure chat stream") - # Build messages with system prompt prepended messages = session.to_openai_messages() if system_prompt: - from openai.types.chat import ChatCompletionSystemMessageParam - system_message = ChatCompletionSystemMessageParam( role="system", content=system_prompt, @@ -1034,314 +1235,38 @@ async def _stream_chat_chunks( messages = [system_message] + messages # Apply context window management - token_count = 0 # Initialize for exception handler - try: - from backend.util.prompt import estimate_token_count + context_result = await _manage_context_window( + messages=messages, + model=model, + api_key=config.api_key, + base_url=config.base_url, + ) - # Convert to dict for token counting - # OpenAI message types are TypedDicts, so they're already dict-like - messages_dict = [] - for msg in messages: - # TypedDict objects are already dicts, just filter None values - if isinstance(msg, dict): - msg_dict = {k: v for k, v in msg.items() if v is not None} - else: - # Fallback for unexpected types - msg_dict = dict(msg) - messages_dict.append(msg_dict) - - # Estimate tokens using appropriate tokenizer - # Normalize model name for token counting (tiktoken only supports OpenAI models) - token_count_model = model - if "/" in model: - # Strip provider prefix (e.g., "anthropic/claude-opus-4.5" -> "claude-opus-4.5") - token_count_model = model.split("/")[-1] - - # For Claude and other non-OpenAI models, approximate with gpt-4o tokenizer - # Most modern LLMs have similar tokenization (~1 token per 4 chars) - if "claude" in token_count_model.lower() or not any( - known in token_count_model.lower() - for known in ["gpt", "o1", "chatgpt", "text-"] - ): - token_count_model = "gpt-4o" - - # Attempt token counting with error handling - try: - token_count = estimate_token_count(messages_dict, model=token_count_model) - except Exception as token_error: - # If token counting fails, use gpt-4o as fallback approximation - logger.warning( - f"Token counting failed for model {token_count_model}: {token_error}. " - "Using gpt-4o approximation." - ) - token_count = estimate_token_count(messages_dict, model="gpt-4o") - - # If over threshold, summarize old messages - if token_count > 120_000: - KEEP_RECENT = 15 - - # Check if we have a system prompt at the start - has_system_prompt = ( - len(messages) > 0 and messages[0].get("role") == "system" - ) - - # Always attempt mitigation when over limit, even with few messages - if messages: - # Split messages based on whether system prompt exists - # Calculate start index for the slice - slice_start = max(0, len(messages_dict) - KEEP_RECENT) - recent_messages = messages_dict[-KEEP_RECENT:] - - # Ensure tool_call/tool_response pairs stay together - # This prevents API errors from orphan tool responses - recent_messages = _ensure_tool_pairs_intact( - recent_messages, messages_dict, slice_start - ) - - if has_system_prompt: - # Keep system prompt separate, summarize everything between system and recent - system_msg = messages[0] - old_messages_dict = messages_dict[1:-KEEP_RECENT] - else: - # No system prompt, summarize everything except recent - system_msg = None - old_messages_dict = messages_dict[:-KEEP_RECENT] - - # Summarize any non-empty old messages (no minimum threshold) - # If we're over the token limit, we need to compress whatever we can - if old_messages_dict: - # Summarize old messages using the same model as chat - summary_text = await _summarize_messages( - old_messages_dict, - model=model, - api_key=config.api_key, - base_url=config.base_url, - ) - - # Build new message list - # Use assistant role (not system) to prevent privilege escalation - # of user-influenced content to instruction-level authority - from openai.types.chat import ChatCompletionAssistantMessageParam - - summary_msg = ChatCompletionAssistantMessageParam( - role="assistant", - content=( - "[Previous conversation summary — for context only]: " - f"{summary_text}" - ), - ) - - # Rebuild messages based on whether we have a system prompt - if has_system_prompt: - # system_prompt + summary + recent_messages - messages = [system_msg, summary_msg] + recent_messages - else: - # summary + recent_messages (no original system prompt) - messages = [summary_msg] + recent_messages - - logger.info( - f"Context summarized: {token_count} tokens, " - f"summarized {len(old_messages_dict)} old messages, " - f"kept last {KEEP_RECENT} messages" - ) - - # Fallback: If still over limit after summarization, progressively drop recent messages - # This handles edge cases where recent messages are extremely large - new_messages_dict = [] - for msg in messages: - if isinstance(msg, dict): - msg_dict = {k: v for k, v in msg.items() if v is not None} - else: - msg_dict = dict(msg) - new_messages_dict.append(msg_dict) - - new_token_count = estimate_token_count( - new_messages_dict, model=token_count_model - ) - - if new_token_count > 120_000: - # Still over limit - progressively reduce KEEP_RECENT - logger.warning( - f"Still over limit after summarization: {new_token_count} tokens. " - "Reducing number of recent messages kept." - ) - - for keep_count in [12, 10, 8, 5, 3, 2, 1, 0]: - if keep_count == 0: - # Try with just system prompt + summary (no recent messages) - if has_system_prompt: - messages = [system_msg, summary_msg] - else: - messages = [summary_msg] - logger.info( - "Trying with 0 recent messages (system + summary only)" - ) - else: - # Slice from ORIGINAL recent_messages to avoid duplicating summary - reduced_recent = ( - recent_messages[-keep_count:] - if len(recent_messages) >= keep_count - else recent_messages - ) - # Ensure tool pairs stay intact in the reduced slice - reduced_slice_start = max( - 0, len(recent_messages) - keep_count - ) - reduced_recent = _ensure_tool_pairs_intact( - reduced_recent, recent_messages, reduced_slice_start - ) - if has_system_prompt: - messages = [ - system_msg, - summary_msg, - ] + reduced_recent - else: - messages = [summary_msg] + reduced_recent - - new_messages_dict = [] - for msg in messages: - if isinstance(msg, dict): - msg_dict = { - k: v for k, v in msg.items() if v is not None - } - else: - msg_dict = dict(msg) - new_messages_dict.append(msg_dict) - - new_token_count = estimate_token_count( - new_messages_dict, model=token_count_model - ) - - if new_token_count <= 120_000: - logger.info( - f"Reduced to {keep_count} recent messages, " - f"now {new_token_count} tokens" - ) - break - else: - logger.error( - f"Unable to reduce token count below threshold even with 0 messages. " - f"Final count: {new_token_count} tokens" - ) - # ABSOLUTE LAST RESORT: Drop system prompt - # This should only happen if summary itself is massive - if has_system_prompt and len(messages) > 1: - messages = messages[1:] # Drop system prompt - logger.critical( - "CRITICAL: Dropped system prompt as absolute last resort. " - "Behavioral consistency may be affected." - ) - # Yield error to user - yield StreamError( - errorText=( - "Warning: System prompt dropped due to size constraints. " - "Assistant behavior may be affected." - ) - ) - else: - # No old messages to summarize - all messages are "recent" - # Apply progressive truncation to reduce token count - logger.warning( - f"Token count {token_count} exceeds threshold but no old messages to summarize. " - f"Applying progressive truncation to recent messages." - ) - - # Create a base list excluding system prompt to avoid duplication - # This is the pool of messages we'll slice from in the loop - # Use messages_dict for type consistency with _ensure_tool_pairs_intact - base_msgs = ( - messages_dict[1:] if has_system_prompt else messages_dict - ) - - # Try progressively smaller keep counts - new_token_count = token_count # Initialize with current count - for keep_count in [12, 10, 8, 5, 3, 2, 1, 0]: - if keep_count == 0: - # Try with just system prompt (no recent messages) - if has_system_prompt: - messages = [system_msg] - logger.info( - "Trying with 0 recent messages (system prompt only)" - ) - else: - # No system prompt and no recent messages = empty messages list - # This is invalid, skip this iteration - continue - else: - if len(base_msgs) < keep_count: - continue # Skip if we don't have enough messages - - # Slice from base_msgs to get recent messages (without system prompt) - recent_messages = base_msgs[-keep_count:] - - # Ensure tool pairs stay intact in the reduced slice - reduced_slice_start = max(0, len(base_msgs) - keep_count) - recent_messages = _ensure_tool_pairs_intact( - recent_messages, base_msgs, reduced_slice_start - ) - - if has_system_prompt: - messages = [system_msg] + recent_messages - else: - messages = recent_messages - - new_messages_dict = [] - for msg in messages: - if msg is None: - continue # Skip None messages (type safety) - if isinstance(msg, dict): - msg_dict = { - k: v for k, v in msg.items() if v is not None - } - else: - msg_dict = dict(msg) - new_messages_dict.append(msg_dict) - - new_token_count = estimate_token_count( - new_messages_dict, model=token_count_model - ) - - if new_token_count <= 120_000: - logger.info( - f"Reduced to {keep_count} recent messages, " - f"now {new_token_count} tokens" - ) - break - else: - # Even with 0 messages still over limit - logger.error( - f"Unable to reduce token count below threshold even with 0 messages. " - f"Final count: {new_token_count} tokens. Messages may be extremely large." - ) - # ABSOLUTE LAST RESORT: Drop system prompt - if has_system_prompt and len(messages) > 1: - messages = messages[1:] # Drop system prompt - logger.critical( - "CRITICAL: Dropped system prompt as absolute last resort. " - "Behavioral consistency may be affected." - ) - # Yield error to user - yield StreamError( - errorText=( - "Warning: System prompt dropped due to size constraints. " - "Assistant behavior may be affected." - ) - ) - - except Exception as e: - logger.error(f"Context summarization failed: {e}", exc_info=True) - # If we were over the token limit, yield error to user - # Don't silently continue with oversized messages that will fail - if token_count > 120_000: + if context_result.error: + if "System prompt dropped" in context_result.error: + # Warning only - continue with reduced context yield StreamError( errorText=( - f"Unable to manage context window (token limit exceeded: {token_count} tokens). " - "Context summarization failed. Please start a new conversation." + "Warning: System prompt dropped due to size constraints. " + "Assistant behavior may be affected." + ) + ) + else: + # Any other error - abort to prevent failed LLM calls + yield StreamError( + errorText=( + f"Context window management failed: {context_result.error}. " + "Please start a new conversation." ) ) yield StreamFinish() return - # Otherwise, continue with original messages (under limit) + + messages = context_result.messages + if context_result.was_compacted: + logger.info( + f"Context compacted for streaming: {context_result.token_count} tokens" + ) # Loop to handle tool calls and continue conversation while True: @@ -1369,14 +1294,6 @@ async def _stream_chat_chunks( :128 ] # OpenRouter limit - # Create the stream with proper types - from typing import cast - - from openai.types.chat import ( - ChatCompletionMessageParam, - ChatCompletionStreamOptionsParam, - ) - stream = await client.chat.completions.create( model=model, messages=cast(list[ChatCompletionMessageParam], messages), @@ -1834,6 +1751,11 @@ async def _execute_long_running_tool( tool_call_id=tool_call_id, result=error_response.model_dump_json(), ) + # Generate LLM continuation so user sees explanation even for errors + try: + await _generate_llm_continuation(session_id=session_id, user_id=user_id) + except Exception as llm_err: + logger.warning(f"Failed to generate LLM continuation for error: {llm_err}") finally: await _mark_operation_completed(tool_call_id) @@ -1895,17 +1817,36 @@ async def _generate_llm_continuation( # Build system prompt system_prompt, _ = await _build_system_prompt(user_id) - # Build messages in OpenAI format messages = session.to_openai_messages() if system_prompt: - from openai.types.chat import ChatCompletionSystemMessageParam - system_message = ChatCompletionSystemMessageParam( role="system", content=system_prompt, ) messages = [system_message] + messages + # Apply context window management to prevent oversized requests + context_result = await _manage_context_window( + messages=messages, + model=config.model, + api_key=config.api_key, + base_url=config.base_url, + ) + + if context_result.error and "System prompt dropped" not in context_result.error: + logger.error( + f"Context window management failed for session {session_id}: " + f"{context_result.error} (tokens={context_result.token_count})" + ) + return + + messages = context_result.messages + if context_result.was_compacted: + logger.info( + f"Context compacted for LLM continuation: " + f"{context_result.token_count} tokens" + ) + # Build extra_body for tracing extra_body: dict[str, Any] = { "posthogProperties": { @@ -1918,19 +1859,54 @@ async def _generate_llm_continuation( if session_id: extra_body["session_id"] = session_id[:128] - # Make non-streaming LLM call (no tools - just text response) - from typing import cast + retry_count = 0 + last_error: Exception | None = None + response = None - from openai.types.chat import ChatCompletionMessageParam + while retry_count <= MAX_RETRIES: + try: + logger.info( + f"Generating LLM continuation for session {session_id}" + f"{f' (retry {retry_count}/{MAX_RETRIES})' if retry_count > 0 else ''}" + ) - # No tools parameter = text-only response (no tool calls) - response = await client.chat.completions.create( - model=config.model, - messages=cast(list[ChatCompletionMessageParam], messages), - extra_body=extra_body, - ) + response = await client.chat.completions.create( + model=config.model, + messages=cast(list[ChatCompletionMessageParam], messages), + extra_body=extra_body, + ) + last_error = None # Clear any previous error on success + break # Success, exit retry loop + except Exception as e: + last_error = e + if _is_retryable_error(e) and retry_count < MAX_RETRIES: + retry_count += 1 + delay = min( + BASE_DELAY_SECONDS * (2 ** (retry_count - 1)), + MAX_DELAY_SECONDS, + ) + logger.warning( + f"Retryable error in LLM continuation: {e!s}. " + f"Retrying in {delay:.1f}s (attempt {retry_count}/{MAX_RETRIES})" + ) + await asyncio.sleep(delay) + continue + else: + # Non-retryable error - log and exit gracefully + logger.error( + f"Non-retryable error in LLM continuation: {e!s}", + exc_info=True, + ) + return - if response.choices and response.choices[0].message.content: + if last_error: + logger.error( + f"Max retries ({MAX_RETRIES}) exceeded for LLM continuation. " + f"Last error: {last_error!s}" + ) + return + + if response and response.choices and response.choices[0].message.content: assistant_content = response.choices[0].message.content # Reload session from DB to avoid race condition with user messages diff --git a/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/__init__.py b/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/__init__.py index 392f642c41..b7650b3cbd 100644 --- a/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/__init__.py +++ b/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/__init__.py @@ -2,27 +2,54 @@ from .core import ( AgentGeneratorNotConfiguredError, + AgentJsonValidationError, + AgentSummary, + DecompositionResult, + DecompositionStep, + LibraryAgentSummary, + MarketplaceAgentSummary, decompose_goal, + enrich_library_agents_from_steps, + extract_search_terms_from_steps, + extract_uuids_from_text, generate_agent, generate_agent_patch, get_agent_as_json, + get_all_relevant_agents_for_generation, + get_library_agent_by_graph_id, + get_library_agent_by_id, + get_library_agents_for_generation, json_to_graph, save_agent_to_library, + search_marketplace_agents_for_generation, ) +from .errors import get_user_message_for_error from .service import health_check as check_external_service_health from .service import is_external_service_configured __all__ = [ - # Core functions + "AgentGeneratorNotConfiguredError", + "AgentJsonValidationError", + "AgentSummary", + "DecompositionResult", + "DecompositionStep", + "LibraryAgentSummary", + "MarketplaceAgentSummary", + "check_external_service_health", "decompose_goal", + "enrich_library_agents_from_steps", + "extract_search_terms_from_steps", + "extract_uuids_from_text", "generate_agent", "generate_agent_patch", - "save_agent_to_library", "get_agent_as_json", - "json_to_graph", - # Exceptions - "AgentGeneratorNotConfiguredError", - # Service + "get_all_relevant_agents_for_generation", + "get_library_agent_by_graph_id", + "get_library_agent_by_id", + "get_library_agents_for_generation", + "get_user_message_for_error", "is_external_service_configured", - "check_external_service_health", + "json_to_graph", + "save_agent_to_library", + "search_marketplace_agents_for_generation", ] diff --git a/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/core.py b/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/core.py index fc15587110..0ddd2aa86b 100644 --- a/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/core.py +++ b/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/core.py @@ -1,11 +1,22 @@ """Core agent generation functions.""" import logging +import re import uuid -from typing import Any +from typing import Any, NotRequired, TypedDict from backend.api.features.library import db as library_db -from backend.data.graph import Graph, Link, Node, create_graph +from backend.api.features.store import db as store_db +from backend.data.graph import ( + Graph, + Link, + Node, + create_graph, + get_graph, + get_graph_all_versions, + get_store_listed_graphs, +) +from backend.util.exceptions import DatabaseError, NotFoundError from .service import ( decompose_goal_external, @@ -16,6 +27,74 @@ from .service import ( logger = logging.getLogger(__name__) +AGENT_EXECUTOR_BLOCK_ID = "e189baac-8c20-45a1-94a7-55177ea42565" + + +class ExecutionSummary(TypedDict): + """Summary of a single execution for quality assessment.""" + + status: str + correctness_score: NotRequired[float] + activity_summary: NotRequired[str] + + +class LibraryAgentSummary(TypedDict): + """Summary of a library agent for sub-agent composition. + + Includes recent executions to help the LLM decide whether to use this agent. + Each execution shows status, correctness_score (0-1), and activity_summary. + """ + + graph_id: str + graph_version: int + name: str + description: str + input_schema: dict[str, Any] + output_schema: dict[str, Any] + recent_executions: NotRequired[list[ExecutionSummary]] + + +class MarketplaceAgentSummary(TypedDict): + """Summary of a marketplace agent for sub-agent composition.""" + + name: str + description: str + sub_heading: str + creator: str + is_marketplace_agent: bool + + +class DecompositionStep(TypedDict, total=False): + """A single step in decomposed instructions.""" + + description: str + action: str + block_name: str + tool: str + name: str + + +class DecompositionResult(TypedDict, total=False): + """Result from decompose_goal - can be instructions, questions, or error.""" + + type: str + steps: list[DecompositionStep] + questions: list[dict[str, Any]] + error: str + error_type: str + + +AgentSummary = LibraryAgentSummary | MarketplaceAgentSummary | dict[str, Any] + + +def _to_dict_list( + agents: list[AgentSummary] | list[dict[str, Any]] | None, +) -> list[dict[str, Any]] | None: + """Convert typed agent summaries to plain dicts for external service calls.""" + if agents is None: + return None + return [dict(a) for a in agents] + class AgentGeneratorNotConfiguredError(Exception): """Raised when the external Agent Generator service is not configured.""" @@ -36,15 +115,422 @@ def _check_service_configured() -> None: ) -async def decompose_goal(description: str, context: str = "") -> dict[str, Any] | None: +_UUID_PATTERN = 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}", + re.IGNORECASE, +) + + +def extract_uuids_from_text(text: str) -> list[str]: + """Extract all UUID v4 strings from text. + + Args: + text: Text that may contain UUIDs (e.g., user's goal description) + + Returns: + List of unique UUIDs found in the text (lowercase) + """ + matches = _UUID_PATTERN.findall(text) + return list({m.lower() for m in matches}) + + +async def get_library_agent_by_id( + user_id: str, agent_id: str +) -> LibraryAgentSummary | None: + """Fetch a specific library agent by its ID (library agent ID or graph_id). + + This function tries multiple lookup strategies: + 1. First tries to find by graph_id (AgentGraph primary key) + 2. If not found, tries to find by library agent ID (LibraryAgent primary key) + + This handles both cases: + - User provides graph_id (e.g., from AgentExecutorBlock) + - User provides library agent ID (e.g., from library URL) + + Args: + user_id: The user ID + agent_id: The ID to look up (can be graph_id or library agent ID) + + Returns: + LibraryAgentSummary if found, None otherwise + """ + try: + agent = await library_db.get_library_agent_by_graph_id(user_id, agent_id) + if agent: + logger.debug(f"Found library agent by graph_id: {agent.name}") + return LibraryAgentSummary( + graph_id=agent.graph_id, + graph_version=agent.graph_version, + name=agent.name, + description=agent.description, + input_schema=agent.input_schema, + output_schema=agent.output_schema, + ) + except DatabaseError: + raise + except Exception as e: + logger.debug(f"Could not fetch library agent by graph_id {agent_id}: {e}") + + try: + agent = await library_db.get_library_agent(agent_id, user_id) + if agent: + logger.debug(f"Found library agent by library_id: {agent.name}") + return LibraryAgentSummary( + graph_id=agent.graph_id, + graph_version=agent.graph_version, + name=agent.name, + description=agent.description, + input_schema=agent.input_schema, + output_schema=agent.output_schema, + ) + except NotFoundError: + logger.debug(f"Library agent not found by library_id: {agent_id}") + except DatabaseError: + raise + except Exception as e: + logger.warning( + f"Could not fetch library agent by library_id {agent_id}: {e}", + exc_info=True, + ) + + return None + + +get_library_agent_by_graph_id = get_library_agent_by_id + + +async def get_library_agents_for_generation( + user_id: str, + search_query: str | None = None, + exclude_graph_id: str | None = None, + max_results: int = 15, +) -> list[LibraryAgentSummary]: + """Fetch user's library agents formatted for Agent Generator. + + Uses search-based fetching to return relevant agents instead of all agents. + This is more scalable for users with large libraries. + + Includes recent_executions list to help the LLM assess agent quality: + - Each execution has status, correctness_score (0-1), and activity_summary + - This gives the LLM concrete examples of recent performance + + Args: + user_id: The user ID + search_query: Optional search term to find relevant agents (user's goal/description) + exclude_graph_id: Optional graph ID to exclude (prevents circular references) + max_results: Maximum number of agents to return (default 15) + + Returns: + List of LibraryAgentSummary with schemas and recent executions for sub-agent composition + """ + try: + response = await library_db.list_library_agents( + user_id=user_id, + search_term=search_query, + page=1, + page_size=max_results, + include_executions=True, + ) + + results: list[LibraryAgentSummary] = [] + for agent in response.agents: + if exclude_graph_id is not None and agent.graph_id == exclude_graph_id: + continue + + summary = LibraryAgentSummary( + graph_id=agent.graph_id, + graph_version=agent.graph_version, + name=agent.name, + description=agent.description, + input_schema=agent.input_schema, + output_schema=agent.output_schema, + ) + if agent.recent_executions: + exec_summaries: list[ExecutionSummary] = [] + for ex in agent.recent_executions: + exec_sum = ExecutionSummary(status=ex.status) + if ex.correctness_score is not None: + exec_sum["correctness_score"] = ex.correctness_score + if ex.activity_summary: + exec_sum["activity_summary"] = ex.activity_summary + exec_summaries.append(exec_sum) + summary["recent_executions"] = exec_summaries + results.append(summary) + return results + except DatabaseError: + raise + except Exception as e: + logger.warning(f"Failed to fetch library agents: {e}") + return [] + + +async def search_marketplace_agents_for_generation( + search_query: str, + max_results: int = 10, +) -> list[LibraryAgentSummary]: + """Search marketplace agents formatted for Agent Generator. + + Fetches marketplace agents and their full schemas so they can be used + as sub-agents in generated workflows. + + Args: + search_query: Search term to find relevant public agents + max_results: Maximum number of agents to return (default 10) + + Returns: + List of LibraryAgentSummary with full input/output schemas + """ + try: + response = await store_db.get_store_agents( + search_query=search_query, + page=1, + page_size=max_results, + ) + + agents_with_graphs = [ + agent for agent in response.agents if agent.agent_graph_id + ] + + if not agents_with_graphs: + return [] + + graph_ids = [agent.agent_graph_id for agent in agents_with_graphs] + graphs = await get_store_listed_graphs(*graph_ids) + + results: list[LibraryAgentSummary] = [] + for agent in agents_with_graphs: + graph_id = agent.agent_graph_id + if graph_id and graph_id in graphs: + graph = graphs[graph_id] + results.append( + LibraryAgentSummary( + graph_id=graph.id, + graph_version=graph.version, + name=agent.agent_name, + description=agent.description, + input_schema=graph.input_schema, + output_schema=graph.output_schema, + ) + ) + return results + except Exception as e: + logger.warning(f"Failed to search marketplace agents: {e}") + return [] + + +async def get_all_relevant_agents_for_generation( + user_id: str, + search_query: str | None = None, + exclude_graph_id: str | None = None, + include_library: bool = True, + include_marketplace: bool = True, + max_library_results: int = 15, + max_marketplace_results: int = 10, +) -> list[AgentSummary]: + """Fetch relevant agents from library and/or marketplace. + + Searches both user's library and marketplace by default. + Explicitly mentioned UUIDs in the search query are always looked up. + + Args: + user_id: The user ID + search_query: Search term to find relevant agents (user's goal/description) + exclude_graph_id: Optional graph ID to exclude (prevents circular references) + include_library: Whether to search user's library (default True) + include_marketplace: Whether to also search marketplace (default True) + max_library_results: Max library agents to return (default 15) + max_marketplace_results: Max marketplace agents to return (default 10) + + Returns: + List of AgentSummary with full schemas (both library and marketplace agents) + """ + agents: list[AgentSummary] = [] + seen_graph_ids: set[str] = set() + + if search_query: + mentioned_uuids = extract_uuids_from_text(search_query) + for graph_id in mentioned_uuids: + if graph_id == exclude_graph_id: + continue + agent = await get_library_agent_by_graph_id(user_id, graph_id) + agent_graph_id = agent.get("graph_id") if agent else None + if agent and agent_graph_id and agent_graph_id not in seen_graph_ids: + agents.append(agent) + seen_graph_ids.add(agent_graph_id) + logger.debug( + f"Found explicitly mentioned agent: {agent.get('name') or 'Unknown'}" + ) + + if include_library: + library_agents = await get_library_agents_for_generation( + user_id=user_id, + search_query=search_query, + exclude_graph_id=exclude_graph_id, + max_results=max_library_results, + ) + for agent in library_agents: + graph_id = agent.get("graph_id") + if graph_id and graph_id not in seen_graph_ids: + agents.append(agent) + seen_graph_ids.add(graph_id) + + if include_marketplace and search_query: + marketplace_agents = await search_marketplace_agents_for_generation( + search_query=search_query, + max_results=max_marketplace_results, + ) + for agent in marketplace_agents: + graph_id = agent.get("graph_id") + if graph_id and graph_id not in seen_graph_ids: + agents.append(agent) + seen_graph_ids.add(graph_id) + + return agents + + +def extract_search_terms_from_steps( + decomposition_result: DecompositionResult | dict[str, Any], +) -> list[str]: + """Extract search terms from decomposed instruction steps. + + Analyzes the decomposition result to extract relevant keywords + for additional library agent searches. + + Args: + decomposition_result: Result from decompose_goal containing steps + + Returns: + List of unique search terms extracted from steps + """ + search_terms: list[str] = [] + + if decomposition_result.get("type") != "instructions": + return search_terms + + steps = decomposition_result.get("steps", []) + if not steps: + return search_terms + + step_keys: list[str] = ["description", "action", "block_name", "tool", "name"] + + for step in steps: + for key in step_keys: + value = step.get(key) # type: ignore[union-attr] + if isinstance(value, str) and len(value) > 3: + search_terms.append(value) + + seen: set[str] = set() + unique_terms: list[str] = [] + for term in search_terms: + term_lower = term.lower() + if term_lower not in seen: + seen.add(term_lower) + unique_terms.append(term) + + return unique_terms + + +async def enrich_library_agents_from_steps( + user_id: str, + decomposition_result: DecompositionResult | dict[str, Any], + existing_agents: list[AgentSummary] | list[dict[str, Any]], + exclude_graph_id: str | None = None, + include_marketplace: bool = True, + max_additional_results: int = 10, +) -> list[AgentSummary] | list[dict[str, Any]]: + """Enrich library agents list with additional searches based on decomposed steps. + + This implements two-phase search: after decomposition, we search for additional + relevant agents based on the specific steps identified. + + Args: + user_id: The user ID + decomposition_result: Result from decompose_goal containing steps + existing_agents: Already fetched library agents from initial search + exclude_graph_id: Optional graph ID to exclude + include_marketplace: Whether to also search marketplace + max_additional_results: Max additional agents per search term (default 10) + + Returns: + Combined list of library agents (existing + newly discovered) + """ + search_terms = extract_search_terms_from_steps(decomposition_result) + + if not search_terms: + return existing_agents + + existing_ids: set[str] = set() + existing_names: set[str] = set() + + for agent in existing_agents: + agent_name = agent.get("name") + if agent_name and isinstance(agent_name, str): + existing_names.add(agent_name.lower()) + graph_id = agent.get("graph_id") # type: ignore[call-overload] + if graph_id and isinstance(graph_id, str): + existing_ids.add(graph_id) + + all_agents: list[AgentSummary] | list[dict[str, Any]] = list(existing_agents) + + for term in search_terms[:3]: + try: + additional_agents = await get_all_relevant_agents_for_generation( + user_id=user_id, + search_query=term, + exclude_graph_id=exclude_graph_id, + include_marketplace=include_marketplace, + max_library_results=max_additional_results, + max_marketplace_results=5, + ) + + for agent in additional_agents: + agent_name = agent.get("name") + if not agent_name or not isinstance(agent_name, str): + continue + agent_name_lower = agent_name.lower() + + if agent_name_lower in existing_names: + continue + + graph_id = agent.get("graph_id") # type: ignore[call-overload] + if graph_id and graph_id in existing_ids: + continue + + all_agents.append(agent) + existing_names.add(agent_name_lower) + if graph_id and isinstance(graph_id, str): + existing_ids.add(graph_id) + + except DatabaseError: + logger.error(f"Database error searching for agents with term '{term}'") + raise + except Exception as e: + logger.warning( + f"Failed to search for additional agents with term '{term}': {e}" + ) + + logger.debug( + f"Enriched library agents: {len(existing_agents)} initial + " + f"{len(all_agents) - len(existing_agents)} additional = {len(all_agents)} total" + ) + + return all_agents + + +async def decompose_goal( + description: str, + context: str = "", + library_agents: list[AgentSummary] | None = None, +) -> DecompositionResult | 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) + library_agents: User's library agents available for sub-agent composition Returns: - Dict with either: + DecompositionResult with either: - {"type": "clarifying_questions", "questions": [...]} - {"type": "instructions", "steps": [...]} Or None on error @@ -54,26 +540,36 @@ async def decompose_goal(description: str, context: str = "") -> dict[str, Any] """ _check_service_configured() logger.info("Calling external Agent Generator service for decompose_goal") - return await decompose_goal_external(description, context) + result = await decompose_goal_external( + description, context, _to_dict_list(library_agents) + ) + return result # type: ignore[return-value] -async def generate_agent(instructions: dict[str, Any]) -> dict[str, Any] | None: +async def generate_agent( + instructions: DecompositionResult | dict[str, Any], + library_agents: list[AgentSummary] | list[dict[str, Any]] | None = None, +) -> dict[str, Any] | None: """Generate agent JSON from instructions. Args: instructions: Structured instructions from decompose_goal + library_agents: User's library agents available for sub-agent composition Returns: - Agent JSON dict or None on error + Agent JSON dict, error dict {"type": "error", ...}, or None on error Raises: AgentGeneratorNotConfiguredError: If the external service is not configured. """ _check_service_configured() logger.info("Calling external Agent Generator service for generate_agent") - result = await generate_agent_external(instructions) + result = await generate_agent_external( + dict(instructions), _to_dict_list(library_agents) + ) if result: - # Ensure required fields + if isinstance(result, dict) and result.get("type") == "error": + return result if "id" not in result: result["id"] = str(uuid.uuid4()) if "version" not in result: @@ -83,6 +579,12 @@ async def generate_agent(instructions: dict[str, Any]) -> dict[str, Any] | None: return result +class AgentJsonValidationError(Exception): + """Raised when agent JSON is invalid or missing required fields.""" + + pass + + def json_to_graph(agent_json: dict[str, Any]) -> Graph: """Convert agent JSON dict to Graph model. @@ -91,25 +593,55 @@ def json_to_graph(agent_json: dict[str, Any]) -> Graph: Returns: Graph ready for saving + + Raises: + AgentJsonValidationError: If required fields are missing from nodes or links """ nodes = [] - for n in agent_json.get("nodes", []): + for idx, n in enumerate(agent_json.get("nodes", [])): + block_id = n.get("block_id") + if not block_id: + node_id = n.get("id", f"index_{idx}") + raise AgentJsonValidationError( + f"Node '{node_id}' is missing required field 'block_id'" + ) node = Node( id=n.get("id", str(uuid.uuid4())), - block_id=n["block_id"], + block_id=block_id, input_default=n.get("input_default", {}), metadata=n.get("metadata", {}), ) nodes.append(node) links = [] - for link_data in agent_json.get("links", []): + for idx, link_data in enumerate(agent_json.get("links", [])): + source_id = link_data.get("source_id") + sink_id = link_data.get("sink_id") + source_name = link_data.get("source_name") + sink_name = link_data.get("sink_name") + + missing_fields = [] + if not source_id: + missing_fields.append("source_id") + if not sink_id: + missing_fields.append("sink_id") + if not source_name: + missing_fields.append("source_name") + if not sink_name: + missing_fields.append("sink_name") + + if missing_fields: + link_id = link_data.get("id", f"index_{idx}") + raise AgentJsonValidationError( + f"Link '{link_id}' is missing required fields: {', '.join(missing_fields)}" + ) + 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"], + source_id=source_id, + sink_id=sink_id, + source_name=source_name, + sink_name=sink_name, is_static=link_data.get("is_static", False), ) links.append(link) @@ -130,22 +662,40 @@ def _reassign_node_ids(graph: Graph) -> None: 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 + link.id = str(uuid.uuid4()) 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] +def _populate_agent_executor_user_ids(agent_json: dict[str, Any], user_id: str) -> None: + """Populate user_id in AgentExecutorBlock nodes. + + The external agent generator creates AgentExecutorBlock nodes with empty user_id. + This function fills in the actual user_id so sub-agents run with correct permissions. + + Args: + agent_json: Agent JSON dict (modified in place) + user_id: User ID to set + """ + for node in agent_json.get("nodes", []): + if node.get("block_id") == AGENT_EXECUTOR_BLOCK_ID: + input_default = node.get("input_default") or {} + if not input_default.get("user_id"): + input_default["user_id"] = user_id + node["input_default"] = input_default + logger.debug( + f"Set user_id for AgentExecutorBlock node {node.get('id')}" + ) + + async def save_agent_to_library( agent_json: dict[str, Any], user_id: str, is_update: bool = False ) -> tuple[Graph, Any]: @@ -159,33 +709,27 @@ async def save_agent_to_library( Returns: Tuple of (created Graph, LibraryAgent) """ - from backend.data.graph import get_graph_all_versions + # Populate user_id in AgentExecutorBlock nodes before conversion + _populate_agent_executor_user_ids(agent_json, user_id) 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, @@ -197,25 +741,31 @@ async def save_agent_to_library( async def get_agent_as_json( - graph_id: str, user_id: str | None + agent_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 + agent_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 + graph = await get_graph(agent_id, version=None, user_id=user_id) + + if not graph and user_id: + try: + library_agent = await library_db.get_library_agent(agent_id, user_id) + graph = await get_graph( + library_agent.graph_id, version=None, user_id=user_id + ) + except NotFoundError: + pass - # 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( @@ -253,7 +803,9 @@ async def get_agent_as_json( async def generate_agent_patch( - update_request: str, current_agent: dict[str, Any] + update_request: str, + current_agent: dict[str, Any], + library_agents: list[AgentSummary] | None = None, ) -> dict[str, Any] | None: """Update an existing agent using natural language. @@ -265,13 +817,17 @@ async def generate_agent_patch( Args: update_request: Natural language description of changes current_agent: Current agent JSON + library_agents: User's library agents available for sub-agent composition Returns: - Updated agent JSON, clarifying questions dict, or None on error + Updated agent JSON, clarifying questions dict {"type": "clarifying_questions", ...}, + error dict {"type": "error", ...}, or None on unexpected error Raises: AgentGeneratorNotConfiguredError: If the external service is not configured. """ _check_service_configured() logger.info("Calling external Agent Generator service for generate_agent_patch") - return await generate_agent_patch_external(update_request, current_agent) + return await generate_agent_patch_external( + update_request, current_agent, _to_dict_list(library_agents) + ) diff --git a/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/errors.py b/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/errors.py new file mode 100644 index 0000000000..282d8cf9aa --- /dev/null +++ b/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/errors.py @@ -0,0 +1,95 @@ +"""Error handling utilities for agent generator.""" + +import re + + +def _sanitize_error_details(details: str) -> str: + """Sanitize error details to remove sensitive information. + + Strips common patterns that could expose internal system info: + - File paths (Unix and Windows) + - Database connection strings + - URLs with credentials + - Stack trace internals + + Args: + details: Raw error details string + + Returns: + Sanitized error details safe for user display + """ + sanitized = re.sub( + r"/[a-zA-Z0-9_./\-]+\.(py|js|ts|json|yaml|yml)", "[path]", details + ) + sanitized = re.sub(r"[A-Z]:\\[a-zA-Z0-9_\\.\\-]+", "[path]", sanitized) + sanitized = re.sub( + r"(postgres|mysql|mongodb|redis)://[^\s]+", "[database_url]", sanitized + ) + sanitized = re.sub(r"https?://[^:]+:[^@]+@[^\s]+", "[url]", sanitized) + sanitized = re.sub(r", line \d+", "", sanitized) + sanitized = re.sub(r'File "[^"]+",?', "", sanitized) + + return sanitized.strip() + + +def get_user_message_for_error( + error_type: str, + operation: str = "process the request", + llm_parse_message: str | None = None, + validation_message: str | None = None, + error_details: str | None = None, +) -> str: + """Get a user-friendly error message based on error type. + + This function maps internal error types to user-friendly messages, + providing a consistent experience across different agent operations. + + Args: + error_type: The error type from the external service + (e.g., "llm_parse_error", "timeout", "rate_limit") + operation: Description of what operation failed, used in the default + message (e.g., "analyze the goal", "generate the agent") + llm_parse_message: Custom message for llm_parse_error type + validation_message: Custom message for validation_error type + error_details: Optional additional details about the error + + Returns: + User-friendly error message suitable for display to the user + """ + base_message = "" + + if error_type == "llm_parse_error": + base_message = ( + llm_parse_message + or "The AI had trouble processing this request. Please try again." + ) + elif error_type == "validation_error": + base_message = ( + validation_message + or "The generated agent failed validation. " + "This usually happens when the agent structure doesn't match " + "what the platform expects. Please try simplifying your goal " + "or breaking it into smaller parts." + ) + elif error_type == "patch_error": + base_message = ( + "Failed to apply the changes. The modification couldn't be " + "validated. Please try a different approach or simplify the change." + ) + elif error_type in ("timeout", "llm_timeout"): + base_message = ( + "The request took too long to process. This can happen with " + "complex agents. Please try again or simplify your goal." + ) + elif error_type in ("rate_limit", "llm_rate_limit"): + base_message = "The service is currently busy. Please try again in a moment." + else: + base_message = f"Failed to {operation}. Please try again." + + if error_details: + details = _sanitize_error_details(error_details) + if len(details) > 200: + details = details[:200] + "..." + base_message += f"\n\nTechnical details: {details}" + + return base_message diff --git a/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/service.py b/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/service.py index a4d2f1af15..c9c960d1ae 100644 --- a/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/service.py +++ b/autogpt_platform/backend/backend/api/features/chat/tools/agent_generator/service.py @@ -14,6 +14,70 @@ from backend.util.settings import Settings logger = logging.getLogger(__name__) + +def _create_error_response( + error_message: str, + error_type: str = "unknown", + details: dict[str, Any] | None = None, +) -> dict[str, Any]: + """Create a standardized error response dict. + + Args: + error_message: Human-readable error message + error_type: Machine-readable error type + details: Optional additional error details + + Returns: + Error dict with type="error" and error details + """ + response: dict[str, Any] = { + "type": "error", + "error": error_message, + "error_type": error_type, + } + if details: + response["details"] = details + return response + + +def _classify_http_error(e: httpx.HTTPStatusError) -> tuple[str, str]: + """Classify an HTTP error into error_type and message. + + Args: + e: The HTTP status error + + Returns: + Tuple of (error_type, error_message) + """ + status = e.response.status_code + if status == 429: + return "rate_limit", f"Agent Generator rate limited: {e}" + elif status == 503: + return "service_unavailable", f"Agent Generator unavailable: {e}" + elif status == 504 or status == 408: + return "timeout", f"Agent Generator timed out: {e}" + else: + return "http_error", f"HTTP error calling Agent Generator: {e}" + + +def _classify_request_error(e: httpx.RequestError) -> tuple[str, str]: + """Classify a request error into error_type and message. + + Args: + e: The request error + + Returns: + Tuple of (error_type, error_message) + """ + error_str = str(e).lower() + if "timeout" in error_str or "timed out" in error_str: + return "timeout", f"Agent Generator request timed out: {e}" + elif "connect" in error_str: + return "connection_error", f"Could not connect to Agent Generator: {e}" + else: + return "request_error", f"Request error calling Agent Generator: {e}" + + _client: httpx.AsyncClient | None = None _settings: Settings | None = None @@ -53,13 +117,16 @@ def _get_client() -> httpx.AsyncClient: async def decompose_goal_external( - description: str, context: str = "" + description: str, + context: str = "", + library_agents: list[dict[str, Any]] | None = None, ) -> dict[str, Any] | None: """Call the external service to decompose a goal. Args: description: Natural language goal description context: Additional context (e.g., answers to previous questions) + library_agents: User's library agents available for sub-agent composition Returns: Dict with either: @@ -67,15 +134,17 @@ async def decompose_goal_external( - {"type": "instructions", "steps": [...]} - {"type": "unachievable_goal", ...} - {"type": "vague_goal", ...} - Or None on error + - {"type": "error", "error": "...", "error_type": "..."} on error + Or None on unexpected error """ client = _get_client() - # Build the request payload - payload: dict[str, Any] = {"description": description} if context: - # The external service uses user_instruction for additional context - payload["user_instruction"] = context + description = f"{description}\n\nAdditional context from user:\n{context}" + + payload: dict[str, Any] = {"description": description} + if library_agents: + payload["library_agents"] = library_agents try: response = await client.post("/api/decompose-description", json=payload) @@ -83,8 +152,13 @@ async def decompose_goal_external( data = response.json() if not data.get("success"): - logger.error(f"External service returned error: {data.get('error')}") - return None + error_msg = data.get("error", "Unknown error from Agent Generator") + error_type = data.get("error_type", "unknown") + logger.error( + f"Agent Generator decomposition failed: {error_msg} " + f"(type: {error_type})" + ) + return _create_error_response(error_msg, error_type) # Map the response to the expected format response_type = data.get("type") @@ -106,88 +180,120 @@ async def decompose_goal_external( "type": "vague_goal", "suggested_goal": data.get("suggested_goal"), } + elif response_type == "error": + # Pass through error from the service + return _create_error_response( + data.get("error", "Unknown error"), + data.get("error_type", "unknown"), + ) else: logger.error( f"Unknown response type from external service: {response_type}" ) - return None + return _create_error_response( + f"Unknown response type from Agent Generator: {response_type}", + "invalid_response", + ) except httpx.HTTPStatusError as e: - logger.error(f"HTTP error calling external agent generator: {e}") - return None + error_type, error_msg = _classify_http_error(e) + logger.error(error_msg) + return _create_error_response(error_msg, error_type) except httpx.RequestError as e: - logger.error(f"Request error calling external agent generator: {e}") - return None + error_type, error_msg = _classify_request_error(e) + logger.error(error_msg) + return _create_error_response(error_msg, error_type) except Exception as e: - logger.error(f"Unexpected error calling external agent generator: {e}") - return None + error_msg = f"Unexpected error calling Agent Generator: {e}" + logger.error(error_msg) + return _create_error_response(error_msg, "unexpected_error") async def generate_agent_external( - instructions: dict[str, Any] + instructions: dict[str, Any], + library_agents: list[dict[str, Any]] | None = None, ) -> dict[str, Any] | None: """Call the external service to generate an agent from instructions. Args: instructions: Structured instructions from decompose_goal + library_agents: User's library agents available for sub-agent composition Returns: - Agent JSON dict or None on error + Agent JSON dict on success, or error dict {"type": "error", ...} on error """ client = _get_client() + payload: dict[str, Any] = {"instructions": instructions} + if library_agents: + payload["library_agents"] = library_agents + try: - response = await client.post( - "/api/generate-agent", json={"instructions": instructions} - ) + response = await client.post("/api/generate-agent", json=payload) response.raise_for_status() data = response.json() if not data.get("success"): - logger.error(f"External service returned error: {data.get('error')}") - return None + error_msg = data.get("error", "Unknown error from Agent Generator") + error_type = data.get("error_type", "unknown") + logger.error( + f"Agent Generator generation failed: {error_msg} (type: {error_type})" + ) + return _create_error_response(error_msg, error_type) return data.get("agent_json") except httpx.HTTPStatusError as e: - logger.error(f"HTTP error calling external agent generator: {e}") - return None + error_type, error_msg = _classify_http_error(e) + logger.error(error_msg) + return _create_error_response(error_msg, error_type) except httpx.RequestError as e: - logger.error(f"Request error calling external agent generator: {e}") - return None + error_type, error_msg = _classify_request_error(e) + logger.error(error_msg) + return _create_error_response(error_msg, error_type) except Exception as e: - logger.error(f"Unexpected error calling external agent generator: {e}") - return None + error_msg = f"Unexpected error calling Agent Generator: {e}" + logger.error(error_msg) + return _create_error_response(error_msg, "unexpected_error") async def generate_agent_patch_external( - update_request: str, current_agent: dict[str, Any] + update_request: str, + current_agent: dict[str, Any], + library_agents: list[dict[str, Any]] | None = None, ) -> dict[str, Any] | None: """Call the external service to generate a patch for an existing agent. Args: update_request: Natural language description of changes current_agent: Current agent JSON + library_agents: User's library agents available for sub-agent composition Returns: - Updated agent JSON, clarifying questions dict, or None on error + Updated agent JSON, clarifying questions dict, or error dict on error """ client = _get_client() + payload: dict[str, Any] = { + "update_request": update_request, + "current_agent_json": current_agent, + } + if library_agents: + payload["library_agents"] = library_agents + try: - response = await client.post( - "/api/update-agent", - json={ - "update_request": update_request, - "current_agent_json": current_agent, - }, - ) + response = await client.post("/api/update-agent", json=payload) response.raise_for_status() data = response.json() if not data.get("success"): - logger.error(f"External service returned error: {data.get('error')}") - return None + error_msg = data.get("error", "Unknown error from Agent Generator") + error_type = data.get("error_type", "unknown") + logger.error( + f"Agent Generator patch generation failed: {error_msg} " + f"(type: {error_type})" + ) + return _create_error_response(error_msg, error_type) # Check if it's clarifying questions if data.get("type") == "clarifying_questions": @@ -196,18 +302,28 @@ async def generate_agent_patch_external( "questions": data.get("questions", []), } + # Check if it's an error passed through + if data.get("type") == "error": + return _create_error_response( + data.get("error", "Unknown error"), + data.get("error_type", "unknown"), + ) + # Otherwise return the updated agent JSON return data.get("agent_json") except httpx.HTTPStatusError as e: - logger.error(f"HTTP error calling external agent generator: {e}") - return None + error_type, error_msg = _classify_http_error(e) + logger.error(error_msg) + return _create_error_response(error_msg, error_type) except httpx.RequestError as e: - logger.error(f"Request error calling external agent generator: {e}") - return None + error_type, error_msg = _classify_request_error(e) + logger.error(error_msg) + return _create_error_response(error_msg, error_type) except Exception as e: - logger.error(f"Unexpected error calling external agent generator: {e}") - return None + error_msg = f"Unexpected error calling Agent Generator: {e}" + logger.error(error_msg) + return _create_error_response(error_msg, "unexpected_error") async def get_blocks_external() -> list[dict[str, Any]] | None: diff --git a/autogpt_platform/backend/backend/api/features/chat/tools/agent_search.py b/autogpt_platform/backend/backend/api/features/chat/tools/agent_search.py index 5fa74ba04e..62d59c470e 100644 --- a/autogpt_platform/backend/backend/api/features/chat/tools/agent_search.py +++ b/autogpt_platform/backend/backend/api/features/chat/tools/agent_search.py @@ -1,6 +1,7 @@ """Shared agent search functionality for find_agent and find_library_agent tools.""" import logging +import re from typing import Literal from backend.api.features.library import db as library_db @@ -19,6 +20,85 @@ logger = logging.getLogger(__name__) SearchSource = Literal["marketplace", "library"] +_UUID_PATTERN = 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}$", + re.IGNORECASE, +) + + +def _is_uuid(text: str) -> bool: + """Check if text is a valid UUID v4.""" + return bool(_UUID_PATTERN.match(text.strip())) + + +async def _get_library_agent_by_id(user_id: str, agent_id: str) -> AgentInfo | None: + """Fetch a library agent by ID (library agent ID or graph_id). + + Tries multiple lookup strategies: + 1. First by graph_id (AgentGraph primary key) + 2. Then by library agent ID (LibraryAgent primary key) + + Args: + user_id: The user ID + agent_id: The ID to look up (can be graph_id or library agent ID) + + Returns: + AgentInfo if found, None otherwise + """ + try: + agent = await library_db.get_library_agent_by_graph_id(user_id, agent_id) + if agent: + logger.debug(f"Found library agent by graph_id: {agent.name}") + return 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: + raise + except Exception as e: + logger.warning( + f"Could not fetch library agent by graph_id {agent_id}: {e}", + exc_info=True, + ) + + try: + agent = await library_db.get_library_agent(agent_id, user_id) + if agent: + logger.debug(f"Found library agent by library_id: {agent.name}") + return 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 NotFoundError: + logger.debug(f"Library agent not found by library_id: {agent_id}") + except DatabaseError: + raise + except Exception as e: + logger.warning( + f"Could not fetch library agent by library_id {agent_id}: {e}", + exc_info=True, + ) + + return None + async def search_agents( query: str, @@ -69,29 +149,37 @@ async def search_agents( 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, - ) + else: + if _is_uuid(query): + logger.info(f"Query looks like UUID, trying direct lookup: {query}") + agent = await _get_library_agent_by_id(user_id, query) # type: ignore[arg-type] + if agent: + agents.append(agent) + logger.info(f"Found agent by direct ID lookup: {agent.name}") + + if not agents: + 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 diff --git a/autogpt_platform/backend/backend/api/features/chat/tools/create_agent.py b/autogpt_platform/backend/backend/api/features/chat/tools/create_agent.py index 6b3784e323..adb2c78fce 100644 --- a/autogpt_platform/backend/backend/api/features/chat/tools/create_agent.py +++ b/autogpt_platform/backend/backend/api/features/chat/tools/create_agent.py @@ -8,7 +8,10 @@ from backend.api.features.chat.model import ChatSession from .agent_generator import ( AgentGeneratorNotConfiguredError, decompose_goal, + enrich_library_agents_from_steps, generate_agent, + get_all_relevant_agents_for_generation, + get_user_message_for_error, save_agent_to_library, ) from .base import BaseTool @@ -102,9 +105,24 @@ class CreateAgentTool(BaseTool): session_id=session_id, ) - # Step 1: Decompose goal into steps + library_agents = None + if user_id: + try: + library_agents = await get_all_relevant_agents_for_generation( + user_id=user_id, + search_query=description, + include_marketplace=True, + ) + logger.debug( + f"Found {len(library_agents)} relevant agents for sub-agent composition" + ) + except Exception as e: + logger.warning(f"Failed to fetch library agents: {e}") + try: - decomposition_result = await decompose_goal(description, context) + decomposition_result = await decompose_goal( + description, context, library_agents + ) except AgentGeneratorNotConfiguredError: return ErrorResponse( message=( @@ -117,15 +135,31 @@ class CreateAgentTool(BaseTool): if decomposition_result is None: return ErrorResponse( - message="Failed to analyze the goal. The agent generation service may be unavailable or timed out. Please try again.", + message="Failed to analyze the goal. The agent generation service may be unavailable. Please try again.", error="decomposition_failed", - details={ - "description": description[:100] - }, # Include context for debugging + details={"description": description[:100]}, + session_id=session_id, + ) + + if decomposition_result.get("type") == "error": + error_msg = decomposition_result.get("error", "Unknown error") + error_type = decomposition_result.get("error_type", "unknown") + user_message = get_user_message_for_error( + error_type, + operation="analyze the goal", + llm_parse_message="The AI had trouble understanding this request. Please try rephrasing your goal.", + ) + return ErrorResponse( + message=user_message, + error=f"decomposition_failed:{error_type}", + details={ + "description": description[:100], + "service_error": error_msg, + "error_type": error_type, + }, session_id=session_id, ) - # Check if LLM returned clarifying questions if decomposition_result.get("type") == "clarifying_questions": questions = decomposition_result.get("questions", []) return ClarificationNeededResponse( @@ -144,7 +178,6 @@ class CreateAgentTool(BaseTool): 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", "") @@ -171,9 +204,22 @@ class CreateAgentTool(BaseTool): session_id=session_id, ) - # Step 2: Generate agent JSON (external service handles fixing and validation) + if user_id and library_agents is not None: + try: + library_agents = await enrich_library_agents_from_steps( + user_id=user_id, + decomposition_result=decomposition_result, + existing_agents=library_agents, + include_marketplace=True, + ) + logger.debug( + f"After enrichment: {len(library_agents)} total agents for sub-agent composition" + ) + except Exception as e: + logger.warning(f"Failed to enrich library agents from steps: {e}") + try: - agent_json = await generate_agent(decomposition_result) + agent_json = await generate_agent(decomposition_result, library_agents) except AgentGeneratorNotConfiguredError: return ErrorResponse( message=( @@ -186,11 +232,34 @@ class CreateAgentTool(BaseTool): if agent_json is None: return ErrorResponse( - message="Failed to generate the agent. The agent generation service may be unavailable or timed out. Please try again.", + message="Failed to generate the agent. The agent generation service may be unavailable. Please try again.", error="generation_failed", + details={"description": description[:100]}, + session_id=session_id, + ) + + if isinstance(agent_json, dict) and agent_json.get("type") == "error": + error_msg = agent_json.get("error", "Unknown error") + error_type = agent_json.get("error_type", "unknown") + user_message = get_user_message_for_error( + error_type, + operation="generate the agent", + llm_parse_message="The AI had trouble generating the agent. Please try again or simplify your goal.", + validation_message=( + "I wasn't able to create a valid agent for this request. " + "The generated workflow had some structural issues. " + "Please try simplifying your goal or breaking it into smaller steps." + ), + error_details=error_msg, + ) + return ErrorResponse( + message=user_message, + error=f"generation_failed:{error_type}", details={ - "description": description[:100] - }, # Include context for debugging + "description": description[:100], + "service_error": error_msg, + "error_type": error_type, + }, session_id=session_id, ) @@ -199,7 +268,6 @@ class CreateAgentTool(BaseTool): node_count = len(agent_json.get("nodes", [])) link_count = len(agent_json.get("links", [])) - # Step 3: Preview or save if not save: return AgentPreviewResponse( message=( @@ -214,7 +282,6 @@ class CreateAgentTool(BaseTool): session_id=session_id, ) - # Save to library if not user_id: return ErrorResponse( message="You must be logged in to save agents.", @@ -232,7 +299,7 @@ class CreateAgentTool(BaseTool): agent_id=created_graph.id, agent_name=created_graph.name, library_agent_id=library_agent.id, - library_agent_link=f"/library/{library_agent.id}", + library_agent_link=f"/library/agents/{library_agent.id}", agent_page_link=f"/build?flowID={created_graph.id}", session_id=session_id, ) diff --git a/autogpt_platform/backend/backend/api/features/chat/tools/edit_agent.py b/autogpt_platform/backend/backend/api/features/chat/tools/edit_agent.py index 7c4da8ad43..2c2c48226b 100644 --- a/autogpt_platform/backend/backend/api/features/chat/tools/edit_agent.py +++ b/autogpt_platform/backend/backend/api/features/chat/tools/edit_agent.py @@ -9,6 +9,8 @@ from .agent_generator import ( AgentGeneratorNotConfiguredError, generate_agent_patch, get_agent_as_json, + get_all_relevant_agents_for_generation, + get_user_message_for_error, save_agent_to_library, ) from .base import BaseTool @@ -116,7 +118,6 @@ class EditAgentTool(BaseTool): 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: @@ -126,14 +127,30 @@ class EditAgentTool(BaseTool): session_id=session_id, ) - # Build the update request with context + library_agents = None + if user_id: + try: + graph_id = current_agent.get("id") + library_agents = await get_all_relevant_agents_for_generation( + user_id=user_id, + search_query=changes, + exclude_graph_id=graph_id, + include_marketplace=True, + ) + logger.debug( + f"Found {len(library_agents)} relevant agents for sub-agent composition" + ) + except Exception as e: + logger.warning(f"Failed to fetch library agents: {e}") + update_request = changes if context: update_request = f"{changes}\n\nAdditional context:\n{context}" - # Step 2: Generate updated agent (external service handles fixing and validation) try: - result = await generate_agent_patch(update_request, current_agent) + result = await generate_agent_patch( + update_request, current_agent, library_agents + ) except AgentGeneratorNotConfiguredError: return ErrorResponse( message=( @@ -152,7 +169,28 @@ class EditAgentTool(BaseTool): session_id=session_id, ) - # Check if LLM returned clarifying questions + if isinstance(result, dict) and result.get("type") == "error": + error_msg = result.get("error", "Unknown error") + error_type = result.get("error_type", "unknown") + user_message = get_user_message_for_error( + error_type, + operation="generate the changes", + llm_parse_message="The AI had trouble generating the changes. Please try again or simplify your request.", + validation_message="The generated changes failed validation. Please try rephrasing your request.", + error_details=error_msg, + ) + return ErrorResponse( + message=user_message, + error=f"update_generation_failed:{error_type}", + details={ + "agent_id": agent_id, + "changes": changes[:100], + "service_error": error_msg, + "error_type": error_type, + }, + session_id=session_id, + ) + if result.get("type") == "clarifying_questions": questions = result.get("questions", []) return ClarificationNeededResponse( @@ -171,7 +209,6 @@ class EditAgentTool(BaseTool): session_id=session_id, ) - # Result is the updated agent JSON updated_agent = result agent_name = updated_agent.get("name", "Updated Agent") @@ -179,7 +216,6 @@ class EditAgentTool(BaseTool): node_count = len(updated_agent.get("nodes", [])) link_count = len(updated_agent.get("links", [])) - # Step 3: Preview or save if not save: return AgentPreviewResponse( message=( @@ -195,7 +231,6 @@ class EditAgentTool(BaseTool): 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.", @@ -213,7 +248,7 @@ class EditAgentTool(BaseTool): agent_id=created_graph.id, agent_name=created_graph.name, library_agent_id=library_agent.id, - library_agent_link=f"/library/{library_agent.id}", + library_agent_link=f"/library/agents/{library_agent.id}", agent_page_link=f"/build?flowID={created_graph.id}", session_id=session_id, ) diff --git a/autogpt_platform/backend/backend/api/features/chat/tools/utils.py b/autogpt_platform/backend/backend/api/features/chat/tools/utils.py index a2ac91dc65..0046d0b249 100644 --- a/autogpt_platform/backend/backend/api/features/chat/tools/utils.py +++ b/autogpt_platform/backend/backend/api/features/chat/tools/utils.py @@ -8,7 +8,7 @@ from backend.api.features.library import model as library_model from backend.api.features.store import db as store_db from backend.data import graph as graph_db from backend.data.graph import GraphModel -from backend.data.model import CredentialsFieldInfo, CredentialsMetaInput +from backend.data.model import Credentials, CredentialsFieldInfo, CredentialsMetaInput from backend.integrations.creds_manager import IntegrationCredentialsManager from backend.util.exceptions import NotFoundError @@ -266,13 +266,14 @@ async def match_user_credentials_to_graph( credential_requirements, _node_fields, ) in aggregated_creds.items(): - # Find first matching credential by provider and type + # Find first matching credential by provider, type, and scopes matching_cred = next( ( cred for cred in available_creds if cred.provider in credential_requirements.provider and cred.type in credential_requirements.supported_types + and _credential_has_required_scopes(cred, credential_requirements) ), None, ) @@ -296,10 +297,17 @@ async def match_user_credentials_to_graph( f"{credential_field_name} (validation failed: {e})" ) else: + # Build a helpful error message including scope requirements + error_parts = [ + f"provider in {list(credential_requirements.provider)}", + f"type in {list(credential_requirements.supported_types)}", + ] + if credential_requirements.required_scopes: + error_parts.append( + f"scopes including {list(credential_requirements.required_scopes)}" + ) missing_creds.append( - f"{credential_field_name} " - f"(requires provider in {list(credential_requirements.provider)}, " - f"type in {list(credential_requirements.supported_types)})" + f"{credential_field_name} (requires {', '.join(error_parts)})" ) logger.info( @@ -309,6 +317,28 @@ async def match_user_credentials_to_graph( return graph_credentials_inputs, missing_creds +def _credential_has_required_scopes( + credential: Credentials, + requirements: CredentialsFieldInfo, +) -> bool: + """ + Check if a credential has all the scopes required by the block. + + For OAuth2 credentials, verifies that the credential's scopes are a superset + of the required scopes. For other credential types, returns True (no scope check). + """ + # Only OAuth2 credentials have scopes to check + if credential.type != "oauth2": + return True + + # If no scopes are required, any credential matches + if not requirements.required_scopes: + return True + + # Check that credential scopes are a superset of required scopes + return set(credential.scopes).issuperset(requirements.required_scopes) + + async def check_user_has_required_credentials( user_id: str, required_credentials: list[CredentialsMetaInput], diff --git a/autogpt_platform/backend/backend/api/features/library/db.py b/autogpt_platform/backend/backend/api/features/library/db.py index 872fe66b28..394f959953 100644 --- a/autogpt_platform/backend/backend/api/features/library/db.py +++ b/autogpt_platform/backend/backend/api/features/library/db.py @@ -39,6 +39,7 @@ async def list_library_agents( sort_by: library_model.LibraryAgentSort = library_model.LibraryAgentSort.UPDATED_AT, page: int = 1, page_size: int = 50, + include_executions: bool = False, ) -> library_model.LibraryAgentResponse: """ Retrieves a paginated list of LibraryAgent records for a given user. @@ -49,6 +50,9 @@ async def list_library_agents( sort_by: Sorting field (createdAt, updatedAt, isFavorite, isCreatedByUser). page: Current page (1-indexed). page_size: Number of items per page. + include_executions: Whether to include execution data for status calculation. + Defaults to False for performance (UI fetches status separately). + Set to True when accurate status/metrics are needed (e.g., agent generator). Returns: A LibraryAgentResponse containing the list of agents and pagination details. @@ -76,7 +80,6 @@ async def list_library_agents( "isArchived": False, } - # Build search filter if applicable if search_term: where_clause["OR"] = [ { @@ -93,7 +96,6 @@ async def list_library_agents( }, ] - # Determine sorting order_by: prisma.types.LibraryAgentOrderByInput | None = None if sort_by == library_model.LibraryAgentSort.CREATED_AT: @@ -105,7 +107,7 @@ async def list_library_agents( library_agents = await prisma.models.LibraryAgent.prisma().find_many( where=where_clause, include=library_agent_include( - user_id, include_nodes=False, include_executions=False + user_id, include_nodes=False, include_executions=include_executions ), order=order_by, skip=(page - 1) * page_size, diff --git a/autogpt_platform/backend/backend/api/features/library/model.py b/autogpt_platform/backend/backend/api/features/library/model.py index 14d7c7be81..c6bc0e0427 100644 --- a/autogpt_platform/backend/backend/api/features/library/model.py +++ b/autogpt_platform/backend/backend/api/features/library/model.py @@ -9,6 +9,7 @@ import pydantic from backend.data.block import BlockInput from backend.data.graph import GraphModel, GraphSettings, GraphTriggerInfo from backend.data.model import CredentialsMetaInput, is_credentials_field_name +from backend.util.json import loads as json_loads from backend.util.models import Pagination if TYPE_CHECKING: @@ -16,10 +17,10 @@ if TYPE_CHECKING: class LibraryAgentStatus(str, Enum): - COMPLETED = "COMPLETED" # All runs completed - HEALTHY = "HEALTHY" # Agent is running (not all runs have completed) - WAITING = "WAITING" # Agent is queued or waiting to start - ERROR = "ERROR" # Agent is in an error state + COMPLETED = "COMPLETED" + HEALTHY = "HEALTHY" + WAITING = "WAITING" + ERROR = "ERROR" class MarketplaceListingCreator(pydantic.BaseModel): @@ -39,6 +40,30 @@ class MarketplaceListing(pydantic.BaseModel): creator: MarketplaceListingCreator +class RecentExecution(pydantic.BaseModel): + """Summary of a recent execution for quality assessment. + + Used by the LLM to understand the agent's recent performance with specific examples + rather than just aggregate statistics. + """ + + status: str + correctness_score: float | None = None + activity_summary: str | None = None + + +def _parse_settings(settings: dict | str | None) -> GraphSettings: + """Parse settings from database, handling both dict and string formats.""" + if settings is None: + return GraphSettings() + try: + if isinstance(settings, str): + settings = json_loads(settings) + return GraphSettings.model_validate(settings) + except Exception: + return GraphSettings() + + class LibraryAgent(pydantic.BaseModel): """ Represents an agent in the library, including metadata for display and @@ -48,7 +73,7 @@ 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 + owner_user_id: str image_url: str | None @@ -64,7 +89,7 @@ class LibraryAgent(pydantic.BaseModel): description: str instructions: str | None = None - input_schema: dict[str, Any] # Should be BlockIOObjectSubSchema in frontend + input_schema: dict[str, Any] output_schema: dict[str, Any] credentials_input_schema: dict[str, Any] | None = pydantic.Field( description="Input schema for credentials required by the agent", @@ -81,25 +106,19 @@ class LibraryAgent(pydantic.BaseModel): ) trigger_setup_info: Optional[GraphTriggerInfo] = None - # Indicates whether there's a new output (based on recent runs) new_output: bool - - # Whether the user can access the underlying graph + execution_count: int = 0 + success_rate: float | None = None + avg_correctness_score: float | None = None + recent_executions: list[RecentExecution] = pydantic.Field( + default_factory=list, + description="List of recent executions with status, score, and summary", + ) can_access_graph: bool - - # Indicates if this agent is the latest version is_latest_version: bool - - # Whether the agent is marked as favorite by the user is_favorite: bool - - # Recommended schedule cron (from marketplace agents) recommended_schedule_cron: str | None = None - - # User-specific settings for this library agent settings: GraphSettings = pydantic.Field(default_factory=GraphSettings) - - # Marketplace listing information if the agent has been published marketplace_listing: Optional["MarketplaceListing"] = None @staticmethod @@ -123,7 +142,6 @@ class LibraryAgent(pydantic.BaseModel): agent_updated_at = agent.AgentGraph.updatedAt lib_agent_updated_at = agent.updatedAt - # Compute updated_at as the latest between library agent and graph updated_at = ( max(agent_updated_at, lib_agent_updated_at) if agent_updated_at @@ -136,7 +154,6 @@ class LibraryAgent(pydantic.BaseModel): creator_name = agent.Creator.name or "Unknown" creator_image_url = agent.Creator.avatarUrl or "" - # Logic to calculate status and new_output week_ago = datetime.datetime.now(datetime.timezone.utc) - datetime.timedelta( days=7 ) @@ -145,13 +162,55 @@ class LibraryAgent(pydantic.BaseModel): status = status_result.status new_output = status_result.new_output - # Check if user can access the graph - can_access_graph = agent.AgentGraph.userId == agent.userId + execution_count = len(executions) + success_rate: float | None = None + avg_correctness_score: float | None = None + if execution_count > 0: + success_count = sum( + 1 + for e in executions + if e.executionStatus == prisma.enums.AgentExecutionStatus.COMPLETED + ) + success_rate = (success_count / execution_count) * 100 - # Hard-coded to True until a method to check is implemented + correctness_scores = [] + for e in executions: + if e.stats and isinstance(e.stats, dict): + score = e.stats.get("correctness_score") + if score is not None and isinstance(score, (int, float)): + correctness_scores.append(float(score)) + if correctness_scores: + avg_correctness_score = sum(correctness_scores) / len( + correctness_scores + ) + + recent_executions: list[RecentExecution] = [] + for e in executions: + exec_score: float | None = None + exec_summary: str | None = None + if e.stats and isinstance(e.stats, dict): + score = e.stats.get("correctness_score") + if score is not None and isinstance(score, (int, float)): + exec_score = float(score) + summary = e.stats.get("activity_status") + if summary is not None and isinstance(summary, str): + exec_summary = summary + exec_status = ( + e.executionStatus.value + if hasattr(e.executionStatus, "value") + else str(e.executionStatus) + ) + recent_executions.append( + RecentExecution( + status=exec_status, + correctness_score=exec_score, + activity_summary=exec_summary, + ) + ) + + can_access_graph = agent.AgentGraph.userId == agent.userId is_latest_version = True - # Build marketplace_listing if available marketplace_listing_data = None if store_listing and store_listing.ActiveVersion and profile: creator_data = MarketplaceListingCreator( @@ -190,11 +249,15 @@ class LibraryAgent(pydantic.BaseModel): has_sensitive_action=graph.has_sensitive_action, trigger_setup_info=graph.trigger_setup_info, new_output=new_output, + execution_count=execution_count, + success_rate=success_rate, + avg_correctness_score=avg_correctness_score, + recent_executions=recent_executions, can_access_graph=can_access_graph, is_latest_version=is_latest_version, is_favorite=agent.isFavorite, recommended_schedule_cron=agent.AgentGraph.recommendedScheduleCron, - settings=GraphSettings.model_validate(agent.settings), + settings=_parse_settings(agent.settings), marketplace_listing=marketplace_listing_data, ) @@ -220,18 +283,15 @@ def _calculate_agent_status( if not executions: return AgentStatusResult(status=LibraryAgentStatus.COMPLETED, new_output=False) - # Track how many times each execution status appears status_counts = {status: 0 for status in prisma.enums.AgentExecutionStatus} new_output = False for execution in executions: - # Check if there's a completed run more recent than `recent_threshold` if execution.createdAt >= recent_threshold: if execution.executionStatus == prisma.enums.AgentExecutionStatus.COMPLETED: new_output = True status_counts[execution.executionStatus] += 1 - # Determine the final status based on counts if status_counts[prisma.enums.AgentExecutionStatus.FAILED] > 0: return AgentStatusResult(status=LibraryAgentStatus.ERROR, new_output=new_output) elif status_counts[prisma.enums.AgentExecutionStatus.QUEUED] > 0: diff --git a/autogpt_platform/backend/backend/api/features/store/db.py b/autogpt_platform/backend/backend/api/features/store/db.py index 956fdfa7da..850a2bc3e9 100644 --- a/autogpt_platform/backend/backend/api/features/store/db.py +++ b/autogpt_platform/backend/backend/api/features/store/db.py @@ -112,6 +112,7 @@ async def get_store_agents( description=agent["description"], runs=agent["runs"], rating=agent["rating"], + agent_graph_id=agent.get("agentGraphId", ""), ) store_agents.append(store_agent) except Exception as e: @@ -170,6 +171,7 @@ async def get_store_agents( description=agent.description, runs=agent.runs, rating=agent.rating, + agent_graph_id=agent.agentGraphId, ) # Add to the list only if creation was successful store_agents.append(store_agent) diff --git a/autogpt_platform/backend/backend/api/features/store/hybrid_search.py b/autogpt_platform/backend/backend/api/features/store/hybrid_search.py index 8b0884bb24..e1b8f402c8 100644 --- a/autogpt_platform/backend/backend/api/features/store/hybrid_search.py +++ b/autogpt_platform/backend/backend/api/features/store/hybrid_search.py @@ -600,6 +600,7 @@ async def hybrid_search( sa.featured, sa.is_available, sa.updated_at, + sa."agentGraphId", -- Searchable text for BM25 reranking COALESCE(sa.agent_name, '') || ' ' || COALESCE(sa.sub_heading, '') || ' ' || COALESCE(sa.description, '') as searchable_text, -- Semantic score @@ -659,6 +660,7 @@ async def hybrid_search( featured, is_available, updated_at, + "agentGraphId", searchable_text, semantic_score, lexical_score, diff --git a/autogpt_platform/backend/backend/api/features/store/model.py b/autogpt_platform/backend/backend/api/features/store/model.py index a3310b96fc..d66b91807d 100644 --- a/autogpt_platform/backend/backend/api/features/store/model.py +++ b/autogpt_platform/backend/backend/api/features/store/model.py @@ -38,6 +38,7 @@ class StoreAgent(pydantic.BaseModel): description: str runs: int rating: float + agent_graph_id: str class StoreAgentsResponse(pydantic.BaseModel): diff --git a/autogpt_platform/backend/backend/api/features/store/model_test.py b/autogpt_platform/backend/backend/api/features/store/model_test.py index fd09a0cf77..c4109f4603 100644 --- a/autogpt_platform/backend/backend/api/features/store/model_test.py +++ b/autogpt_platform/backend/backend/api/features/store/model_test.py @@ -26,11 +26,13 @@ def test_store_agent(): description="Test description", runs=50, rating=4.5, + agent_graph_id="test-graph-id", ) assert agent.slug == "test-agent" assert agent.agent_name == "Test Agent" assert agent.runs == 50 assert agent.rating == 4.5 + assert agent.agent_graph_id == "test-graph-id" def test_store_agents_response(): @@ -46,6 +48,7 @@ def test_store_agents_response(): description="Test description", runs=50, rating=4.5, + agent_graph_id="test-graph-id", ) ], pagination=store_model.Pagination( diff --git a/autogpt_platform/backend/backend/api/features/store/routes_test.py b/autogpt_platform/backend/backend/api/features/store/routes_test.py index 36431c20ec..fcef3f845a 100644 --- a/autogpt_platform/backend/backend/api/features/store/routes_test.py +++ b/autogpt_platform/backend/backend/api/features/store/routes_test.py @@ -82,6 +82,7 @@ def test_get_agents_featured( description="Featured agent description", runs=100, rating=4.5, + agent_graph_id="test-graph-1", ) ], pagination=store_model.Pagination( @@ -127,6 +128,7 @@ def test_get_agents_by_creator( description="Creator agent description", runs=50, rating=4.0, + agent_graph_id="test-graph-2", ) ], pagination=store_model.Pagination( @@ -172,6 +174,7 @@ def test_get_agents_sorted( description="Top agent description", runs=1000, rating=5.0, + agent_graph_id="test-graph-3", ) ], pagination=store_model.Pagination( @@ -217,6 +220,7 @@ def test_get_agents_search( description="Specific search term description", runs=75, rating=4.2, + agent_graph_id="test-graph-search", ) ], pagination=store_model.Pagination( @@ -262,6 +266,7 @@ def test_get_agents_category( description="Category agent description", runs=60, rating=4.1, + agent_graph_id="test-graph-category", ) ], pagination=store_model.Pagination( @@ -306,6 +311,7 @@ def test_get_agents_pagination( description=f"Agent {i} description", runs=i * 10, rating=4.0, + agent_graph_id="test-graph-2", ) for i in range(5) ], diff --git a/autogpt_platform/backend/backend/api/features/store/test_cache_delete.py b/autogpt_platform/backend/backend/api/features/store/test_cache_delete.py index dd9be1f4ab..298c51d47c 100644 --- a/autogpt_platform/backend/backend/api/features/store/test_cache_delete.py +++ b/autogpt_platform/backend/backend/api/features/store/test_cache_delete.py @@ -33,6 +33,7 @@ class TestCacheDeletion: description="Test description", runs=100, rating=4.5, + agent_graph_id="test-graph-id", ) ], pagination=Pagination( diff --git a/autogpt_platform/backend/backend/blocks/llm.py b/autogpt_platform/backend/backend/blocks/llm.py index fdcd7f3568..732fb1354c 100644 --- a/autogpt_platform/backend/backend/blocks/llm.py +++ b/autogpt_platform/backend/backend/blocks/llm.py @@ -115,7 +115,6 @@ class LlmModel(str, Enum, metaclass=LlmModelMeta): CLAUDE_4_5_OPUS = "claude-opus-4-5-20251101" CLAUDE_4_5_SONNET = "claude-sonnet-4-5-20250929" CLAUDE_4_5_HAIKU = "claude-haiku-4-5-20251001" - CLAUDE_3_7_SONNET = "claude-3-7-sonnet-20250219" CLAUDE_3_HAIKU = "claude-3-haiku-20240307" # AI/ML API models AIML_API_QWEN2_5_72B = "Qwen/Qwen2.5-72B-Instruct-Turbo" @@ -280,9 +279,6 @@ MODEL_METADATA = { LlmModel.CLAUDE_4_5_HAIKU: ModelMetadata( "anthropic", 200000, 64000, "Claude Haiku 4.5", "Anthropic", "Anthropic", 2 ), # claude-haiku-4-5-20251001 - LlmModel.CLAUDE_3_7_SONNET: ModelMetadata( - "anthropic", 200000, 64000, "Claude 3.7 Sonnet", "Anthropic", "Anthropic", 2 - ), # claude-3-7-sonnet-20250219 LlmModel.CLAUDE_3_HAIKU: ModelMetadata( "anthropic", 200000, 4096, "Claude 3 Haiku", "Anthropic", "Anthropic", 1 ), # claude-3-haiku-20240307 diff --git a/autogpt_platform/backend/backend/blocks/stagehand/blocks.py b/autogpt_platform/backend/backend/blocks/stagehand/blocks.py index be1d736962..4d5d6bf4f3 100644 --- a/autogpt_platform/backend/backend/blocks/stagehand/blocks.py +++ b/autogpt_platform/backend/backend/blocks/stagehand/blocks.py @@ -83,7 +83,7 @@ class StagehandRecommendedLlmModel(str, Enum): GPT41_MINI = "gpt-4.1-mini-2025-04-14" # Anthropic - CLAUDE_3_7_SONNET = "claude-3-7-sonnet-20250219" + CLAUDE_4_5_SONNET = "claude-sonnet-4-5-20250929" @property def provider_name(self) -> str: @@ -137,7 +137,7 @@ class StagehandObserveBlock(Block): model: StagehandRecommendedLlmModel = SchemaField( title="LLM Model", description="LLM to use for Stagehand (provider is inferred)", - default=StagehandRecommendedLlmModel.CLAUDE_3_7_SONNET, + default=StagehandRecommendedLlmModel.CLAUDE_4_5_SONNET, advanced=False, ) model_credentials: AICredentials = AICredentialsField() @@ -230,7 +230,7 @@ class StagehandActBlock(Block): model: StagehandRecommendedLlmModel = SchemaField( title="LLM Model", description="LLM to use for Stagehand (provider is inferred)", - default=StagehandRecommendedLlmModel.CLAUDE_3_7_SONNET, + default=StagehandRecommendedLlmModel.CLAUDE_4_5_SONNET, advanced=False, ) model_credentials: AICredentials = AICredentialsField() @@ -330,7 +330,7 @@ class StagehandExtractBlock(Block): model: StagehandRecommendedLlmModel = SchemaField( title="LLM Model", description="LLM to use for Stagehand (provider is inferred)", - default=StagehandRecommendedLlmModel.CLAUDE_3_7_SONNET, + default=StagehandRecommendedLlmModel.CLAUDE_4_5_SONNET, advanced=False, ) model_credentials: AICredentials = AICredentialsField() diff --git a/autogpt_platform/backend/backend/data/block_cost_config.py b/autogpt_platform/backend/backend/data/block_cost_config.py index 1b54ae0942..f46cc726f0 100644 --- a/autogpt_platform/backend/backend/data/block_cost_config.py +++ b/autogpt_platform/backend/backend/data/block_cost_config.py @@ -81,7 +81,6 @@ MODEL_COST: dict[LlmModel, int] = { LlmModel.CLAUDE_4_5_HAIKU: 4, LlmModel.CLAUDE_4_5_OPUS: 14, LlmModel.CLAUDE_4_5_SONNET: 9, - LlmModel.CLAUDE_3_7_SONNET: 5, LlmModel.CLAUDE_3_HAIKU: 1, LlmModel.AIML_API_QWEN2_5_72B: 1, LlmModel.AIML_API_LLAMA3_1_70B: 1, diff --git a/autogpt_platform/backend/backend/data/graph.py b/autogpt_platform/backend/backend/data/graph.py index c1f38f81d5..ee6cd2e4b0 100644 --- a/autogpt_platform/backend/backend/data/graph.py +++ b/autogpt_platform/backend/backend/data/graph.py @@ -1028,6 +1028,39 @@ async def get_graph( return GraphModel.from_db(graph, for_export) +async def get_store_listed_graphs(*graph_ids: str) -> dict[str, GraphModel]: + """Batch-fetch multiple store-listed graphs by their IDs. + + Only returns graphs that have approved store listings (publicly available). + Does not require permission checks since store-listed graphs are public. + + Args: + *graph_ids: Variable number of graph IDs to fetch + + Returns: + Dict mapping graph_id to GraphModel for graphs with approved store listings + """ + if not graph_ids: + return {} + + store_listings = await StoreListingVersion.prisma().find_many( + where={ + "agentGraphId": {"in": list(graph_ids)}, + "submissionStatus": SubmissionStatus.APPROVED, + "isDeleted": False, + }, + include={"AgentGraph": {"include": AGENT_GRAPH_INCLUDE}}, + distinct=["agentGraphId"], + order={"agentGraphVersion": "desc"}, + ) + + return { + listing.agentGraphId: GraphModel.from_db(listing.AgentGraph) + for listing in store_listings + if listing.AgentGraph + } + + async def get_graph_as_admin( graph_id: str, version: int | None = None, diff --git a/autogpt_platform/backend/backend/data/model.py b/autogpt_platform/backend/backend/data/model.py index 2cc73f6b7b..331126fbd6 100644 --- a/autogpt_platform/backend/backend/data/model.py +++ b/autogpt_platform/backend/backend/data/model.py @@ -666,10 +666,16 @@ class CredentialsFieldInfo(BaseModel, Generic[CP, CT]): if not (self.discriminator and self.discriminator_mapping): return self + try: + provider = self.discriminator_mapping[discriminator_value] + except KeyError: + raise ValueError( + f"Model '{discriminator_value}' is not supported. " + "It may have been deprecated. Please update your agent configuration." + ) + return CredentialsFieldInfo( - credentials_provider=frozenset( - [self.discriminator_mapping[discriminator_value]] - ), + credentials_provider=frozenset([provider]), credentials_types=self.supported_types, credentials_scopes=self.required_scopes, discriminator=self.discriminator, diff --git a/autogpt_platform/backend/backend/integrations/webhooks/utils_test.py b/autogpt_platform/backend/backend/integrations/webhooks/utils_test.py new file mode 100644 index 0000000000..bc502a8e44 --- /dev/null +++ b/autogpt_platform/backend/backend/integrations/webhooks/utils_test.py @@ -0,0 +1,39 @@ +from urllib.parse import urlparse + +import fastapi +from fastapi.routing import APIRoute + +from backend.api.features.integrations.router import router as integrations_router +from backend.integrations.providers import ProviderName +from backend.integrations.webhooks import utils as webhooks_utils + + +def test_webhook_ingress_url_matches_route(monkeypatch) -> None: + app = fastapi.FastAPI() + app.include_router(integrations_router, prefix="/api/integrations") + + provider = ProviderName.GITHUB + webhook_id = "webhook_123" + base_url = "https://example.com" + + monkeypatch.setattr(webhooks_utils.app_config, "platform_base_url", base_url) + + route = next( + route + for route in integrations_router.routes + if isinstance(route, APIRoute) + and route.path == "/{provider}/webhooks/{webhook_id}/ingress" + and "POST" in route.methods + ) + expected_path = f"/api/integrations{route.path}".format( + provider=provider.value, + webhook_id=webhook_id, + ) + actual_url = urlparse(webhooks_utils.webhook_ingress_url(provider, webhook_id)) + expected_base = urlparse(base_url) + + assert (actual_url.scheme, actual_url.netloc) == ( + expected_base.scheme, + expected_base.netloc, + ) + assert actual_url.path == expected_path diff --git a/autogpt_platform/backend/migrations/20260126120000_migrate_claude_3_7_to_4_5_sonnet/migration.sql b/autogpt_platform/backend/migrations/20260126120000_migrate_claude_3_7_to_4_5_sonnet/migration.sql new file mode 100644 index 0000000000..5746c80820 --- /dev/null +++ b/autogpt_platform/backend/migrations/20260126120000_migrate_claude_3_7_to_4_5_sonnet/migration.sql @@ -0,0 +1,22 @@ +-- Migrate Claude 3.7 Sonnet to Claude 4.5 Sonnet +-- This updates all AgentNode blocks that use the deprecated Claude 3.7 Sonnet model +-- Anthropic is retiring claude-3-7-sonnet-20250219 on February 19, 2026 + +-- Update AgentNode constant inputs +UPDATE "AgentNode" +SET "constantInput" = JSONB_SET( + "constantInput"::jsonb, + '{model}', + '"claude-sonnet-4-5-20250929"'::jsonb + ) +WHERE "constantInput"::jsonb->>'model' = 'claude-3-7-sonnet-20250219'; + +-- Update AgentPreset input overrides (stored in AgentNodeExecutionInputOutput) +UPDATE "AgentNodeExecutionInputOutput" +SET "data" = JSONB_SET( + "data"::jsonb, + '{model}', + '"claude-sonnet-4-5-20250929"'::jsonb + ) +WHERE "agentPresetId" IS NOT NULL + AND "data"::jsonb->>'model' = 'claude-3-7-sonnet-20250219'; diff --git a/autogpt_platform/backend/snapshots/agts_by_creator b/autogpt_platform/backend/snapshots/agts_by_creator index 4d6dd12920..3f2e128a0d 100644 --- a/autogpt_platform/backend/snapshots/agts_by_creator +++ b/autogpt_platform/backend/snapshots/agts_by_creator @@ -9,7 +9,8 @@ "sub_heading": "Creator agent subheading", "description": "Creator agent description", "runs": 50, - "rating": 4.0 + "rating": 4.0, + "agent_graph_id": "test-graph-2" } ], "pagination": { diff --git a/autogpt_platform/backend/snapshots/agts_category b/autogpt_platform/backend/snapshots/agts_category index f65925ead3..4d0531763c 100644 --- a/autogpt_platform/backend/snapshots/agts_category +++ b/autogpt_platform/backend/snapshots/agts_category @@ -9,7 +9,8 @@ "sub_heading": "Category agent subheading", "description": "Category agent description", "runs": 60, - "rating": 4.1 + "rating": 4.1, + "agent_graph_id": "test-graph-category" } ], "pagination": { diff --git a/autogpt_platform/backend/snapshots/agts_pagination b/autogpt_platform/backend/snapshots/agts_pagination index 82e7f5f9bf..7b946157fb 100644 --- a/autogpt_platform/backend/snapshots/agts_pagination +++ b/autogpt_platform/backend/snapshots/agts_pagination @@ -9,7 +9,8 @@ "sub_heading": "Agent 0 subheading", "description": "Agent 0 description", "runs": 0, - "rating": 4.0 + "rating": 4.0, + "agent_graph_id": "test-graph-2" }, { "slug": "agent-1", @@ -20,7 +21,8 @@ "sub_heading": "Agent 1 subheading", "description": "Agent 1 description", "runs": 10, - "rating": 4.0 + "rating": 4.0, + "agent_graph_id": "test-graph-2" }, { "slug": "agent-2", @@ -31,7 +33,8 @@ "sub_heading": "Agent 2 subheading", "description": "Agent 2 description", "runs": 20, - "rating": 4.0 + "rating": 4.0, + "agent_graph_id": "test-graph-2" }, { "slug": "agent-3", @@ -42,7 +45,8 @@ "sub_heading": "Agent 3 subheading", "description": "Agent 3 description", "runs": 30, - "rating": 4.0 + "rating": 4.0, + "agent_graph_id": "test-graph-2" }, { "slug": "agent-4", @@ -53,7 +57,8 @@ "sub_heading": "Agent 4 subheading", "description": "Agent 4 description", "runs": 40, - "rating": 4.0 + "rating": 4.0, + "agent_graph_id": "test-graph-2" } ], "pagination": { diff --git a/autogpt_platform/backend/snapshots/agts_search b/autogpt_platform/backend/snapshots/agts_search index ca3f504584..ae9cc116bc 100644 --- a/autogpt_platform/backend/snapshots/agts_search +++ b/autogpt_platform/backend/snapshots/agts_search @@ -9,7 +9,8 @@ "sub_heading": "Search agent subheading", "description": "Specific search term description", "runs": 75, - "rating": 4.2 + "rating": 4.2, + "agent_graph_id": "test-graph-search" } ], "pagination": { diff --git a/autogpt_platform/backend/snapshots/agts_sorted b/autogpt_platform/backend/snapshots/agts_sorted index cddead76a5..b182256b2c 100644 --- a/autogpt_platform/backend/snapshots/agts_sorted +++ b/autogpt_platform/backend/snapshots/agts_sorted @@ -9,7 +9,8 @@ "sub_heading": "Top agent subheading", "description": "Top agent description", "runs": 1000, - "rating": 5.0 + "rating": 5.0, + "agent_graph_id": "test-graph-3" } ], "pagination": { diff --git a/autogpt_platform/backend/snapshots/feat_agts b/autogpt_platform/backend/snapshots/feat_agts index d57996a768..4f85786434 100644 --- a/autogpt_platform/backend/snapshots/feat_agts +++ b/autogpt_platform/backend/snapshots/feat_agts @@ -9,7 +9,8 @@ "sub_heading": "Featured agent subheading", "description": "Featured agent description", "runs": 100, - "rating": 4.5 + "rating": 4.5, + "agent_graph_id": "test-graph-1" } ], "pagination": { diff --git a/autogpt_platform/backend/snapshots/lib_agts_search b/autogpt_platform/backend/snapshots/lib_agts_search index 67c307b09e..3ce8402b63 100644 --- a/autogpt_platform/backend/snapshots/lib_agts_search +++ b/autogpt_platform/backend/snapshots/lib_agts_search @@ -31,6 +31,10 @@ "has_sensitive_action": false, "trigger_setup_info": null, "new_output": false, + "execution_count": 0, + "success_rate": null, + "avg_correctness_score": null, + "recent_executions": [], "can_access_graph": true, "is_latest_version": true, "is_favorite": false, @@ -72,6 +76,10 @@ "has_sensitive_action": false, "trigger_setup_info": null, "new_output": false, + "execution_count": 0, + "success_rate": null, + "avg_correctness_score": null, + "recent_executions": [], "can_access_graph": false, "is_latest_version": true, "is_favorite": false, diff --git a/autogpt_platform/backend/test/agent_generator/test_core_integration.py b/autogpt_platform/backend/test/agent_generator/test_core_integration.py index bdcc24ba79..05ce4a3aff 100644 --- a/autogpt_platform/backend/test/agent_generator/test_core_integration.py +++ b/autogpt_platform/backend/test/agent_generator/test_core_integration.py @@ -57,7 +57,8 @@ class TestDecomposeGoal: result = await core.decompose_goal("Build a chatbot") - mock_external.assert_called_once_with("Build a chatbot", "") + # library_agents defaults to None + mock_external.assert_called_once_with("Build a chatbot", "", None) assert result == expected_result @pytest.mark.asyncio @@ -74,7 +75,8 @@ class TestDecomposeGoal: await core.decompose_goal("Build a chatbot", "Use Python") - mock_external.assert_called_once_with("Build a chatbot", "Use Python") + # library_agents defaults to None + mock_external.assert_called_once_with("Build a chatbot", "Use Python", None) @pytest.mark.asyncio async def test_returns_none_on_service_failure(self): @@ -109,7 +111,8 @@ class TestGenerateAgent: instructions = {"type": "instructions", "steps": ["Step 1"]} result = await core.generate_agent(instructions) - mock_external.assert_called_once_with(instructions) + # library_agents defaults to None + mock_external.assert_called_once_with(instructions, None) # Result should have id, version, is_active added if not present assert result is not None assert result["name"] == "Test Agent" @@ -174,7 +177,8 @@ class TestGenerateAgentPatch: current_agent = {"nodes": [], "links": []} result = await core.generate_agent_patch("Add a node", current_agent) - mock_external.assert_called_once_with("Add a node", current_agent) + # library_agents defaults to None + mock_external.assert_called_once_with("Add a node", current_agent, None) assert result == expected_result @pytest.mark.asyncio diff --git a/autogpt_platform/backend/test/agent_generator/test_library_agents.py b/autogpt_platform/backend/test/agent_generator/test_library_agents.py new file mode 100644 index 0000000000..8387339582 --- /dev/null +++ b/autogpt_platform/backend/test/agent_generator/test_library_agents.py @@ -0,0 +1,857 @@ +""" +Tests for library agent fetching functionality in agent generator. + +This test suite verifies the search-based library agent fetching, +including the combination of library and marketplace agents. +""" + +from unittest.mock import AsyncMock, MagicMock, patch + +import pytest + +from backend.api.features.chat.tools.agent_generator import core + + +class TestGetLibraryAgentsForGeneration: + """Test get_library_agents_for_generation function.""" + + @pytest.mark.asyncio + async def test_fetches_agents_with_search_term(self): + """Test that search_term is passed to the library db.""" + # Create a mock agent with proper attribute values + mock_agent = MagicMock() + mock_agent.graph_id = "agent-123" + mock_agent.graph_version = 1 + mock_agent.name = "Email Agent" + mock_agent.description = "Sends emails" + mock_agent.input_schema = {"properties": {}} + mock_agent.output_schema = {"properties": {}} + mock_agent.recent_executions = [] + + mock_response = MagicMock() + mock_response.agents = [mock_agent] + + with patch.object( + core.library_db, + "list_library_agents", + new_callable=AsyncMock, + return_value=mock_response, + ) as mock_list: + result = await core.get_library_agents_for_generation( + user_id="user-123", + search_query="send email", + ) + + mock_list.assert_called_once_with( + user_id="user-123", + search_term="send email", + page=1, + page_size=15, + include_executions=True, + ) + + # Verify result format + assert len(result) == 1 + assert result[0]["graph_id"] == "agent-123" + assert result[0]["name"] == "Email Agent" + + @pytest.mark.asyncio + async def test_excludes_specified_graph_id(self): + """Test that agents with excluded graph_id are filtered out.""" + mock_response = MagicMock() + mock_response.agents = [ + MagicMock( + graph_id="agent-123", + graph_version=1, + name="Agent 1", + description="First agent", + input_schema={}, + output_schema={}, + recent_executions=[], + ), + MagicMock( + graph_id="agent-456", + graph_version=1, + name="Agent 2", + description="Second agent", + input_schema={}, + output_schema={}, + recent_executions=[], + ), + ] + + with patch.object( + core.library_db, + "list_library_agents", + new_callable=AsyncMock, + return_value=mock_response, + ): + result = await core.get_library_agents_for_generation( + user_id="user-123", + exclude_graph_id="agent-123", + ) + + # Verify the excluded agent is not in results + assert len(result) == 1 + assert result[0]["graph_id"] == "agent-456" + + @pytest.mark.asyncio + async def test_respects_max_results(self): + """Test that max_results parameter limits the page_size.""" + mock_response = MagicMock() + mock_response.agents = [] + + with patch.object( + core.library_db, + "list_library_agents", + new_callable=AsyncMock, + return_value=mock_response, + ) as mock_list: + await core.get_library_agents_for_generation( + user_id="user-123", + max_results=5, + ) + + mock_list.assert_called_once_with( + user_id="user-123", + search_term=None, + page=1, + page_size=5, + include_executions=True, + ) + + +class TestSearchMarketplaceAgentsForGeneration: + """Test search_marketplace_agents_for_generation function.""" + + @pytest.mark.asyncio + async def test_searches_marketplace_with_query(self): + """Test that marketplace is searched with the query.""" + mock_response = MagicMock() + mock_response.agents = [ + MagicMock( + agent_name="Public Agent", + description="A public agent", + sub_heading="Does something useful", + creator="creator-1", + agent_graph_id="graph-123", + ) + ] + + mock_graph = MagicMock() + mock_graph.id = "graph-123" + mock_graph.version = 1 + mock_graph.input_schema = {"type": "object"} + mock_graph.output_schema = {"type": "object"} + + with ( + patch( + "backend.api.features.store.db.get_store_agents", + new_callable=AsyncMock, + return_value=mock_response, + ) as mock_search, + patch( + "backend.api.features.chat.tools.agent_generator.core.get_store_listed_graphs", + new_callable=AsyncMock, + return_value={"graph-123": mock_graph}, + ), + ): + result = await core.search_marketplace_agents_for_generation( + search_query="automation", + max_results=10, + ) + + mock_search.assert_called_once_with( + search_query="automation", + page=1, + page_size=10, + ) + + assert len(result) == 1 + assert result[0]["name"] == "Public Agent" + assert result[0]["graph_id"] == "graph-123" + + @pytest.mark.asyncio + async def test_handles_marketplace_error_gracefully(self): + """Test that marketplace errors don't crash the function.""" + with patch( + "backend.api.features.store.db.get_store_agents", + new_callable=AsyncMock, + side_effect=Exception("Marketplace unavailable"), + ): + result = await core.search_marketplace_agents_for_generation( + search_query="test" + ) + + # Should return empty list, not raise exception + assert result == [] + + +class TestGetAllRelevantAgentsForGeneration: + """Test get_all_relevant_agents_for_generation function.""" + + @pytest.mark.asyncio + async def test_combines_library_and_marketplace_agents(self): + """Test that agents from both sources are combined.""" + library_agents = [ + { + "graph_id": "lib-123", + "graph_version": 1, + "name": "Library Agent", + "description": "From library", + "input_schema": {}, + "output_schema": {}, + } + ] + + marketplace_agents = [ + { + "graph_id": "market-456", + "graph_version": 1, + "name": "Market Agent", + "description": "From marketplace", + "input_schema": {}, + "output_schema": {}, + } + ] + + with patch.object( + core, + "get_library_agents_for_generation", + new_callable=AsyncMock, + return_value=library_agents, + ): + with patch.object( + core, + "search_marketplace_agents_for_generation", + new_callable=AsyncMock, + return_value=marketplace_agents, + ): + result = await core.get_all_relevant_agents_for_generation( + user_id="user-123", + search_query="test query", + include_marketplace=True, + ) + + # Library agents should come first + assert len(result) == 2 + assert result[0]["name"] == "Library Agent" + assert result[1]["name"] == "Market Agent" + + @pytest.mark.asyncio + async def test_deduplicates_by_graph_id(self): + """Test that marketplace agents with same graph_id as library are excluded.""" + library_agents = [ + { + "graph_id": "shared-123", + "graph_version": 1, + "name": "Shared Agent", + "description": "From library", + "input_schema": {}, + "output_schema": {}, + } + ] + + marketplace_agents = [ + { + "graph_id": "shared-123", # Same graph_id, should be deduplicated + "graph_version": 1, + "name": "Shared Agent", + "description": "From marketplace", + "input_schema": {}, + "output_schema": {}, + }, + { + "graph_id": "unique-456", + "graph_version": 1, + "name": "Unique Agent", + "description": "Only in marketplace", + "input_schema": {}, + "output_schema": {}, + }, + ] + + with patch.object( + core, + "get_library_agents_for_generation", + new_callable=AsyncMock, + return_value=library_agents, + ): + with patch.object( + core, + "search_marketplace_agents_for_generation", + new_callable=AsyncMock, + return_value=marketplace_agents, + ): + result = await core.get_all_relevant_agents_for_generation( + user_id="user-123", + search_query="test", + include_marketplace=True, + ) + + # Shared Agent from marketplace should be excluded by graph_id + assert len(result) == 2 + names = [a["name"] for a in result] + assert "Shared Agent" in names + assert "Unique Agent" in names + + @pytest.mark.asyncio + async def test_skips_marketplace_when_disabled(self): + """Test that marketplace is not searched when include_marketplace=False.""" + library_agents = [ + { + "graph_id": "lib-123", + "graph_version": 1, + "name": "Library Agent", + "description": "From library", + "input_schema": {}, + "output_schema": {}, + } + ] + + with patch.object( + core, + "get_library_agents_for_generation", + new_callable=AsyncMock, + return_value=library_agents, + ): + with patch.object( + core, + "search_marketplace_agents_for_generation", + new_callable=AsyncMock, + ) as mock_marketplace: + result = await core.get_all_relevant_agents_for_generation( + user_id="user-123", + search_query="test", + include_marketplace=False, + ) + + # Marketplace should not be called + mock_marketplace.assert_not_called() + assert len(result) == 1 + + @pytest.mark.asyncio + async def test_skips_marketplace_when_no_search_query(self): + """Test that marketplace is not searched without a search query.""" + library_agents = [ + { + "graph_id": "lib-123", + "graph_version": 1, + "name": "Library Agent", + "description": "From library", + "input_schema": {}, + "output_schema": {}, + } + ] + + with patch.object( + core, + "get_library_agents_for_generation", + new_callable=AsyncMock, + return_value=library_agents, + ): + with patch.object( + core, + "search_marketplace_agents_for_generation", + new_callable=AsyncMock, + ) as mock_marketplace: + result = await core.get_all_relevant_agents_for_generation( + user_id="user-123", + search_query=None, # No search query + include_marketplace=True, + ) + + # Marketplace should not be called without search query + mock_marketplace.assert_not_called() + assert len(result) == 1 + + +class TestExtractSearchTermsFromSteps: + """Test extract_search_terms_from_steps function.""" + + def test_extracts_terms_from_instructions_type(self): + """Test extraction from valid instructions decomposition result.""" + decomposition_result = { + "type": "instructions", + "steps": [ + { + "description": "Send an email notification", + "block_name": "GmailSendBlock", + }, + {"description": "Fetch weather data", "action": "Get weather API"}, + ], + } + + result = core.extract_search_terms_from_steps(decomposition_result) + + assert "Send an email notification" in result + assert "GmailSendBlock" in result + assert "Fetch weather data" in result + assert "Get weather API" in result + + def test_returns_empty_for_non_instructions_type(self): + """Test that non-instructions types return empty list.""" + decomposition_result = { + "type": "clarifying_questions", + "questions": [{"question": "What email?"}], + } + + result = core.extract_search_terms_from_steps(decomposition_result) + + assert result == [] + + def test_deduplicates_terms_case_insensitively(self): + """Test that duplicate terms are removed (case-insensitive).""" + decomposition_result = { + "type": "instructions", + "steps": [ + {"description": "Send Email", "name": "send email"}, + {"description": "Other task"}, + ], + } + + result = core.extract_search_terms_from_steps(decomposition_result) + + # Should only have one "send email" variant + email_terms = [t for t in result if "email" in t.lower()] + assert len(email_terms) == 1 + + def test_filters_short_terms(self): + """Test that terms with 3 or fewer characters are filtered out.""" + decomposition_result = { + "type": "instructions", + "steps": [ + {"description": "ab", "action": "xyz"}, # Both too short + {"description": "Valid term here"}, + ], + } + + result = core.extract_search_terms_from_steps(decomposition_result) + + assert "ab" not in result + assert "xyz" not in result + assert "Valid term here" in result + + def test_handles_empty_steps(self): + """Test handling of empty steps list.""" + decomposition_result = { + "type": "instructions", + "steps": [], + } + + result = core.extract_search_terms_from_steps(decomposition_result) + + assert result == [] + + +class TestEnrichLibraryAgentsFromSteps: + """Test enrich_library_agents_from_steps function.""" + + @pytest.mark.asyncio + async def test_enriches_with_additional_agents(self): + """Test that additional agents are found based on steps.""" + existing_agents = [ + { + "graph_id": "existing-123", + "graph_version": 1, + "name": "Existing Agent", + "description": "Already fetched", + "input_schema": {}, + "output_schema": {}, + } + ] + + additional_agents = [ + { + "graph_id": "new-456", + "graph_version": 1, + "name": "Email Agent", + "description": "For sending emails", + "input_schema": {}, + "output_schema": {}, + } + ] + + decomposition_result = { + "type": "instructions", + "steps": [ + {"description": "Send email notification"}, + ], + } + + with patch.object( + core, + "get_all_relevant_agents_for_generation", + new_callable=AsyncMock, + return_value=additional_agents, + ): + result = await core.enrich_library_agents_from_steps( + user_id="user-123", + decomposition_result=decomposition_result, + existing_agents=existing_agents, + ) + + # Should have both existing and new agents + assert len(result) == 2 + names = [a["name"] for a in result] + assert "Existing Agent" in names + assert "Email Agent" in names + + @pytest.mark.asyncio + async def test_deduplicates_by_graph_id(self): + """Test that agents with same graph_id are not duplicated.""" + existing_agents = [ + { + "graph_id": "agent-123", + "graph_version": 1, + "name": "Existing Agent", + "description": "Already fetched", + "input_schema": {}, + "output_schema": {}, + } + ] + + # Additional search returns same agent + additional_agents = [ + { + "graph_id": "agent-123", # Same ID + "graph_version": 1, + "name": "Existing Agent Copy", + "description": "Same agent different name", + "input_schema": {}, + "output_schema": {}, + } + ] + + decomposition_result = { + "type": "instructions", + "steps": [{"description": "Some action"}], + } + + with patch.object( + core, + "get_all_relevant_agents_for_generation", + new_callable=AsyncMock, + return_value=additional_agents, + ): + result = await core.enrich_library_agents_from_steps( + user_id="user-123", + decomposition_result=decomposition_result, + existing_agents=existing_agents, + ) + + # Should not duplicate + assert len(result) == 1 + + @pytest.mark.asyncio + async def test_deduplicates_by_name(self): + """Test that agents with same name are not duplicated.""" + existing_agents = [ + { + "graph_id": "agent-123", + "graph_version": 1, + "name": "Email Agent", + "description": "Already fetched", + "input_schema": {}, + "output_schema": {}, + } + ] + + # Additional search returns agent with same name but different ID + additional_agents = [ + { + "graph_id": "agent-456", # Different ID + "graph_version": 1, + "name": "Email Agent", # Same name + "description": "Different agent same name", + "input_schema": {}, + "output_schema": {}, + } + ] + + decomposition_result = { + "type": "instructions", + "steps": [{"description": "Send email"}], + } + + with patch.object( + core, + "get_all_relevant_agents_for_generation", + new_callable=AsyncMock, + return_value=additional_agents, + ): + result = await core.enrich_library_agents_from_steps( + user_id="user-123", + decomposition_result=decomposition_result, + existing_agents=existing_agents, + ) + + # Should not duplicate by name + assert len(result) == 1 + assert result[0].get("graph_id") == "agent-123" # Original kept + + @pytest.mark.asyncio + async def test_returns_existing_when_no_steps(self): + """Test that existing agents are returned when no search terms extracted.""" + existing_agents = [ + { + "graph_id": "existing-123", + "graph_version": 1, + "name": "Existing Agent", + "description": "Already fetched", + "input_schema": {}, + "output_schema": {}, + } + ] + + decomposition_result = { + "type": "clarifying_questions", # Not instructions type + "questions": [], + } + + result = await core.enrich_library_agents_from_steps( + user_id="user-123", + decomposition_result=decomposition_result, + existing_agents=existing_agents, + ) + + # Should return existing unchanged + assert result == existing_agents + + @pytest.mark.asyncio + async def test_limits_search_terms_to_three(self): + """Test that only first 3 search terms are used.""" + existing_agents = [] + + decomposition_result = { + "type": "instructions", + "steps": [ + {"description": "First action"}, + {"description": "Second action"}, + {"description": "Third action"}, + {"description": "Fourth action"}, + {"description": "Fifth action"}, + ], + } + + call_count = 0 + + async def mock_get_agents(*args, **kwargs): + nonlocal call_count + call_count += 1 + return [] + + with patch.object( + core, + "get_all_relevant_agents_for_generation", + side_effect=mock_get_agents, + ): + await core.enrich_library_agents_from_steps( + user_id="user-123", + decomposition_result=decomposition_result, + existing_agents=existing_agents, + ) + + # Should only make 3 calls (limited to first 3 terms) + assert call_count == 3 + + +class TestExtractUuidsFromText: + """Test extract_uuids_from_text function.""" + + def test_extracts_single_uuid(self): + """Test extraction of a single UUID from text.""" + text = "Use my agent 46631191-e8a8-486f-ad90-84f89738321d for this task" + result = core.extract_uuids_from_text(text) + assert len(result) == 1 + assert "46631191-e8a8-486f-ad90-84f89738321d" in result + + def test_extracts_multiple_uuids(self): + """Test extraction of multiple UUIDs from text.""" + text = ( + "Combine agents 11111111-1111-4111-8111-111111111111 " + "and 22222222-2222-4222-9222-222222222222" + ) + result = core.extract_uuids_from_text(text) + assert len(result) == 2 + assert "11111111-1111-4111-8111-111111111111" in result + assert "22222222-2222-4222-9222-222222222222" in result + + def test_deduplicates_uuids(self): + """Test that duplicate UUIDs are deduplicated.""" + text = ( + "Use 46631191-e8a8-486f-ad90-84f89738321d twice: " + "46631191-e8a8-486f-ad90-84f89738321d" + ) + result = core.extract_uuids_from_text(text) + assert len(result) == 1 + + def test_normalizes_to_lowercase(self): + """Test that UUIDs are normalized to lowercase.""" + text = "Use 46631191-E8A8-486F-AD90-84F89738321D" + result = core.extract_uuids_from_text(text) + assert result[0] == "46631191-e8a8-486f-ad90-84f89738321d" + + def test_returns_empty_for_no_uuids(self): + """Test that empty list is returned when no UUIDs found.""" + text = "Create an email agent that sends notifications" + result = core.extract_uuids_from_text(text) + assert result == [] + + def test_ignores_invalid_uuids(self): + """Test that invalid UUID-like strings are ignored.""" + text = "Not a valid UUID: 12345678-1234-1234-1234-123456789abc" + result = core.extract_uuids_from_text(text) + # UUID v4 requires specific patterns (4 in third group, 8/9/a/b in fourth) + assert len(result) == 0 + + +class TestGetLibraryAgentById: + """Test get_library_agent_by_id function (and its alias get_library_agent_by_graph_id).""" + + @pytest.mark.asyncio + async def test_returns_agent_when_found_by_graph_id(self): + """Test that agent is returned when found by graph_id.""" + mock_agent = MagicMock() + mock_agent.graph_id = "agent-123" + mock_agent.graph_version = 1 + mock_agent.name = "Test Agent" + mock_agent.description = "Test description" + mock_agent.input_schema = {"properties": {}} + mock_agent.output_schema = {"properties": {}} + + with patch.object( + core.library_db, + "get_library_agent_by_graph_id", + new_callable=AsyncMock, + return_value=mock_agent, + ): + result = await core.get_library_agent_by_id("user-123", "agent-123") + + assert result is not None + assert result["graph_id"] == "agent-123" + assert result["name"] == "Test Agent" + + @pytest.mark.asyncio + async def test_falls_back_to_library_agent_id(self): + """Test that lookup falls back to library agent ID when graph_id not found.""" + mock_agent = MagicMock() + mock_agent.graph_id = "graph-456" # Different from the lookup ID + mock_agent.graph_version = 1 + mock_agent.name = "Library Agent" + mock_agent.description = "Found by library ID" + mock_agent.input_schema = {"properties": {}} + mock_agent.output_schema = {"properties": {}} + + with ( + patch.object( + core.library_db, + "get_library_agent_by_graph_id", + new_callable=AsyncMock, + return_value=None, # Not found by graph_id + ), + patch.object( + core.library_db, + "get_library_agent", + new_callable=AsyncMock, + return_value=mock_agent, # Found by library ID + ), + ): + result = await core.get_library_agent_by_id("user-123", "library-id-123") + + assert result is not None + assert result["graph_id"] == "graph-456" + assert result["name"] == "Library Agent" + + @pytest.mark.asyncio + async def test_returns_none_when_not_found_by_either_method(self): + """Test that None is returned when agent not found by either method.""" + with ( + patch.object( + core.library_db, + "get_library_agent_by_graph_id", + new_callable=AsyncMock, + return_value=None, + ), + patch.object( + core.library_db, + "get_library_agent", + new_callable=AsyncMock, + side_effect=core.NotFoundError("Not found"), + ), + ): + result = await core.get_library_agent_by_id("user-123", "nonexistent") + + assert result is None + + @pytest.mark.asyncio + async def test_returns_none_on_exception(self): + """Test that None is returned when exception occurs in both lookups.""" + with ( + patch.object( + core.library_db, + "get_library_agent_by_graph_id", + new_callable=AsyncMock, + side_effect=Exception("Database error"), + ), + patch.object( + core.library_db, + "get_library_agent", + new_callable=AsyncMock, + side_effect=Exception("Database error"), + ), + ): + result = await core.get_library_agent_by_id("user-123", "agent-123") + + assert result is None + + @pytest.mark.asyncio + async def test_alias_works(self): + """Test that get_library_agent_by_graph_id is an alias for get_library_agent_by_id.""" + assert core.get_library_agent_by_graph_id is core.get_library_agent_by_id + + +class TestGetAllRelevantAgentsWithUuids: + """Test UUID extraction in get_all_relevant_agents_for_generation.""" + + @pytest.mark.asyncio + async def test_fetches_explicitly_mentioned_agents(self): + """Test that agents mentioned by UUID are fetched directly.""" + mock_agent = MagicMock() + mock_agent.graph_id = "46631191-e8a8-486f-ad90-84f89738321d" + mock_agent.graph_version = 1 + mock_agent.name = "Mentioned Agent" + mock_agent.description = "Explicitly mentioned" + mock_agent.input_schema = {} + mock_agent.output_schema = {} + + mock_response = MagicMock() + mock_response.agents = [] + + with ( + patch.object( + core.library_db, + "get_library_agent_by_graph_id", + new_callable=AsyncMock, + return_value=mock_agent, + ), + patch.object( + core.library_db, + "list_library_agents", + new_callable=AsyncMock, + return_value=mock_response, + ), + ): + result = await core.get_all_relevant_agents_for_generation( + user_id="user-123", + search_query="Use agent 46631191-e8a8-486f-ad90-84f89738321d", + include_marketplace=False, + ) + + assert len(result) == 1 + assert result[0].get("graph_id") == "46631191-e8a8-486f-ad90-84f89738321d" + assert result[0].get("name") == "Mentioned Agent" + + +if __name__ == "__main__": + pytest.main([__file__, "-v"]) diff --git a/autogpt_platform/backend/test/agent_generator/test_service.py b/autogpt_platform/backend/test/agent_generator/test_service.py index 81ff794532..cc37c428c0 100644 --- a/autogpt_platform/backend/test/agent_generator/test_service.py +++ b/autogpt_platform/backend/test/agent_generator/test_service.py @@ -102,7 +102,7 @@ class TestDecomposeGoalExternal: @pytest.mark.asyncio async def test_decompose_goal_with_context(self): - """Test decomposition with additional context.""" + """Test decomposition with additional context enriched into description.""" mock_response = MagicMock() mock_response.json.return_value = { "success": True, @@ -119,9 +119,12 @@ class TestDecomposeGoalExternal: "Build a chatbot", context="Use Python" ) + expected_description = ( + "Build a chatbot\n\nAdditional context from user:\nUse Python" + ) mock_client.post.assert_called_once_with( "/api/decompose-description", - json={"description": "Build a chatbot", "user_instruction": "Use Python"}, + json={"description": expected_description}, ) @pytest.mark.asyncio @@ -151,15 +154,20 @@ class TestDecomposeGoalExternal: @pytest.mark.asyncio async def test_decompose_goal_handles_http_error(self): """Test decomposition handles HTTP errors gracefully.""" + mock_response = MagicMock() + mock_response.status_code = 500 mock_client = AsyncMock() mock_client.post.side_effect = httpx.HTTPStatusError( - "Server error", request=MagicMock(), response=MagicMock() + "Server error", request=MagicMock(), response=mock_response ) with patch.object(service, "_get_client", return_value=mock_client): result = await service.decompose_goal_external("Build a chatbot") - assert result is None + assert result is not None + assert result.get("type") == "error" + assert result.get("error_type") == "http_error" + assert "Server error" in result.get("error", "") @pytest.mark.asyncio async def test_decompose_goal_handles_request_error(self): @@ -170,7 +178,10 @@ class TestDecomposeGoalExternal: with patch.object(service, "_get_client", return_value=mock_client): result = await service.decompose_goal_external("Build a chatbot") - assert result is None + assert result is not None + assert result.get("type") == "error" + assert result.get("error_type") == "connection_error" + assert "Connection failed" in result.get("error", "") @pytest.mark.asyncio async def test_decompose_goal_handles_service_error(self): @@ -179,6 +190,7 @@ class TestDecomposeGoalExternal: mock_response.json.return_value = { "success": False, "error": "Internal error", + "error_type": "internal_error", } mock_response.raise_for_status = MagicMock() @@ -188,7 +200,10 @@ class TestDecomposeGoalExternal: with patch.object(service, "_get_client", return_value=mock_client): result = await service.decompose_goal_external("Build a chatbot") - assert result is None + assert result is not None + assert result.get("type") == "error" + assert result.get("error") == "Internal error" + assert result.get("error_type") == "internal_error" class TestGenerateAgentExternal: @@ -236,7 +251,10 @@ class TestGenerateAgentExternal: with patch.object(service, "_get_client", return_value=mock_client): result = await service.generate_agent_external({"steps": []}) - assert result is None + assert result is not None + assert result.get("type") == "error" + assert result.get("error_type") == "connection_error" + assert "Connection failed" in result.get("error", "") class TestGenerateAgentPatchExternal: @@ -418,5 +436,139 @@ class TestGetBlocksExternal: assert result is None +class TestLibraryAgentsPassthrough: + """Test that library_agents are passed correctly in all requests.""" + + def setup_method(self): + """Reset client singleton before each test.""" + service._settings = None + service._client = None + + @pytest.mark.asyncio + async def test_decompose_goal_passes_library_agents(self): + """Test that library_agents are included in decompose goal payload.""" + library_agents = [ + { + "graph_id": "agent-123", + "graph_version": 1, + "name": "Email Sender", + "description": "Sends emails", + "input_schema": {"properties": {"to": {"type": "string"}}}, + "output_schema": {"properties": {"sent": {"type": "boolean"}}}, + }, + ] + + mock_response = MagicMock() + mock_response.json.return_value = { + "success": True, + "type": "instructions", + "steps": ["Step 1"], + } + mock_response.raise_for_status = MagicMock() + + mock_client = AsyncMock() + mock_client.post.return_value = mock_response + + with patch.object(service, "_get_client", return_value=mock_client): + await service.decompose_goal_external( + "Send an email", + library_agents=library_agents, + ) + + # Verify library_agents was passed in the payload + call_args = mock_client.post.call_args + assert call_args[1]["json"]["library_agents"] == library_agents + + @pytest.mark.asyncio + async def test_generate_agent_passes_library_agents(self): + """Test that library_agents are included in generate agent payload.""" + library_agents = [ + { + "graph_id": "agent-456", + "graph_version": 2, + "name": "Data Fetcher", + "description": "Fetches data from API", + "input_schema": {"properties": {"url": {"type": "string"}}}, + "output_schema": {"properties": {"data": {"type": "object"}}}, + }, + ] + + mock_response = MagicMock() + mock_response.json.return_value = { + "success": True, + "agent_json": {"name": "Test Agent", "nodes": []}, + } + mock_response.raise_for_status = MagicMock() + + mock_client = AsyncMock() + mock_client.post.return_value = mock_response + + with patch.object(service, "_get_client", return_value=mock_client): + await service.generate_agent_external( + {"steps": ["Step 1"]}, + library_agents=library_agents, + ) + + # Verify library_agents was passed in the payload + call_args = mock_client.post.call_args + assert call_args[1]["json"]["library_agents"] == library_agents + + @pytest.mark.asyncio + async def test_generate_agent_patch_passes_library_agents(self): + """Test that library_agents are included in patch generation payload.""" + library_agents = [ + { + "graph_id": "agent-789", + "graph_version": 1, + "name": "Slack Notifier", + "description": "Sends Slack messages", + "input_schema": {"properties": {"message": {"type": "string"}}}, + "output_schema": {"properties": {"success": {"type": "boolean"}}}, + }, + ] + + mock_response = MagicMock() + mock_response.json.return_value = { + "success": True, + "agent_json": {"name": "Updated Agent", "nodes": []}, + } + mock_response.raise_for_status = MagicMock() + + mock_client = AsyncMock() + mock_client.post.return_value = mock_response + + with patch.object(service, "_get_client", return_value=mock_client): + await service.generate_agent_patch_external( + "Add error handling", + {"name": "Original Agent", "nodes": []}, + library_agents=library_agents, + ) + + # Verify library_agents was passed in the payload + call_args = mock_client.post.call_args + assert call_args[1]["json"]["library_agents"] == library_agents + + @pytest.mark.asyncio + async def test_decompose_goal_without_library_agents(self): + """Test that decompose goal works without library_agents.""" + mock_response = MagicMock() + mock_response.json.return_value = { + "success": True, + "type": "instructions", + "steps": ["Step 1"], + } + mock_response.raise_for_status = MagicMock() + + mock_client = AsyncMock() + mock_client.post.return_value = mock_response + + with patch.object(service, "_get_client", return_value=mock_client): + await service.decompose_goal_external("Build a workflow") + + # Verify library_agents was NOT passed when not provided + call_args = mock_client.post.call_args + assert "library_agents" not in call_args[1]["json"] + + if __name__ == "__main__": pytest.main([__file__, "-v"]) diff --git a/autogpt_platform/backend/test/e2e_test_data.py b/autogpt_platform/backend/test/e2e_test_data.py index d7576cdad3..7288197a90 100644 --- a/autogpt_platform/backend/test/e2e_test_data.py +++ b/autogpt_platform/backend/test/e2e_test_data.py @@ -43,19 +43,24 @@ faker = Faker() # Constants for data generation limits (reduced for E2E tests) NUM_USERS = 15 NUM_AGENT_BLOCKS = 30 -MIN_GRAPHS_PER_USER = 15 -MAX_GRAPHS_PER_USER = 15 +MIN_GRAPHS_PER_USER = 25 +MAX_GRAPHS_PER_USER = 25 MIN_NODES_PER_GRAPH = 3 MAX_NODES_PER_GRAPH = 6 MIN_PRESETS_PER_USER = 2 MAX_PRESETS_PER_USER = 3 -MIN_AGENTS_PER_USER = 15 -MAX_AGENTS_PER_USER = 15 +MIN_AGENTS_PER_USER = 25 +MAX_AGENTS_PER_USER = 25 MIN_EXECUTIONS_PER_GRAPH = 2 MAX_EXECUTIONS_PER_GRAPH = 8 MIN_REVIEWS_PER_VERSION = 2 MAX_REVIEWS_PER_VERSION = 5 +# Guaranteed minimums for marketplace tests (deterministic) +GUARANTEED_FEATURED_AGENTS = 8 +GUARANTEED_FEATURED_CREATORS = 5 +GUARANTEED_TOP_AGENTS = 10 + def get_image(): """Generate a consistent image URL using picsum.photos service.""" @@ -385,7 +390,7 @@ class TestDataCreator: library_agents = [] for user in self.users: - num_agents = 10 # Create exactly 10 agents per user + num_agents = random.randint(MIN_AGENTS_PER_USER, MAX_AGENTS_PER_USER) # Get available graphs for this user user_graphs = [ @@ -507,14 +512,17 @@ class TestDataCreator: existing_profiles, min(num_creators, len(existing_profiles)) ) - # Mark about 50% of creators as featured (more for testing) - num_featured = max(2, int(num_creators * 0.5)) + # Guarantee at least GUARANTEED_FEATURED_CREATORS featured creators + num_featured = max(GUARANTEED_FEATURED_CREATORS, int(num_creators * 0.5)) num_featured = min( num_featured, len(selected_profiles) ) # Don't exceed available profiles featured_profile_ids = set( random.sample([p.id for p in selected_profiles], num_featured) ) + print( + f"🎯 Creating {num_featured} featured creators (min: {GUARANTEED_FEATURED_CREATORS})" + ) for profile in selected_profiles: try: @@ -545,21 +553,25 @@ class TestDataCreator: return profiles async def create_test_store_submissions(self) -> List[Dict[str, Any]]: - """Create test store submissions using the API function.""" + """Create test store submissions using the API function. + + DETERMINISTIC: Guarantees minimum featured agents for E2E tests. + """ print("Creating test store submissions...") submissions = [] approved_submissions = [] + featured_count = 0 + submission_counter = 0 - # Create a special test submission for test123@gmail.com + # Create a special test submission for test123@gmail.com (ALWAYS approved + featured) test_user = next( (user for user in self.users if user["email"] == "test123@gmail.com"), None ) - if test_user: - # Special test data for consistent testing + if test_user and self.agent_graphs: test_submission_data = { "user_id": test_user["id"], - "agent_id": self.agent_graphs[0]["id"], # Use first available graph + "agent_id": self.agent_graphs[0]["id"], "agent_version": 1, "slug": "test-agent-submission", "name": "Test Agent Submission", @@ -580,37 +592,24 @@ class TestDataCreator: submissions.append(test_submission.model_dump()) print("✅ Created special test store submission for test123@gmail.com") - # Randomly approve, reject, or leave pending the test submission + # ALWAYS approve and feature the test submission if test_submission.store_listing_version_id: - random_value = random.random() - if random_value < 0.4: # 40% chance to approve - approved_submission = await review_store_submission( - store_listing_version_id=test_submission.store_listing_version_id, - is_approved=True, - external_comments="Test submission approved", - internal_comments="Auto-approved test submission", - reviewer_id=test_user["id"], - ) - approved_submissions.append(approved_submission.model_dump()) - print("✅ Approved test store submission") + approved_submission = await review_store_submission( + store_listing_version_id=test_submission.store_listing_version_id, + is_approved=True, + external_comments="Test submission approved", + internal_comments="Auto-approved test submission", + reviewer_id=test_user["id"], + ) + approved_submissions.append(approved_submission.model_dump()) + print("✅ Approved test store submission") - # Mark approved submission as featured - await prisma.storelistingversion.update( - where={"id": test_submission.store_listing_version_id}, - data={"isFeatured": True}, - ) - print("🌟 Marked test agent as FEATURED") - elif random_value < 0.7: # 30% chance to reject (40% to 70%) - await review_store_submission( - store_listing_version_id=test_submission.store_listing_version_id, - is_approved=False, - external_comments="Test submission rejected - needs improvements", - internal_comments="Auto-rejected test submission for E2E testing", - reviewer_id=test_user["id"], - ) - print("❌ Rejected test store submission") - else: # 30% chance to leave pending (70% to 100%) - print("⏳ Left test submission pending for review") + await prisma.storelistingversion.update( + where={"id": test_submission.store_listing_version_id}, + data={"isFeatured": True}, + ) + featured_count += 1 + print("🌟 Marked test agent as FEATURED") except Exception as e: print(f"Error creating test store submission: {e}") @@ -620,7 +619,6 @@ class TestDataCreator: # Create regular submissions for all users for user in self.users: - # Get available graphs for this specific user user_graphs = [ g for g in self.agent_graphs if g.get("userId") == user["id"] ] @@ -631,18 +629,17 @@ class TestDataCreator: ) continue - # Create exactly 4 store submissions per user for submission_index in range(4): graph = random.choice(user_graphs) + submission_counter += 1 try: print( - f"Creating store submission for user {user['id']} with graph {graph['id']} (owner: {graph.get('userId')})" + f"Creating store submission for user {user['id']} with graph {graph['id']}" ) - # Use the API function to create store submission with correct parameters submission = await create_store_submission( - user_id=user["id"], # Must match graph's userId + user_id=user["id"], agent_id=graph["id"], agent_version=graph.get("version", 1), slug=faker.slug(), @@ -651,22 +648,24 @@ class TestDataCreator: video_url=get_video_url() if random.random() < 0.3 else None, image_urls=[get_image() for _ in range(3)], description=faker.text(), - categories=[ - get_category() - ], # Single category from predefined list + categories=[get_category()], changes_summary="Initial E2E test submission", ) submissions.append(submission.model_dump()) print(f"✅ Created store submission: {submission.name}") - # Randomly approve, reject, or leave pending the submission if submission.store_listing_version_id: - random_value = random.random() - if random_value < 0.4: # 40% chance to approve - try: - # Pick a random user as the reviewer (admin) - reviewer_id = random.choice(self.users)["id"] + # DETERMINISTIC: First N submissions are always approved + # First GUARANTEED_FEATURED_AGENTS of those are always featured + should_approve = ( + submission_counter <= GUARANTEED_TOP_AGENTS + or random.random() < 0.4 + ) + should_feature = featured_count < GUARANTEED_FEATURED_AGENTS + if should_approve: + try: + reviewer_id = random.choice(self.users)["id"] approved_submission = await review_store_submission( store_listing_version_id=submission.store_listing_version_id, is_approved=True, @@ -681,16 +680,7 @@ class TestDataCreator: f"✅ Approved store submission: {submission.name}" ) - # Mark some agents as featured during creation (30% chance) - # More likely for creators and first submissions - is_creator = user["id"] in [ - p.get("userId") for p in self.profiles - ] - feature_chance = ( - 0.5 if is_creator else 0.2 - ) # 50% for creators, 20% for others - - if random.random() < feature_chance: + if should_feature: try: await prisma.storelistingversion.update( where={ @@ -698,8 +688,25 @@ class TestDataCreator: }, data={"isFeatured": True}, ) + featured_count += 1 print( - f"🌟 Marked agent as FEATURED: {submission.name}" + f"🌟 Marked agent as FEATURED ({featured_count}/{GUARANTEED_FEATURED_AGENTS}): {submission.name}" + ) + except Exception as e: + print( + f"Warning: Could not mark submission as featured: {e}" + ) + elif random.random() < 0.2: + try: + await prisma.storelistingversion.update( + where={ + "id": submission.store_listing_version_id + }, + data={"isFeatured": True}, + ) + featured_count += 1 + print( + f"🌟 Marked agent as FEATURED (bonus): {submission.name}" ) except Exception as e: print( @@ -710,11 +717,9 @@ class TestDataCreator: print( f"Warning: Could not approve submission {submission.name}: {e}" ) - elif random_value < 0.7: # 30% chance to reject (40% to 70%) + elif random.random() < 0.5: try: - # Pick a random user as the reviewer (admin) reviewer_id = random.choice(self.users)["id"] - await review_store_submission( store_listing_version_id=submission.store_listing_version_id, is_approved=False, @@ -729,7 +734,7 @@ class TestDataCreator: print( f"Warning: Could not reject submission {submission.name}: {e}" ) - else: # 30% chance to leave pending (70% to 100%) + else: print( f"⏳ Left submission pending for review: {submission.name}" ) @@ -743,9 +748,13 @@ class TestDataCreator: traceback.print_exc() continue + print("\n📊 Store Submissions Summary:") + print(f" Created: {len(submissions)}") + print(f" Approved: {len(approved_submissions)}") print( - f"Created {len(submissions)} store submissions, approved {len(approved_submissions)}" + f" Featured: {featured_count} (guaranteed min: {GUARANTEED_FEATURED_AGENTS})" ) + self.store_submissions = submissions return submissions @@ -825,12 +834,15 @@ class TestDataCreator: print(f"✅ Agent blocks available: {len(self.agent_blocks)}") print(f"✅ Agent graphs created: {len(self.agent_graphs)}") print(f"✅ Library agents created: {len(self.library_agents)}") - print(f"✅ Creator profiles updated: {len(self.profiles)} (some featured)") - print( - f"✅ Store submissions created: {len(self.store_submissions)} (some marked as featured during creation)" - ) + print(f"✅ Creator profiles updated: {len(self.profiles)}") + print(f"✅ Store submissions created: {len(self.store_submissions)}") print(f"✅ API keys created: {len(self.api_keys)}") print(f"✅ Presets created: {len(self.presets)}") + print("\n🎯 Deterministic Guarantees:") + print(f" • Featured agents: >= {GUARANTEED_FEATURED_AGENTS}") + print(f" • Featured creators: >= {GUARANTEED_FEATURED_CREATORS}") + print(f" • Top agents (approved): >= {GUARANTEED_TOP_AGENTS}") + print(f" • Library agents per user: >= {MIN_AGENTS_PER_USER}") print("\n🚀 Your E2E test database is ready to use!") diff --git a/autogpt_platform/frontend/.env.default b/autogpt_platform/frontend/.env.default index af250fb8bf..7a9d81e39e 100644 --- a/autogpt_platform/frontend/.env.default +++ b/autogpt_platform/frontend/.env.default @@ -34,3 +34,6 @@ NEXT_PUBLIC_PREVIEW_STEALING_DEV= # PostHog Analytics NEXT_PUBLIC_POSTHOG_KEY= NEXT_PUBLIC_POSTHOG_HOST=https://eu.i.posthog.com + +# OpenAI (for voice transcription) +OPENAI_API_KEY= diff --git a/autogpt_platform/frontend/CLAUDE.md b/autogpt_platform/frontend/CLAUDE.md new file mode 100644 index 0000000000..b58f1ad6aa --- /dev/null +++ b/autogpt_platform/frontend/CLAUDE.md @@ -0,0 +1,76 @@ +# CLAUDE.md - Frontend + +This file provides guidance to Claude Code when working with the frontend. + +## Essential Commands + +```bash +# Install dependencies +pnpm i + +# Generate API client from OpenAPI spec +pnpm generate:api + +# Start development server +pnpm dev + +# Run E2E tests +pnpm test + +# Run Storybook for component development +pnpm storybook + +# Build production +pnpm build + +# Format and lint +pnpm format + +# Type checking +pnpm types +``` + +### Code Style + +- Fully capitalize acronyms in symbols, e.g. `graphID`, `useBackendAPI` +- Use function declarations (not arrow functions) for components/handlers + +## Architecture + +- **Framework**: Next.js 15 App Router (client-first approach) +- **Data Fetching**: Type-safe generated API hooks via Orval + React Query +- **State Management**: React Query for server state, co-located UI state in components/hooks +- **Component Structure**: Separate render logic (`.tsx`) from business logic (`use*.ts` hooks) +- **Workflow Builder**: Visual graph editor using @xyflow/react +- **UI Components**: shadcn/ui (Radix UI primitives) with Tailwind CSS styling +- **Icons**: Phosphor Icons only +- **Feature Flags**: LaunchDarkly integration +- **Error Handling**: ErrorCard for render errors, toast for mutations, Sentry for exceptions +- **Testing**: Playwright for E2E, Storybook for component development + +## Environment Configuration + +`.env.default` (defaults) → `.env` (user overrides) + +## Feature Development + +See @CONTRIBUTING.md for complete patterns. Quick reference: + +1. **Pages**: Create in `src/app/(platform)/feature-name/page.tsx` + - Extract component logic into custom hooks grouped by concern, not by component. Each hook should represent a cohesive domain of functionality (e.g., useSearch, useFilters, usePagination) rather than bundling all state into one useComponentState hook. + - Put each hook in its own `.ts` file + - Put sub-components in local `components/` folder + - Component props should be `type Props = { ... }` (not exported) unless it needs to be used outside the component +2. **Components**: Structure as `ComponentName/ComponentName.tsx` + `useComponentName.ts` + `helpers.ts` + - Use design system components from `src/components/` (atoms, molecules, organisms) + - Never use `src/components/__legacy__/*` +3. **Data fetching**: Use generated API hooks from `@/app/api/__generated__/endpoints/` + - Regenerate with `pnpm generate:api` + - Pattern: `use{Method}{Version}{OperationName}` +4. **Styling**: Tailwind CSS only, use design tokens, Phosphor Icons only +5. **Testing**: Add Storybook stories for new components, Playwright for E2E +6. **Code conventions**: + - Use function declarations (not arrow functions) for components/handlers + - Do not use `useCallback` or `useMemo` unless asked to optimise a given function + - Do not type hook returns, let Typescript infer as much as possible + - Never type with `any` unless a variable/attribute can ACTUALLY be of any type diff --git a/autogpt_platform/frontend/src/app/(platform)/build/components/legacy-builder/CustomNode/CustomNode.tsx b/autogpt_platform/frontend/src/app/(platform)/build/components/legacy-builder/CustomNode/CustomNode.tsx index 94e917a4ac..834603cc4a 100644 --- a/autogpt_platform/frontend/src/app/(platform)/build/components/legacy-builder/CustomNode/CustomNode.tsx +++ b/autogpt_platform/frontend/src/app/(platform)/build/components/legacy-builder/CustomNode/CustomNode.tsx @@ -857,7 +857,7 @@ export const CustomNode = React.memo( })(); const hasAdvancedFields = - data.inputSchema && + data.inputSchema?.properties && Object.entries(data.inputSchema.properties).some(([key, value]) => { return ( value.advanced === true && !data.inputSchema.required?.includes(key) diff --git a/autogpt_platform/frontend/src/app/(platform)/copilot/components/CopilotShell/components/SessionsList/useSessionsPagination.ts b/autogpt_platform/frontend/src/app/(platform)/copilot/components/CopilotShell/components/SessionsList/useSessionsPagination.ts index 11ddd937af..61e3e6f37f 100644 --- a/autogpt_platform/frontend/src/app/(platform)/copilot/components/CopilotShell/components/SessionsList/useSessionsPagination.ts +++ b/autogpt_platform/frontend/src/app/(platform)/copilot/components/CopilotShell/components/SessionsList/useSessionsPagination.ts @@ -73,9 +73,9 @@ export function useSessionsPagination({ enabled }: UseSessionsPaginationArgs) { }; const reset = () => { + // Only reset the offset - keep existing sessions visible during refetch + // The effect will replace sessions when new data arrives at offset 0 setOffset(0); - setAccumulatedSessions([]); - setTotalCount(null); }; return { diff --git a/autogpt_platform/frontend/src/app/api/openapi.json b/autogpt_platform/frontend/src/app/api/openapi.json index bf8f58fe22..7b0cc410b4 100644 --- a/autogpt_platform/frontend/src/app/api/openapi.json +++ b/autogpt_platform/frontend/src/app/api/openapi.json @@ -7967,6 +7967,25 @@ ] }, "new_output": { "type": "boolean", "title": "New Output" }, + "execution_count": { + "type": "integer", + "title": "Execution Count", + "default": 0 + }, + "success_rate": { + "anyOf": [{ "type": "number" }, { "type": "null" }], + "title": "Success Rate" + }, + "avg_correctness_score": { + "anyOf": [{ "type": "number" }, { "type": "null" }], + "title": "Avg Correctness Score" + }, + "recent_executions": { + "items": { "$ref": "#/components/schemas/RecentExecution" }, + "type": "array", + "title": "Recent Executions", + "description": "List of recent executions with status, score, and summary" + }, "can_access_graph": { "type": "boolean", "title": "Can Access Graph" @@ -9360,6 +9379,23 @@ "required": ["providers", "pagination"], "title": "ProviderResponse" }, + "RecentExecution": { + "properties": { + "status": { "type": "string", "title": "Status" }, + "correctness_score": { + "anyOf": [{ "type": "number" }, { "type": "null" }], + "title": "Correctness Score" + }, + "activity_summary": { + "anyOf": [{ "type": "string" }, { "type": "null" }], + "title": "Activity Summary" + } + }, + "type": "object", + "required": ["status"], + "title": "RecentExecution", + "description": "Summary of a recent execution for quality assessment.\n\nUsed by the LLM to understand the agent's recent performance with specific examples\nrather than just aggregate statistics." + }, "RefundRequest": { "properties": { "id": { "type": "string", "title": "Id" }, @@ -9783,7 +9819,8 @@ "sub_heading": { "type": "string", "title": "Sub Heading" }, "description": { "type": "string", "title": "Description" }, "runs": { "type": "integer", "title": "Runs" }, - "rating": { "type": "number", "title": "Rating" } + "rating": { "type": "number", "title": "Rating" }, + "agent_graph_id": { "type": "string", "title": "Agent Graph Id" } }, "type": "object", "required": [ @@ -9795,7 +9832,8 @@ "sub_heading", "description", "runs", - "rating" + "rating", + "agent_graph_id" ], "title": "StoreAgent" }, diff --git a/autogpt_platform/frontend/src/app/api/transcribe/route.ts b/autogpt_platform/frontend/src/app/api/transcribe/route.ts new file mode 100644 index 0000000000..10c182cdfa --- /dev/null +++ b/autogpt_platform/frontend/src/app/api/transcribe/route.ts @@ -0,0 +1,77 @@ +import { getServerAuthToken } from "@/lib/autogpt-server-api/helpers"; +import { NextRequest, NextResponse } from "next/server"; + +const WHISPER_API_URL = "https://api.openai.com/v1/audio/transcriptions"; +const MAX_FILE_SIZE = 25 * 1024 * 1024; // 25MB - Whisper's limit + +function getExtensionFromMimeType(mimeType: string): string { + const subtype = mimeType.split("/")[1]?.split(";")[0]; + return subtype || "webm"; +} + +export async function POST(request: NextRequest) { + const token = await getServerAuthToken(); + + if (!token || token === "no-token-found") { + return NextResponse.json({ error: "Unauthorized" }, { status: 401 }); + } + + const apiKey = process.env.OPENAI_API_KEY; + + if (!apiKey) { + return NextResponse.json( + { error: "OpenAI API key not configured" }, + { status: 401 }, + ); + } + + try { + const formData = await request.formData(); + const audioFile = formData.get("audio"); + + if (!audioFile || !(audioFile instanceof Blob)) { + return NextResponse.json( + { error: "No audio file provided" }, + { status: 400 }, + ); + } + + if (audioFile.size > MAX_FILE_SIZE) { + return NextResponse.json( + { error: "File too large. Maximum size is 25MB." }, + { status: 413 }, + ); + } + + const ext = getExtensionFromMimeType(audioFile.type); + const whisperFormData = new FormData(); + whisperFormData.append("file", audioFile, `recording.${ext}`); + whisperFormData.append("model", "whisper-1"); + + const response = await fetch(WHISPER_API_URL, { + method: "POST", + headers: { + Authorization: `Bearer ${apiKey}`, + }, + body: whisperFormData, + }); + + if (!response.ok) { + const errorData = await response.json().catch(() => ({})); + console.error("Whisper API error:", errorData); + return NextResponse.json( + { error: errorData.error?.message || "Transcription failed" }, + { status: response.status }, + ); + } + + const result = await response.json(); + return NextResponse.json({ text: result.text }); + } catch (error) { + console.error("Transcription error:", error); + return NextResponse.json( + { error: "Failed to process audio" }, + { status: 500 }, + ); + } +} diff --git a/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/ChatInput.tsx b/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/ChatInput.tsx index c45e8dc250..beb4678e73 100644 --- a/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/ChatInput.tsx +++ b/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/ChatInput.tsx @@ -1,7 +1,14 @@ import { Button } from "@/components/atoms/Button/Button"; import { cn } from "@/lib/utils"; -import { ArrowUpIcon, StopIcon } from "@phosphor-icons/react"; +import { + ArrowUpIcon, + CircleNotchIcon, + MicrophoneIcon, + StopIcon, +} from "@phosphor-icons/react"; +import { RecordingIndicator } from "./components/RecordingIndicator"; import { useChatInput } from "./useChatInput"; +import { useVoiceRecording } from "./useVoiceRecording"; export interface Props { onSend: (message: string) => void; @@ -21,13 +28,37 @@ export function ChatInput({ className, }: Props) { const inputId = "chat-input"; - const { value, handleKeyDown, handleSubmit, handleChange, hasMultipleLines } = - useChatInput({ - onSend, - disabled: disabled || isStreaming, - maxRows: 4, - inputId, - }); + const { + value, + setValue, + handleKeyDown: baseHandleKeyDown, + handleSubmit, + handleChange, + hasMultipleLines, + } = useChatInput({ + onSend, + disabled: disabled || isStreaming, + maxRows: 4, + inputId, + }); + + const { + isRecording, + isTranscribing, + elapsedTime, + toggleRecording, + handleKeyDown, + showMicButton, + isInputDisabled, + audioStream, + } = useVoiceRecording({ + setValue, + disabled: disabled || isStreaming, + isStreaming, + value, + baseHandleKeyDown, + inputId, + }); return (
@@ -35,8 +66,11 @@ export function ChatInput({
@@ -46,48 +80,94 @@ export function ChatInput({ value={value} onChange={handleChange} onKeyDown={handleKeyDown} - placeholder={placeholder} - disabled={disabled || isStreaming} + placeholder={ + isTranscribing + ? "Transcribing..." + : isRecording + ? "" + : placeholder + } + disabled={isInputDisabled} rows={1} className={cn( "w-full resize-none overflow-y-auto border-0 bg-transparent text-[1rem] leading-6 text-black", "placeholder:text-zinc-400", "focus:outline-none focus:ring-0", "disabled:text-zinc-500", - hasMultipleLines ? "pb-6 pl-4 pr-4 pt-2" : "pb-4 pl-4 pr-14 pt-4", + hasMultipleLines + ? "pb-6 pl-4 pr-4 pt-2" + : showMicButton + ? "pb-4 pl-14 pr-14 pt-4" + : "pb-4 pl-4 pr-14 pt-4", )} /> + {isRecording && !value && ( +
+ +
+ )}
- Press Enter to send, Shift+Enter for new line + Press Enter to send, Shift+Enter for new line, Space to record voice - {isStreaming ? ( - - ) : ( - + {showMicButton && ( +
+ +
)} + +
+ {isStreaming ? ( + + ) : ( + + )} +
); diff --git a/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/components/AudioWaveform.tsx b/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/components/AudioWaveform.tsx new file mode 100644 index 0000000000..10cbb3fc9f --- /dev/null +++ b/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/components/AudioWaveform.tsx @@ -0,0 +1,142 @@ +"use client"; + +import { useEffect, useRef, useState } from "react"; + +interface Props { + stream: MediaStream | null; + barCount?: number; + barWidth?: number; + barGap?: number; + barColor?: string; + minBarHeight?: number; + maxBarHeight?: number; +} + +export function AudioWaveform({ + stream, + barCount = 24, + barWidth = 3, + barGap = 2, + barColor = "#ef4444", // red-500 + minBarHeight = 4, + maxBarHeight = 32, +}: Props) { + const [bars, setBars] = useState(() => + Array(barCount).fill(minBarHeight), + ); + const analyserRef = useRef(null); + const audioContextRef = useRef(null); + const sourceRef = useRef(null); + const animationRef = useRef(null); + + useEffect(() => { + if (!stream) { + setBars(Array(barCount).fill(minBarHeight)); + return; + } + + // Create audio context and analyser + const audioContext = new AudioContext(); + const analyser = audioContext.createAnalyser(); + analyser.fftSize = 512; + analyser.smoothingTimeConstant = 0.8; + + // Connect the stream to the analyser + const source = audioContext.createMediaStreamSource(stream); + source.connect(analyser); + + audioContextRef.current = audioContext; + analyserRef.current = analyser; + sourceRef.current = source; + + const timeData = new Uint8Array(analyser.frequencyBinCount); + + const updateBars = () => { + if (!analyserRef.current) return; + + analyserRef.current.getByteTimeDomainData(timeData); + + // Distribute time-domain data across bars + // This shows waveform amplitude, making all bars respond to audio + const newBars: number[] = []; + const samplesPerBar = timeData.length / barCount; + + for (let i = 0; i < barCount; i++) { + // Sample waveform data for this bar + let maxAmplitude = 0; + const startIdx = Math.floor(i * samplesPerBar); + const endIdx = Math.floor((i + 1) * samplesPerBar); + + for (let j = startIdx; j < endIdx && j < timeData.length; j++) { + // Convert to amplitude (distance from center 128) + const amplitude = Math.abs(timeData[j] - 128); + maxAmplitude = Math.max(maxAmplitude, amplitude); + } + + // Map amplitude (0-128) to bar height + const normalized = (maxAmplitude / 128) * 255; + const height = + minBarHeight + (normalized / 255) * (maxBarHeight - minBarHeight); + newBars.push(height); + } + + setBars(newBars); + animationRef.current = requestAnimationFrame(updateBars); + }; + + updateBars(); + + return () => { + if (animationRef.current) { + cancelAnimationFrame(animationRef.current); + } + if (sourceRef.current) { + sourceRef.current.disconnect(); + } + if (audioContextRef.current) { + audioContextRef.current.close(); + } + analyserRef.current = null; + audioContextRef.current = null; + sourceRef.current = null; + }; + }, [stream, barCount, minBarHeight, maxBarHeight]); + + const totalWidth = barCount * barWidth + (barCount - 1) * barGap; + + return ( +
+ {bars.map((height, i) => { + const barHeight = Math.max(minBarHeight, height); + return ( +
+
+
+ ); + })} +
+ ); +} diff --git a/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/components/RecordingIndicator.tsx b/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/components/RecordingIndicator.tsx new file mode 100644 index 0000000000..0be0d069bb --- /dev/null +++ b/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/components/RecordingIndicator.tsx @@ -0,0 +1,26 @@ +import { formatElapsedTime } from "../helpers"; +import { AudioWaveform } from "./AudioWaveform"; + +type Props = { + elapsedTime: number; + audioStream: MediaStream | null; +}; + +export function RecordingIndicator({ elapsedTime, audioStream }: Props) { + return ( +
+ + + {formatElapsedTime(elapsedTime)} + +
+ ); +} diff --git a/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/helpers.ts b/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/helpers.ts new file mode 100644 index 0000000000..26bae8c9d9 --- /dev/null +++ b/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/helpers.ts @@ -0,0 +1,6 @@ +export function formatElapsedTime(ms: number): string { + const seconds = Math.floor(ms / 1000); + const minutes = Math.floor(seconds / 60); + const remainingSeconds = seconds % 60; + return `${minutes}:${remainingSeconds.toString().padStart(2, "0")}`; +} diff --git a/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/useChatInput.ts b/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/useChatInput.ts index 6fa8e7252b..a053e6080f 100644 --- a/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/useChatInput.ts +++ b/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/useChatInput.ts @@ -6,7 +6,7 @@ import { useState, } from "react"; -interface UseChatInputArgs { +interface Args { onSend: (message: string) => void; disabled?: boolean; maxRows?: number; @@ -18,7 +18,7 @@ export function useChatInput({ disabled = false, maxRows = 5, inputId = "chat-input", -}: UseChatInputArgs) { +}: Args) { const [value, setValue] = useState(""); const [hasMultipleLines, setHasMultipleLines] = useState(false); diff --git a/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/useVoiceRecording.ts b/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/useVoiceRecording.ts new file mode 100644 index 0000000000..4de74ef2e9 --- /dev/null +++ b/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatInput/useVoiceRecording.ts @@ -0,0 +1,251 @@ +import { useToast } from "@/components/molecules/Toast/use-toast"; +import React, { + KeyboardEvent, + useCallback, + useEffect, + useRef, + useState, +} from "react"; + +const MAX_RECORDING_DURATION = 2 * 60 * 1000; // 2 minutes in ms + +interface Args { + setValue: React.Dispatch>; + disabled?: boolean; + isStreaming?: boolean; + value: string; + baseHandleKeyDown: (event: KeyboardEvent) => void; + inputId?: string; +} + +export function useVoiceRecording({ + setValue, + disabled = false, + isStreaming = false, + value, + baseHandleKeyDown, + inputId, +}: Args) { + const [isRecording, setIsRecording] = useState(false); + const [isTranscribing, setIsTranscribing] = useState(false); + const [error, setError] = useState(null); + const [elapsedTime, setElapsedTime] = useState(0); + + const mediaRecorderRef = useRef(null); + const chunksRef = useRef([]); + const timerRef = useRef(null); + const startTimeRef = useRef(0); + const streamRef = useRef(null); + const isRecordingRef = useRef(false); + + const isSupported = + typeof window !== "undefined" && + !!(navigator.mediaDevices && navigator.mediaDevices.getUserMedia); + + const clearTimer = useCallback(() => { + if (timerRef.current) { + clearInterval(timerRef.current); + timerRef.current = null; + } + }, []); + + const cleanup = useCallback(() => { + clearTimer(); + if (streamRef.current) { + streamRef.current.getTracks().forEach((track) => track.stop()); + streamRef.current = null; + } + mediaRecorderRef.current = null; + chunksRef.current = []; + setElapsedTime(0); + }, [clearTimer]); + + const handleTranscription = useCallback( + (text: string) => { + setValue((prev) => { + const trimmedPrev = prev.trim(); + if (trimmedPrev) { + return `${trimmedPrev} ${text}`; + } + return text; + }); + }, + [setValue], + ); + + const transcribeAudio = useCallback( + async (audioBlob: Blob) => { + setIsTranscribing(true); + setError(null); + + try { + const formData = new FormData(); + formData.append("audio", audioBlob); + + const response = await fetch("/api/transcribe", { + method: "POST", + body: formData, + }); + + if (!response.ok) { + const data = await response.json().catch(() => ({})); + throw new Error(data.error || "Transcription failed"); + } + + const data = await response.json(); + if (data.text) { + handleTranscription(data.text); + } + } catch (err) { + const message = + err instanceof Error ? err.message : "Transcription failed"; + setError(message); + console.error("Transcription error:", err); + } finally { + setIsTranscribing(false); + } + }, + [handleTranscription, inputId], + ); + + const stopRecording = useCallback(() => { + if (mediaRecorderRef.current && isRecordingRef.current) { + mediaRecorderRef.current.stop(); + isRecordingRef.current = false; + setIsRecording(false); + clearTimer(); + } + }, [clearTimer]); + + const startRecording = useCallback(async () => { + if (disabled || isRecordingRef.current || isTranscribing) return; + + setError(null); + chunksRef.current = []; + + try { + const stream = await navigator.mediaDevices.getUserMedia({ audio: true }); + streamRef.current = stream; + + const mediaRecorder = new MediaRecorder(stream, { + mimeType: MediaRecorder.isTypeSupported("audio/webm") + ? "audio/webm" + : "audio/mp4", + }); + + mediaRecorderRef.current = mediaRecorder; + + mediaRecorder.ondataavailable = (event) => { + if (event.data.size > 0) { + chunksRef.current.push(event.data); + } + }; + + mediaRecorder.onstop = async () => { + const audioBlob = new Blob(chunksRef.current, { + type: mediaRecorder.mimeType, + }); + + // Cleanup stream + if (streamRef.current) { + streamRef.current.getTracks().forEach((track) => track.stop()); + streamRef.current = null; + } + + if (audioBlob.size > 0) { + await transcribeAudio(audioBlob); + } + }; + + mediaRecorder.start(1000); // Collect data every second + isRecordingRef.current = true; + setIsRecording(true); + startTimeRef.current = Date.now(); + + // Start elapsed time timer + timerRef.current = setInterval(() => { + const elapsed = Date.now() - startTimeRef.current; + setElapsedTime(elapsed); + + // Auto-stop at max duration + if (elapsed >= MAX_RECORDING_DURATION) { + stopRecording(); + } + }, 100); + } catch (err) { + console.error("Failed to start recording:", err); + if (err instanceof DOMException && err.name === "NotAllowedError") { + setError("Microphone permission denied"); + } else { + setError("Failed to access microphone"); + } + cleanup(); + } + }, [disabled, isTranscribing, stopRecording, transcribeAudio, cleanup]); + + const toggleRecording = useCallback(() => { + if (isRecording) { + stopRecording(); + } else { + startRecording(); + } + }, [isRecording, startRecording, stopRecording]); + + const { toast } = useToast(); + + useEffect(() => { + if (error) { + toast({ + title: "Voice recording failed", + description: error, + variant: "destructive", + }); + } + }, [error, toast]); + + useEffect(() => { + if (!isTranscribing && inputId) { + const inputElement = document.getElementById(inputId); + if (inputElement) { + inputElement.focus(); + } + } + }, [isTranscribing, inputId]); + + const handleKeyDown = useCallback( + (event: KeyboardEvent) => { + if (event.key === " " && !value.trim() && !isTranscribing) { + event.preventDefault(); + toggleRecording(); + return; + } + baseHandleKeyDown(event); + }, + [value, isTranscribing, toggleRecording, baseHandleKeyDown], + ); + + const showMicButton = isSupported && !isStreaming; + const isInputDisabled = disabled || isStreaming || isTranscribing; + + // Cleanup on unmount + useEffect(() => { + return () => { + cleanup(); + }; + }, [cleanup]); + + return { + isRecording, + isTranscribing, + error, + elapsedTime, + startRecording, + stopRecording, + toggleRecording, + isSupported, + handleKeyDown, + showMicButton, + isInputDisabled, + audioStream: streamRef.current, + }; +} diff --git a/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatMessage/ChatMessage.tsx b/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatMessage/ChatMessage.tsx index c922d0da76..2ac433a272 100644 --- a/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatMessage/ChatMessage.tsx +++ b/autogpt_platform/frontend/src/components/contextual/Chat/components/ChatMessage/ChatMessage.tsx @@ -156,11 +156,19 @@ export function ChatMessage({ } if (isClarificationNeeded && message.type === "clarification_needed") { + const hasUserReplyAfter = + index >= 0 && + messages + .slice(index + 1) + .some((m) => m.type === "message" && m.role === "user"); + return ( ); diff --git a/autogpt_platform/frontend/src/components/contextual/Chat/components/ClarificationQuestionsWidget/ClarificationQuestionsWidget.tsx b/autogpt_platform/frontend/src/components/contextual/Chat/components/ClarificationQuestionsWidget/ClarificationQuestionsWidget.tsx index a3bd17dd3f..3b225d1ef1 100644 --- a/autogpt_platform/frontend/src/components/contextual/Chat/components/ClarificationQuestionsWidget/ClarificationQuestionsWidget.tsx +++ b/autogpt_platform/frontend/src/components/contextual/Chat/components/ClarificationQuestionsWidget/ClarificationQuestionsWidget.tsx @@ -6,7 +6,7 @@ import { Input } from "@/components/atoms/Input/Input"; import { Text } from "@/components/atoms/Text/Text"; import { cn } from "@/lib/utils"; import { CheckCircleIcon, QuestionIcon } from "@phosphor-icons/react"; -import { useState } from "react"; +import { useState, useEffect, useRef } from "react"; export interface ClarifyingQuestion { question: string; @@ -17,39 +17,96 @@ export interface ClarifyingQuestion { interface Props { questions: ClarifyingQuestion[]; message: string; + sessionId?: string; onSubmitAnswers: (answers: Record) => void; onCancel?: () => void; + isAnswered?: boolean; className?: string; } +function getStorageKey(sessionId?: string): string | null { + if (!sessionId) return null; + return `clarification_answers_${sessionId}`; +} + export function ClarificationQuestionsWidget({ questions, message, + sessionId, onSubmitAnswers, onCancel, + isAnswered = false, className, }: Props) { const [answers, setAnswers] = useState>({}); const [isSubmitted, setIsSubmitted] = useState(false); + const lastSessionIdRef = useRef(undefined); + + useEffect(() => { + const storageKey = getStorageKey(sessionId); + if (!storageKey) { + setAnswers({}); + setIsSubmitted(false); + lastSessionIdRef.current = sessionId; + return; + } + + try { + const saved = localStorage.getItem(storageKey); + if (saved) { + const parsed = JSON.parse(saved) as Record; + setAnswers(parsed); + } else { + setAnswers({}); + } + setIsSubmitted(false); + } catch { + setAnswers({}); + setIsSubmitted(false); + } + lastSessionIdRef.current = sessionId; + }, [sessionId]); + + useEffect(() => { + if (lastSessionIdRef.current !== sessionId) { + return; + } + const storageKey = getStorageKey(sessionId); + if (!storageKey) return; + + const hasAnswers = Object.values(answers).some((v) => v.trim()); + try { + if (hasAnswers) { + localStorage.setItem(storageKey, JSON.stringify(answers)); + } else { + localStorage.removeItem(storageKey); + } + } catch {} + }, [answers, sessionId]); function handleAnswerChange(keyword: string, value: string) { setAnswers((prev) => ({ ...prev, [keyword]: value })); } function handleSubmit() { - // Check if all questions are answered const allAnswered = questions.every((q) => answers[q.keyword]?.trim()); if (!allAnswered) { return; } setIsSubmitted(true); onSubmitAnswers(answers); + + const storageKey = getStorageKey(sessionId); + try { + if (storageKey) { + localStorage.removeItem(storageKey); + } + } catch {} } const allAnswered = questions.every((q) => answers[q.keyword]?.trim()); - // Show submitted state after answers are submitted - if (isSubmitted) { + if (isAnswered || isSubmitted) { return (
; - if (response.error) return stripInternalReasoning(String(response.error)); if (response.message) return stripInternalReasoning(String(response.message)); + if (response.error) return stripInternalReasoning(String(response.error)); } return "An error occurred"; } @@ -363,8 +363,8 @@ export function formatToolResponse(result: unknown, toolName: string): string { case "error": const errorMsg = - (response.error as string) || response.message || "An error occurred"; - return `Error: ${errorMsg}`; + (response.message as string) || response.error || "An error occurred"; + return stripInternalReasoning(String(errorMsg)); case "no_results": const suggestions = (response.suggestions as string[]) || []; diff --git a/autogpt_platform/frontend/src/lib/autogpt-server-api/types.ts b/autogpt_platform/frontend/src/lib/autogpt-server-api/types.ts index 2d583d2062..74855f5e28 100644 --- a/autogpt_platform/frontend/src/lib/autogpt-server-api/types.ts +++ b/autogpt_platform/frontend/src/lib/autogpt-server-api/types.ts @@ -516,7 +516,7 @@ export type GraphValidationErrorResponse = { /* *** LIBRARY *** */ -/* Mirror of backend/server/v2/library/model.py:LibraryAgent */ +/* Mirror of backend/api/features/library/model.py:LibraryAgent */ export type LibraryAgent = { id: LibraryAgentID; graph_id: GraphID; @@ -616,7 +616,7 @@ export enum LibraryAgentSortEnum { /* *** CREDENTIALS *** */ -/* Mirror of backend/server/integrations/router.py:CredentialsMetaResponse */ +/* Mirror of backend/api/features/integrations/router.py:CredentialsMetaResponse */ export type CredentialsMetaResponse = { id: string; provider: CredentialsProviderName; @@ -628,13 +628,13 @@ export type CredentialsMetaResponse = { is_system?: boolean; }; -/* Mirror of backend/server/integrations/router.py:CredentialsDeletionResponse */ +/* Mirror of backend/api/features/integrations/router.py:CredentialsDeletionResponse */ export type CredentialsDeleteResponse = { deleted: true; revoked: boolean | null; }; -/* Mirror of backend/server/integrations/router.py:CredentialsDeletionNeedsConfirmationResponse */ +/* Mirror of backend/api/features/integrations/router.py:CredentialsDeletionNeedsConfirmationResponse */ export type CredentialsDeleteNeedConfirmationResponse = { deleted: false; need_confirmation: true; @@ -888,7 +888,7 @@ export type Schedule = { export type ScheduleID = Brand; -/* Mirror of backend/server/routers/v1.py:ScheduleCreationRequest */ +/* Mirror of backend/api/features/v1.py:ScheduleCreationRequest */ export type ScheduleCreatable = { graph_id: GraphID; graph_version: number; diff --git a/autogpt_platform/frontend/src/tests/library.spec.ts b/autogpt_platform/frontend/src/tests/library.spec.ts index 1972e94522..52941785e3 100644 --- a/autogpt_platform/frontend/src/tests/library.spec.ts +++ b/autogpt_platform/frontend/src/tests/library.spec.ts @@ -59,12 +59,13 @@ test.describe("Library", () => { }); test("pagination works correctly", async ({ page }, testInfo) => { - test.setTimeout(testInfo.timeout * 3); // Increase timeout for pagination operations + test.setTimeout(testInfo.timeout * 3); await page.goto("/library"); + const PAGE_SIZE = 20; const paginationResult = await libraryPage.testPagination(); - if (paginationResult.initialCount >= 10) { + if (paginationResult.initialCount >= PAGE_SIZE) { expect(paginationResult.finalCount).toBeGreaterThanOrEqual( paginationResult.initialCount, ); @@ -133,7 +134,10 @@ test.describe("Library", () => { test.expect(clearedSearchValue).toBe(""); }); - test("pagination while searching works correctly", async ({ page }) => { + test("pagination while searching works correctly", async ({ + page, + }, testInfo) => { + test.setTimeout(testInfo.timeout * 3); await page.goto("/library"); const allAgents = await libraryPage.getAgents(); @@ -152,9 +156,10 @@ test.describe("Library", () => { ); expect(matchingResults.length).toEqual(initialSearchResults.length); + const PAGE_SIZE = 20; const searchPaginationResult = await libraryPage.testPagination(); - if (searchPaginationResult.initialCount >= 10) { + if (searchPaginationResult.initialCount >= PAGE_SIZE) { expect(searchPaginationResult.finalCount).toBeGreaterThanOrEqual( searchPaginationResult.initialCount, ); diff --git a/autogpt_platform/frontend/src/tests/marketplace-creator.spec.ts b/autogpt_platform/frontend/src/tests/marketplace-creator.spec.ts index 3558f0672c..a41b652afb 100644 --- a/autogpt_platform/frontend/src/tests/marketplace-creator.spec.ts +++ b/autogpt_platform/frontend/src/tests/marketplace-creator.spec.ts @@ -69,9 +69,12 @@ test.describe("Marketplace Creator Page – Basic Functionality", () => { await marketplacePage.getFirstCreatorProfile(page); await firstCreatorProfile.click(); await page.waitForURL("**/marketplace/creator/**"); + await page.waitForLoadState("networkidle").catch(() => {}); + const firstAgent = page .locator('[data-testid="store-card"]:visible') .first(); + await firstAgent.waitFor({ state: "visible", timeout: 30000 }); await firstAgent.click(); await page.waitForURL("**/marketplace/agent/**"); diff --git a/autogpt_platform/frontend/src/tests/marketplace.spec.ts b/autogpt_platform/frontend/src/tests/marketplace.spec.ts index 774713dc82..44d89bf351 100644 --- a/autogpt_platform/frontend/src/tests/marketplace.spec.ts +++ b/autogpt_platform/frontend/src/tests/marketplace.spec.ts @@ -77,7 +77,6 @@ test.describe("Marketplace – Basic Functionality", () => { const firstFeaturedAgent = await marketplacePage.getFirstFeaturedAgent(page); - await firstFeaturedAgent.waitFor({ state: "visible" }); await firstFeaturedAgent.click(); await page.waitForURL("**/marketplace/agent/**"); await matchesUrl(page, /\/marketplace\/agent\/.+/); @@ -116,7 +115,15 @@ test.describe("Marketplace – Basic Functionality", () => { const searchTerm = page.getByText("DummyInput").first(); await isVisible(searchTerm); - await page.waitForTimeout(10000); + await page.waitForLoadState("networkidle").catch(() => {}); + + await page + .waitForFunction( + () => + document.querySelectorAll('[data-testid="store-card"]').length > 0, + { timeout: 15000 }, + ) + .catch(() => console.log("No search results appeared within timeout")); const results = await marketplacePage.getSearchResultsCount(page); expect(results).toBeGreaterThan(0); diff --git a/autogpt_platform/frontend/src/tests/pages/library.page.ts b/autogpt_platform/frontend/src/tests/pages/library.page.ts index 3a7695ec3a..03e98598b4 100644 --- a/autogpt_platform/frontend/src/tests/pages/library.page.ts +++ b/autogpt_platform/frontend/src/tests/pages/library.page.ts @@ -300,21 +300,27 @@ export class LibraryPage extends BasePage { async scrollToLoadMore(): Promise { console.log(`scrolling to load more agents`); - // Get initial agent count - const initialCount = await this.getAgentCount(); - console.log(`Initial agent count: ${initialCount}`); + const initialCount = await this.getAgentCountByListLength(); + console.log(`Initial agent count (DOM cards): ${initialCount}`); - // Scroll down to trigger pagination await this.scrollToBottom(); - // Wait for potential new agents to load - await this.page.waitForTimeout(2000); + await this.page + .waitForLoadState("networkidle", { timeout: 10000 }) + .catch(() => console.log("Network idle timeout, continuing...")); - // Check if more agents loaded - const newCount = await this.getAgentCount(); - console.log(`New agent count after scroll: ${newCount}`); + await this.page + .waitForFunction( + (prevCount) => + document.querySelectorAll('[data-testid="library-agent-card"]') + .length > prevCount, + initialCount, + { timeout: 5000 }, + ) + .catch(() => {}); - return; + const newCount = await this.getAgentCountByListLength(); + console.log(`New agent count after scroll (DOM cards): ${newCount}`); } async testPagination(): Promise<{ diff --git a/autogpt_platform/frontend/src/tests/pages/marketplace.page.ts b/autogpt_platform/frontend/src/tests/pages/marketplace.page.ts index 20f60c371a..115a7b2f12 100644 --- a/autogpt_platform/frontend/src/tests/pages/marketplace.page.ts +++ b/autogpt_platform/frontend/src/tests/pages/marketplace.page.ts @@ -9,6 +9,7 @@ export class MarketplacePage extends BasePage { async goto(page: Page) { await page.goto("/marketplace"); + await page.waitForLoadState("networkidle").catch(() => {}); } async getMarketplaceTitle(page: Page) { @@ -109,16 +110,24 @@ export class MarketplacePage extends BasePage { async getFirstFeaturedAgent(page: Page) { const { getId } = getSelectors(page); - return getId("featured-store-card").first(); + const card = getId("featured-store-card").first(); + await card.waitFor({ state: "visible", timeout: 30000 }); + return card; } async getFirstTopAgent() { - return this.page.locator('[data-testid="store-card"]:visible').first(); + const card = this.page + .locator('[data-testid="store-card"]:visible') + .first(); + await card.waitFor({ state: "visible", timeout: 30000 }); + return card; } async getFirstCreatorProfile(page: Page) { const { getId } = getSelectors(page); - return getId("creator-card").first(); + const card = getId("creator-card").first(); + await card.waitFor({ state: "visible", timeout: 30000 }); + return card; } async getSearchResultsCount(page: Page) { diff --git a/docs/integrations/block-integrations/llm.md b/docs/integrations/block-integrations/llm.md index f4d69b912b..6a0a9e0987 100644 --- a/docs/integrations/block-integrations/llm.md +++ b/docs/integrations/block-integrations/llm.md @@ -65,7 +65,7 @@ The result routes data to yes_output or no_output, enabling intelligent branchin | condition | A plaintext English description of the condition to evaluate | str | Yes | | yes_value | (Optional) Value to output if the condition is true. If not provided, input_value will be used. | Yes Value | No | | no_value | (Optional) Value to output if the condition is false. If not provided, input_value will be used. | No Value | No | -| model | The language model to use for evaluating the condition. | "o3-mini" \| "o3-2025-04-16" \| "o1" \| "o1-mini" \| "gpt-5.2-2025-12-11" \| "gpt-5.1-2025-11-13" \| "gpt-5-2025-08-07" \| "gpt-5-mini-2025-08-07" \| "gpt-5-nano-2025-08-07" \| "gpt-5-chat-latest" \| "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "gpt-4o-mini" \| "gpt-4o" \| "gpt-4-turbo" \| "gpt-3.5-turbo" \| "claude-opus-4-1-20250805" \| "claude-opus-4-20250514" \| "claude-sonnet-4-20250514" \| "claude-opus-4-5-20251101" \| "claude-sonnet-4-5-20250929" \| "claude-haiku-4-5-20251001" \| "claude-3-7-sonnet-20250219" \| "claude-3-haiku-20240307" \| "Qwen/Qwen2.5-72B-Instruct-Turbo" \| "nvidia/llama-3.1-nemotron-70b-instruct" \| "meta-llama/Llama-3.3-70B-Instruct-Turbo" \| "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" \| "meta-llama/Llama-3.2-3B-Instruct-Turbo" \| "llama-3.3-70b-versatile" \| "llama-3.1-8b-instant" \| "llama3.3" \| "llama3.2" \| "llama3" \| "llama3.1:405b" \| "dolphin-mistral:latest" \| "openai/gpt-oss-120b" \| "openai/gpt-oss-20b" \| "google/gemini-2.5-pro-preview-03-25" \| "google/gemini-3-pro-preview" \| "google/gemini-2.5-flash" \| "google/gemini-2.0-flash-001" \| "google/gemini-2.5-flash-lite-preview-06-17" \| "google/gemini-2.0-flash-lite-001" \| "mistralai/mistral-nemo" \| "cohere/command-r-08-2024" \| "cohere/command-r-plus-08-2024" \| "deepseek/deepseek-chat" \| "deepseek/deepseek-r1-0528" \| "perplexity/sonar" \| "perplexity/sonar-pro" \| "perplexity/sonar-deep-research" \| "nousresearch/hermes-3-llama-3.1-405b" \| "nousresearch/hermes-3-llama-3.1-70b" \| "amazon/nova-lite-v1" \| "amazon/nova-micro-v1" \| "amazon/nova-pro-v1" \| "microsoft/wizardlm-2-8x22b" \| "gryphe/mythomax-l2-13b" \| "meta-llama/llama-4-scout" \| "meta-llama/llama-4-maverick" \| "x-ai/grok-4" \| "x-ai/grok-4-fast" \| "x-ai/grok-4.1-fast" \| "x-ai/grok-code-fast-1" \| "moonshotai/kimi-k2" \| "qwen/qwen3-235b-a22b-thinking-2507" \| "qwen/qwen3-coder" \| "Llama-4-Scout-17B-16E-Instruct-FP8" \| "Llama-4-Maverick-17B-128E-Instruct-FP8" \| "Llama-3.3-8B-Instruct" \| "Llama-3.3-70B-Instruct" \| "v0-1.5-md" \| "v0-1.5-lg" \| "v0-1.0-md" | No | +| model | The language model to use for evaluating the condition. | "o3-mini" \| "o3-2025-04-16" \| "o1" \| "o1-mini" \| "gpt-5.2-2025-12-11" \| "gpt-5.1-2025-11-13" \| "gpt-5-2025-08-07" \| "gpt-5-mini-2025-08-07" \| "gpt-5-nano-2025-08-07" \| "gpt-5-chat-latest" \| "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "gpt-4o-mini" \| "gpt-4o" \| "gpt-4-turbo" \| "gpt-3.5-turbo" \| "claude-opus-4-1-20250805" \| "claude-opus-4-20250514" \| "claude-sonnet-4-20250514" \| "claude-opus-4-5-20251101" \| "claude-sonnet-4-5-20250929" \| "claude-haiku-4-5-20251001" \| "claude-3-haiku-20240307" \| "Qwen/Qwen2.5-72B-Instruct-Turbo" \| "nvidia/llama-3.1-nemotron-70b-instruct" \| "meta-llama/Llama-3.3-70B-Instruct-Turbo" \| "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" \| "meta-llama/Llama-3.2-3B-Instruct-Turbo" \| "llama-3.3-70b-versatile" \| "llama-3.1-8b-instant" \| "llama3.3" \| "llama3.2" \| "llama3" \| "llama3.1:405b" \| "dolphin-mistral:latest" \| "openai/gpt-oss-120b" \| "openai/gpt-oss-20b" \| "google/gemini-2.5-pro-preview-03-25" \| "google/gemini-3-pro-preview" \| "google/gemini-2.5-flash" \| "google/gemini-2.0-flash-001" \| "google/gemini-2.5-flash-lite-preview-06-17" \| "google/gemini-2.0-flash-lite-001" \| "mistralai/mistral-nemo" \| "cohere/command-r-08-2024" \| "cohere/command-r-plus-08-2024" \| "deepseek/deepseek-chat" \| "deepseek/deepseek-r1-0528" \| "perplexity/sonar" \| "perplexity/sonar-pro" \| "perplexity/sonar-deep-research" \| "nousresearch/hermes-3-llama-3.1-405b" \| "nousresearch/hermes-3-llama-3.1-70b" \| "amazon/nova-lite-v1" \| "amazon/nova-micro-v1" \| "amazon/nova-pro-v1" \| "microsoft/wizardlm-2-8x22b" \| "gryphe/mythomax-l2-13b" \| "meta-llama/llama-4-scout" \| "meta-llama/llama-4-maverick" \| "x-ai/grok-4" \| "x-ai/grok-4-fast" \| "x-ai/grok-4.1-fast" \| "x-ai/grok-code-fast-1" \| "moonshotai/kimi-k2" \| "qwen/qwen3-235b-a22b-thinking-2507" \| "qwen/qwen3-coder" \| "Llama-4-Scout-17B-16E-Instruct-FP8" \| "Llama-4-Maverick-17B-128E-Instruct-FP8" \| "Llama-3.3-8B-Instruct" \| "Llama-3.3-70B-Instruct" \| "v0-1.5-md" \| "v0-1.5-lg" \| "v0-1.0-md" | No | ### Outputs @@ -103,7 +103,7 @@ The block sends the entire conversation history to the chosen LLM, including sys |-------|-------------|------|----------| | prompt | The prompt to send to the language model. | str | No | | messages | List of messages in the conversation. | List[Any] | Yes | -| model | The language model to use for the conversation. | "o3-mini" \| "o3-2025-04-16" \| "o1" \| "o1-mini" \| "gpt-5.2-2025-12-11" \| "gpt-5.1-2025-11-13" \| "gpt-5-2025-08-07" \| "gpt-5-mini-2025-08-07" \| "gpt-5-nano-2025-08-07" \| "gpt-5-chat-latest" \| "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "gpt-4o-mini" \| "gpt-4o" \| "gpt-4-turbo" \| "gpt-3.5-turbo" \| "claude-opus-4-1-20250805" \| "claude-opus-4-20250514" \| "claude-sonnet-4-20250514" \| "claude-opus-4-5-20251101" \| "claude-sonnet-4-5-20250929" \| "claude-haiku-4-5-20251001" \| "claude-3-7-sonnet-20250219" \| "claude-3-haiku-20240307" \| "Qwen/Qwen2.5-72B-Instruct-Turbo" \| "nvidia/llama-3.1-nemotron-70b-instruct" \| "meta-llama/Llama-3.3-70B-Instruct-Turbo" \| "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" \| "meta-llama/Llama-3.2-3B-Instruct-Turbo" \| "llama-3.3-70b-versatile" \| "llama-3.1-8b-instant" \| "llama3.3" \| "llama3.2" \| "llama3" \| "llama3.1:405b" \| "dolphin-mistral:latest" \| "openai/gpt-oss-120b" \| "openai/gpt-oss-20b" \| "google/gemini-2.5-pro-preview-03-25" \| "google/gemini-3-pro-preview" \| "google/gemini-2.5-flash" \| "google/gemini-2.0-flash-001" \| "google/gemini-2.5-flash-lite-preview-06-17" \| "google/gemini-2.0-flash-lite-001" \| "mistralai/mistral-nemo" \| "cohere/command-r-08-2024" \| "cohere/command-r-plus-08-2024" \| "deepseek/deepseek-chat" \| "deepseek/deepseek-r1-0528" \| "perplexity/sonar" \| "perplexity/sonar-pro" \| "perplexity/sonar-deep-research" \| "nousresearch/hermes-3-llama-3.1-405b" \| "nousresearch/hermes-3-llama-3.1-70b" \| "amazon/nova-lite-v1" \| "amazon/nova-micro-v1" \| "amazon/nova-pro-v1" \| "microsoft/wizardlm-2-8x22b" \| "gryphe/mythomax-l2-13b" \| "meta-llama/llama-4-scout" \| "meta-llama/llama-4-maverick" \| "x-ai/grok-4" \| "x-ai/grok-4-fast" \| "x-ai/grok-4.1-fast" \| "x-ai/grok-code-fast-1" \| "moonshotai/kimi-k2" \| "qwen/qwen3-235b-a22b-thinking-2507" \| "qwen/qwen3-coder" \| "Llama-4-Scout-17B-16E-Instruct-FP8" \| "Llama-4-Maverick-17B-128E-Instruct-FP8" \| "Llama-3.3-8B-Instruct" \| "Llama-3.3-70B-Instruct" \| "v0-1.5-md" \| "v0-1.5-lg" \| "v0-1.0-md" | No | +| model | The language model to use for the conversation. | "o3-mini" \| "o3-2025-04-16" \| "o1" \| "o1-mini" \| "gpt-5.2-2025-12-11" \| "gpt-5.1-2025-11-13" \| "gpt-5-2025-08-07" \| "gpt-5-mini-2025-08-07" \| "gpt-5-nano-2025-08-07" \| "gpt-5-chat-latest" \| "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "gpt-4o-mini" \| "gpt-4o" \| "gpt-4-turbo" \| "gpt-3.5-turbo" \| "claude-opus-4-1-20250805" \| "claude-opus-4-20250514" \| "claude-sonnet-4-20250514" \| "claude-opus-4-5-20251101" \| "claude-sonnet-4-5-20250929" \| "claude-haiku-4-5-20251001" \| "claude-3-haiku-20240307" \| "Qwen/Qwen2.5-72B-Instruct-Turbo" \| "nvidia/llama-3.1-nemotron-70b-instruct" \| "meta-llama/Llama-3.3-70B-Instruct-Turbo" \| "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" \| "meta-llama/Llama-3.2-3B-Instruct-Turbo" \| "llama-3.3-70b-versatile" \| "llama-3.1-8b-instant" \| "llama3.3" \| "llama3.2" \| "llama3" \| "llama3.1:405b" \| "dolphin-mistral:latest" \| "openai/gpt-oss-120b" \| "openai/gpt-oss-20b" \| "google/gemini-2.5-pro-preview-03-25" \| "google/gemini-3-pro-preview" \| "google/gemini-2.5-flash" \| "google/gemini-2.0-flash-001" \| "google/gemini-2.5-flash-lite-preview-06-17" \| "google/gemini-2.0-flash-lite-001" \| "mistralai/mistral-nemo" \| "cohere/command-r-08-2024" \| "cohere/command-r-plus-08-2024" \| "deepseek/deepseek-chat" \| "deepseek/deepseek-r1-0528" \| "perplexity/sonar" \| "perplexity/sonar-pro" \| "perplexity/sonar-deep-research" \| "nousresearch/hermes-3-llama-3.1-405b" \| "nousresearch/hermes-3-llama-3.1-70b" \| "amazon/nova-lite-v1" \| "amazon/nova-micro-v1" \| "amazon/nova-pro-v1" \| "microsoft/wizardlm-2-8x22b" \| "gryphe/mythomax-l2-13b" \| "meta-llama/llama-4-scout" \| "meta-llama/llama-4-maverick" \| "x-ai/grok-4" \| "x-ai/grok-4-fast" \| "x-ai/grok-4.1-fast" \| "x-ai/grok-code-fast-1" \| "moonshotai/kimi-k2" \| "qwen/qwen3-235b-a22b-thinking-2507" \| "qwen/qwen3-coder" \| "Llama-4-Scout-17B-16E-Instruct-FP8" \| "Llama-4-Maverick-17B-128E-Instruct-FP8" \| "Llama-3.3-8B-Instruct" \| "Llama-3.3-70B-Instruct" \| "v0-1.5-md" \| "v0-1.5-lg" \| "v0-1.0-md" | No | | max_tokens | The maximum number of tokens to generate in the chat completion. | int | No | | ollama_host | Ollama host for local models | str | No | @@ -257,7 +257,7 @@ The block formulates a prompt based on the given focus or source data, sends it |-------|-------------|------|----------| | focus | The focus of the list to generate. | str | No | | source_data | The data to generate the list from. | str | No | -| model | The language model to use for generating the list. | "o3-mini" \| "o3-2025-04-16" \| "o1" \| "o1-mini" \| "gpt-5.2-2025-12-11" \| "gpt-5.1-2025-11-13" \| "gpt-5-2025-08-07" \| "gpt-5-mini-2025-08-07" \| "gpt-5-nano-2025-08-07" \| "gpt-5-chat-latest" \| "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "gpt-4o-mini" \| "gpt-4o" \| "gpt-4-turbo" \| "gpt-3.5-turbo" \| "claude-opus-4-1-20250805" \| "claude-opus-4-20250514" \| "claude-sonnet-4-20250514" \| "claude-opus-4-5-20251101" \| "claude-sonnet-4-5-20250929" \| "claude-haiku-4-5-20251001" \| "claude-3-7-sonnet-20250219" \| "claude-3-haiku-20240307" \| "Qwen/Qwen2.5-72B-Instruct-Turbo" \| "nvidia/llama-3.1-nemotron-70b-instruct" \| "meta-llama/Llama-3.3-70B-Instruct-Turbo" \| "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" \| "meta-llama/Llama-3.2-3B-Instruct-Turbo" \| "llama-3.3-70b-versatile" \| "llama-3.1-8b-instant" \| "llama3.3" \| "llama3.2" \| "llama3" \| "llama3.1:405b" \| "dolphin-mistral:latest" \| "openai/gpt-oss-120b" \| "openai/gpt-oss-20b" \| "google/gemini-2.5-pro-preview-03-25" \| "google/gemini-3-pro-preview" \| "google/gemini-2.5-flash" \| "google/gemini-2.0-flash-001" \| "google/gemini-2.5-flash-lite-preview-06-17" \| "google/gemini-2.0-flash-lite-001" \| "mistralai/mistral-nemo" \| "cohere/command-r-08-2024" \| "cohere/command-r-plus-08-2024" \| "deepseek/deepseek-chat" \| "deepseek/deepseek-r1-0528" \| "perplexity/sonar" \| "perplexity/sonar-pro" \| "perplexity/sonar-deep-research" \| "nousresearch/hermes-3-llama-3.1-405b" \| "nousresearch/hermes-3-llama-3.1-70b" \| "amazon/nova-lite-v1" \| "amazon/nova-micro-v1" \| "amazon/nova-pro-v1" \| "microsoft/wizardlm-2-8x22b" \| "gryphe/mythomax-l2-13b" \| "meta-llama/llama-4-scout" \| "meta-llama/llama-4-maverick" \| "x-ai/grok-4" \| "x-ai/grok-4-fast" \| "x-ai/grok-4.1-fast" \| "x-ai/grok-code-fast-1" \| "moonshotai/kimi-k2" \| "qwen/qwen3-235b-a22b-thinking-2507" \| "qwen/qwen3-coder" \| "Llama-4-Scout-17B-16E-Instruct-FP8" \| "Llama-4-Maverick-17B-128E-Instruct-FP8" \| "Llama-3.3-8B-Instruct" \| "Llama-3.3-70B-Instruct" \| "v0-1.5-md" \| "v0-1.5-lg" \| "v0-1.0-md" | No | +| model | The language model to use for generating the list. | "o3-mini" \| "o3-2025-04-16" \| "o1" \| "o1-mini" \| "gpt-5.2-2025-12-11" \| "gpt-5.1-2025-11-13" \| "gpt-5-2025-08-07" \| "gpt-5-mini-2025-08-07" \| "gpt-5-nano-2025-08-07" \| "gpt-5-chat-latest" \| "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "gpt-4o-mini" \| "gpt-4o" \| "gpt-4-turbo" \| "gpt-3.5-turbo" \| "claude-opus-4-1-20250805" \| "claude-opus-4-20250514" \| "claude-sonnet-4-20250514" \| "claude-opus-4-5-20251101" \| "claude-sonnet-4-5-20250929" \| "claude-haiku-4-5-20251001" \| "claude-3-haiku-20240307" \| "Qwen/Qwen2.5-72B-Instruct-Turbo" \| "nvidia/llama-3.1-nemotron-70b-instruct" \| "meta-llama/Llama-3.3-70B-Instruct-Turbo" \| "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" \| "meta-llama/Llama-3.2-3B-Instruct-Turbo" \| "llama-3.3-70b-versatile" \| "llama-3.1-8b-instant" \| "llama3.3" \| "llama3.2" \| "llama3" \| "llama3.1:405b" \| "dolphin-mistral:latest" \| "openai/gpt-oss-120b" \| "openai/gpt-oss-20b" \| "google/gemini-2.5-pro-preview-03-25" \| "google/gemini-3-pro-preview" \| "google/gemini-2.5-flash" \| "google/gemini-2.0-flash-001" \| "google/gemini-2.5-flash-lite-preview-06-17" \| "google/gemini-2.0-flash-lite-001" \| "mistralai/mistral-nemo" \| "cohere/command-r-08-2024" \| "cohere/command-r-plus-08-2024" \| "deepseek/deepseek-chat" \| "deepseek/deepseek-r1-0528" \| "perplexity/sonar" \| "perplexity/sonar-pro" \| "perplexity/sonar-deep-research" \| "nousresearch/hermes-3-llama-3.1-405b" \| "nousresearch/hermes-3-llama-3.1-70b" \| "amazon/nova-lite-v1" \| "amazon/nova-micro-v1" \| "amazon/nova-pro-v1" \| "microsoft/wizardlm-2-8x22b" \| "gryphe/mythomax-l2-13b" \| "meta-llama/llama-4-scout" \| "meta-llama/llama-4-maverick" \| "x-ai/grok-4" \| "x-ai/grok-4-fast" \| "x-ai/grok-4.1-fast" \| "x-ai/grok-code-fast-1" \| "moonshotai/kimi-k2" \| "qwen/qwen3-235b-a22b-thinking-2507" \| "qwen/qwen3-coder" \| "Llama-4-Scout-17B-16E-Instruct-FP8" \| "Llama-4-Maverick-17B-128E-Instruct-FP8" \| "Llama-3.3-8B-Instruct" \| "Llama-3.3-70B-Instruct" \| "v0-1.5-md" \| "v0-1.5-lg" \| "v0-1.0-md" | No | | max_retries | Maximum number of retries for generating a valid list. | int | No | | force_json_output | Whether to force the LLM to produce a JSON-only response. This can increase the block's reliability, but may also reduce the quality of the response because it prohibits the LLM from reasoning before providing its JSON response. | bool | No | | max_tokens | The maximum number of tokens to generate in the chat completion. | int | No | @@ -424,7 +424,7 @@ The block sends the input prompt to a chosen LLM, along with any system prompts | prompt | The prompt to send to the language model. | str | Yes | | expected_format | Expected format of the response. If provided, the response will be validated against this format. The keys should be the expected fields in the response, and the values should be the description of the field. | Dict[str, str] | Yes | | list_result | Whether the response should be a list of objects in the expected format. | bool | No | -| model | The language model to use for answering the prompt. | "o3-mini" \| "o3-2025-04-16" \| "o1" \| "o1-mini" \| "gpt-5.2-2025-12-11" \| "gpt-5.1-2025-11-13" \| "gpt-5-2025-08-07" \| "gpt-5-mini-2025-08-07" \| "gpt-5-nano-2025-08-07" \| "gpt-5-chat-latest" \| "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "gpt-4o-mini" \| "gpt-4o" \| "gpt-4-turbo" \| "gpt-3.5-turbo" \| "claude-opus-4-1-20250805" \| "claude-opus-4-20250514" \| "claude-sonnet-4-20250514" \| "claude-opus-4-5-20251101" \| "claude-sonnet-4-5-20250929" \| "claude-haiku-4-5-20251001" \| "claude-3-7-sonnet-20250219" \| "claude-3-haiku-20240307" \| "Qwen/Qwen2.5-72B-Instruct-Turbo" \| "nvidia/llama-3.1-nemotron-70b-instruct" \| "meta-llama/Llama-3.3-70B-Instruct-Turbo" \| "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" \| "meta-llama/Llama-3.2-3B-Instruct-Turbo" \| "llama-3.3-70b-versatile" \| "llama-3.1-8b-instant" \| "llama3.3" \| "llama3.2" \| "llama3" \| "llama3.1:405b" \| "dolphin-mistral:latest" \| "openai/gpt-oss-120b" \| "openai/gpt-oss-20b" \| "google/gemini-2.5-pro-preview-03-25" \| "google/gemini-3-pro-preview" \| "google/gemini-2.5-flash" \| "google/gemini-2.0-flash-001" \| "google/gemini-2.5-flash-lite-preview-06-17" \| "google/gemini-2.0-flash-lite-001" \| "mistralai/mistral-nemo" \| "cohere/command-r-08-2024" \| "cohere/command-r-plus-08-2024" \| "deepseek/deepseek-chat" \| "deepseek/deepseek-r1-0528" \| "perplexity/sonar" \| "perplexity/sonar-pro" \| "perplexity/sonar-deep-research" \| "nousresearch/hermes-3-llama-3.1-405b" \| "nousresearch/hermes-3-llama-3.1-70b" \| "amazon/nova-lite-v1" \| "amazon/nova-micro-v1" \| "amazon/nova-pro-v1" \| "microsoft/wizardlm-2-8x22b" \| "gryphe/mythomax-l2-13b" \| "meta-llama/llama-4-scout" \| "meta-llama/llama-4-maverick" \| "x-ai/grok-4" \| "x-ai/grok-4-fast" \| "x-ai/grok-4.1-fast" \| "x-ai/grok-code-fast-1" \| "moonshotai/kimi-k2" \| "qwen/qwen3-235b-a22b-thinking-2507" \| "qwen/qwen3-coder" \| "Llama-4-Scout-17B-16E-Instruct-FP8" \| "Llama-4-Maverick-17B-128E-Instruct-FP8" \| "Llama-3.3-8B-Instruct" \| "Llama-3.3-70B-Instruct" \| "v0-1.5-md" \| "v0-1.5-lg" \| "v0-1.0-md" | No | +| model | The language model to use for answering the prompt. | "o3-mini" \| "o3-2025-04-16" \| "o1" \| "o1-mini" \| "gpt-5.2-2025-12-11" \| "gpt-5.1-2025-11-13" \| "gpt-5-2025-08-07" \| "gpt-5-mini-2025-08-07" \| "gpt-5-nano-2025-08-07" \| "gpt-5-chat-latest" \| "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "gpt-4o-mini" \| "gpt-4o" \| "gpt-4-turbo" \| "gpt-3.5-turbo" \| "claude-opus-4-1-20250805" \| "claude-opus-4-20250514" \| "claude-sonnet-4-20250514" \| "claude-opus-4-5-20251101" \| "claude-sonnet-4-5-20250929" \| "claude-haiku-4-5-20251001" \| "claude-3-haiku-20240307" \| "Qwen/Qwen2.5-72B-Instruct-Turbo" \| "nvidia/llama-3.1-nemotron-70b-instruct" \| "meta-llama/Llama-3.3-70B-Instruct-Turbo" \| "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" \| "meta-llama/Llama-3.2-3B-Instruct-Turbo" \| "llama-3.3-70b-versatile" \| "llama-3.1-8b-instant" \| "llama3.3" \| "llama3.2" \| "llama3" \| "llama3.1:405b" \| "dolphin-mistral:latest" \| "openai/gpt-oss-120b" \| "openai/gpt-oss-20b" \| "google/gemini-2.5-pro-preview-03-25" \| "google/gemini-3-pro-preview" \| "google/gemini-2.5-flash" \| "google/gemini-2.0-flash-001" \| "google/gemini-2.5-flash-lite-preview-06-17" \| "google/gemini-2.0-flash-lite-001" \| "mistralai/mistral-nemo" \| "cohere/command-r-08-2024" \| "cohere/command-r-plus-08-2024" \| "deepseek/deepseek-chat" \| "deepseek/deepseek-r1-0528" \| "perplexity/sonar" \| "perplexity/sonar-pro" \| "perplexity/sonar-deep-research" \| "nousresearch/hermes-3-llama-3.1-405b" \| "nousresearch/hermes-3-llama-3.1-70b" \| "amazon/nova-lite-v1" \| "amazon/nova-micro-v1" \| "amazon/nova-pro-v1" \| "microsoft/wizardlm-2-8x22b" \| "gryphe/mythomax-l2-13b" \| "meta-llama/llama-4-scout" \| "meta-llama/llama-4-maverick" \| "x-ai/grok-4" \| "x-ai/grok-4-fast" \| "x-ai/grok-4.1-fast" \| "x-ai/grok-code-fast-1" \| "moonshotai/kimi-k2" \| "qwen/qwen3-235b-a22b-thinking-2507" \| "qwen/qwen3-coder" \| "Llama-4-Scout-17B-16E-Instruct-FP8" \| "Llama-4-Maverick-17B-128E-Instruct-FP8" \| "Llama-3.3-8B-Instruct" \| "Llama-3.3-70B-Instruct" \| "v0-1.5-md" \| "v0-1.5-lg" \| "v0-1.0-md" | No | | force_json_output | Whether to force the LLM to produce a JSON-only response. This can increase the block's reliability, but may also reduce the quality of the response because it prohibits the LLM from reasoning before providing its JSON response. | bool | No | | sys_prompt | The system prompt to provide additional context to the model. | str | No | | conversation_history | The conversation history to provide context for the prompt. | List[Dict[str, Any]] | No | @@ -464,7 +464,7 @@ The block sends the input prompt to a chosen LLM, processes the response, and re | Input | Description | Type | Required | |-------|-------------|------|----------| | prompt | The prompt to send to the language model. You can use any of the {keys} from Prompt Values to fill in the prompt with values from the prompt values dictionary by putting them in curly braces. | str | Yes | -| model | The language model to use for answering the prompt. | "o3-mini" \| "o3-2025-04-16" \| "o1" \| "o1-mini" \| "gpt-5.2-2025-12-11" \| "gpt-5.1-2025-11-13" \| "gpt-5-2025-08-07" \| "gpt-5-mini-2025-08-07" \| "gpt-5-nano-2025-08-07" \| "gpt-5-chat-latest" \| "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "gpt-4o-mini" \| "gpt-4o" \| "gpt-4-turbo" \| "gpt-3.5-turbo" \| "claude-opus-4-1-20250805" \| "claude-opus-4-20250514" \| "claude-sonnet-4-20250514" \| "claude-opus-4-5-20251101" \| "claude-sonnet-4-5-20250929" \| "claude-haiku-4-5-20251001" \| "claude-3-7-sonnet-20250219" \| "claude-3-haiku-20240307" \| "Qwen/Qwen2.5-72B-Instruct-Turbo" \| "nvidia/llama-3.1-nemotron-70b-instruct" \| "meta-llama/Llama-3.3-70B-Instruct-Turbo" \| "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" \| "meta-llama/Llama-3.2-3B-Instruct-Turbo" \| "llama-3.3-70b-versatile" \| "llama-3.1-8b-instant" \| "llama3.3" \| "llama3.2" \| "llama3" \| "llama3.1:405b" \| "dolphin-mistral:latest" \| "openai/gpt-oss-120b" \| "openai/gpt-oss-20b" \| "google/gemini-2.5-pro-preview-03-25" \| "google/gemini-3-pro-preview" \| "google/gemini-2.5-flash" \| "google/gemini-2.0-flash-001" \| "google/gemini-2.5-flash-lite-preview-06-17" \| "google/gemini-2.0-flash-lite-001" \| "mistralai/mistral-nemo" \| "cohere/command-r-08-2024" \| "cohere/command-r-plus-08-2024" \| "deepseek/deepseek-chat" \| "deepseek/deepseek-r1-0528" \| "perplexity/sonar" \| "perplexity/sonar-pro" \| "perplexity/sonar-deep-research" \| "nousresearch/hermes-3-llama-3.1-405b" \| "nousresearch/hermes-3-llama-3.1-70b" \| "amazon/nova-lite-v1" \| "amazon/nova-micro-v1" \| "amazon/nova-pro-v1" \| "microsoft/wizardlm-2-8x22b" \| "gryphe/mythomax-l2-13b" \| "meta-llama/llama-4-scout" \| "meta-llama/llama-4-maverick" \| "x-ai/grok-4" \| "x-ai/grok-4-fast" \| "x-ai/grok-4.1-fast" \| "x-ai/grok-code-fast-1" \| "moonshotai/kimi-k2" \| "qwen/qwen3-235b-a22b-thinking-2507" \| "qwen/qwen3-coder" \| "Llama-4-Scout-17B-16E-Instruct-FP8" \| "Llama-4-Maverick-17B-128E-Instruct-FP8" \| "Llama-3.3-8B-Instruct" \| "Llama-3.3-70B-Instruct" \| "v0-1.5-md" \| "v0-1.5-lg" \| "v0-1.0-md" | No | +| model | The language model to use for answering the prompt. | "o3-mini" \| "o3-2025-04-16" \| "o1" \| "o1-mini" \| "gpt-5.2-2025-12-11" \| "gpt-5.1-2025-11-13" \| "gpt-5-2025-08-07" \| "gpt-5-mini-2025-08-07" \| "gpt-5-nano-2025-08-07" \| "gpt-5-chat-latest" \| "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "gpt-4o-mini" \| "gpt-4o" \| "gpt-4-turbo" \| "gpt-3.5-turbo" \| "claude-opus-4-1-20250805" \| "claude-opus-4-20250514" \| "claude-sonnet-4-20250514" \| "claude-opus-4-5-20251101" \| "claude-sonnet-4-5-20250929" \| "claude-haiku-4-5-20251001" \| "claude-3-haiku-20240307" \| "Qwen/Qwen2.5-72B-Instruct-Turbo" \| "nvidia/llama-3.1-nemotron-70b-instruct" \| "meta-llama/Llama-3.3-70B-Instruct-Turbo" \| "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" \| "meta-llama/Llama-3.2-3B-Instruct-Turbo" \| "llama-3.3-70b-versatile" \| "llama-3.1-8b-instant" \| "llama3.3" \| "llama3.2" \| "llama3" \| "llama3.1:405b" \| "dolphin-mistral:latest" \| "openai/gpt-oss-120b" \| "openai/gpt-oss-20b" \| "google/gemini-2.5-pro-preview-03-25" \| "google/gemini-3-pro-preview" \| "google/gemini-2.5-flash" \| "google/gemini-2.0-flash-001" \| "google/gemini-2.5-flash-lite-preview-06-17" \| "google/gemini-2.0-flash-lite-001" \| "mistralai/mistral-nemo" \| "cohere/command-r-08-2024" \| "cohere/command-r-plus-08-2024" \| "deepseek/deepseek-chat" \| "deepseek/deepseek-r1-0528" \| "perplexity/sonar" \| "perplexity/sonar-pro" \| "perplexity/sonar-deep-research" \| "nousresearch/hermes-3-llama-3.1-405b" \| "nousresearch/hermes-3-llama-3.1-70b" \| "amazon/nova-lite-v1" \| "amazon/nova-micro-v1" \| "amazon/nova-pro-v1" \| "microsoft/wizardlm-2-8x22b" \| "gryphe/mythomax-l2-13b" \| "meta-llama/llama-4-scout" \| "meta-llama/llama-4-maverick" \| "x-ai/grok-4" \| "x-ai/grok-4-fast" \| "x-ai/grok-4.1-fast" \| "x-ai/grok-code-fast-1" \| "moonshotai/kimi-k2" \| "qwen/qwen3-235b-a22b-thinking-2507" \| "qwen/qwen3-coder" \| "Llama-4-Scout-17B-16E-Instruct-FP8" \| "Llama-4-Maverick-17B-128E-Instruct-FP8" \| "Llama-3.3-8B-Instruct" \| "Llama-3.3-70B-Instruct" \| "v0-1.5-md" \| "v0-1.5-lg" \| "v0-1.0-md" | No | | sys_prompt | The system prompt to provide additional context to the model. | str | No | | retry | Number of times to retry the LLM call if the response does not match the expected format. | int | No | | prompt_values | Values used to fill in the prompt. The values can be used in the prompt by putting them in a double curly braces, e.g. {{variable_name}}. | Dict[str, str] | No | @@ -501,7 +501,7 @@ The block splits the input text into smaller chunks, sends each chunk to an LLM | Input | Description | Type | Required | |-------|-------------|------|----------| | text | The text to summarize. | str | Yes | -| model | The language model to use for summarizing the text. | "o3-mini" \| "o3-2025-04-16" \| "o1" \| "o1-mini" \| "gpt-5.2-2025-12-11" \| "gpt-5.1-2025-11-13" \| "gpt-5-2025-08-07" \| "gpt-5-mini-2025-08-07" \| "gpt-5-nano-2025-08-07" \| "gpt-5-chat-latest" \| "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "gpt-4o-mini" \| "gpt-4o" \| "gpt-4-turbo" \| "gpt-3.5-turbo" \| "claude-opus-4-1-20250805" \| "claude-opus-4-20250514" \| "claude-sonnet-4-20250514" \| "claude-opus-4-5-20251101" \| "claude-sonnet-4-5-20250929" \| "claude-haiku-4-5-20251001" \| "claude-3-7-sonnet-20250219" \| "claude-3-haiku-20240307" \| "Qwen/Qwen2.5-72B-Instruct-Turbo" \| "nvidia/llama-3.1-nemotron-70b-instruct" \| "meta-llama/Llama-3.3-70B-Instruct-Turbo" \| "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" \| "meta-llama/Llama-3.2-3B-Instruct-Turbo" \| "llama-3.3-70b-versatile" \| "llama-3.1-8b-instant" \| "llama3.3" \| "llama3.2" \| "llama3" \| "llama3.1:405b" \| "dolphin-mistral:latest" \| "openai/gpt-oss-120b" \| "openai/gpt-oss-20b" \| "google/gemini-2.5-pro-preview-03-25" \| "google/gemini-3-pro-preview" \| "google/gemini-2.5-flash" \| "google/gemini-2.0-flash-001" \| "google/gemini-2.5-flash-lite-preview-06-17" \| "google/gemini-2.0-flash-lite-001" \| "mistralai/mistral-nemo" \| "cohere/command-r-08-2024" \| "cohere/command-r-plus-08-2024" \| "deepseek/deepseek-chat" \| "deepseek/deepseek-r1-0528" \| "perplexity/sonar" \| "perplexity/sonar-pro" \| "perplexity/sonar-deep-research" \| "nousresearch/hermes-3-llama-3.1-405b" \| "nousresearch/hermes-3-llama-3.1-70b" \| "amazon/nova-lite-v1" \| "amazon/nova-micro-v1" \| "amazon/nova-pro-v1" \| "microsoft/wizardlm-2-8x22b" \| "gryphe/mythomax-l2-13b" \| "meta-llama/llama-4-scout" \| "meta-llama/llama-4-maverick" \| "x-ai/grok-4" \| "x-ai/grok-4-fast" \| "x-ai/grok-4.1-fast" \| "x-ai/grok-code-fast-1" \| "moonshotai/kimi-k2" \| "qwen/qwen3-235b-a22b-thinking-2507" \| "qwen/qwen3-coder" \| "Llama-4-Scout-17B-16E-Instruct-FP8" \| "Llama-4-Maverick-17B-128E-Instruct-FP8" \| "Llama-3.3-8B-Instruct" \| "Llama-3.3-70B-Instruct" \| "v0-1.5-md" \| "v0-1.5-lg" \| "v0-1.0-md" | No | +| model | The language model to use for summarizing the text. | "o3-mini" \| "o3-2025-04-16" \| "o1" \| "o1-mini" \| "gpt-5.2-2025-12-11" \| "gpt-5.1-2025-11-13" \| "gpt-5-2025-08-07" \| "gpt-5-mini-2025-08-07" \| "gpt-5-nano-2025-08-07" \| "gpt-5-chat-latest" \| "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "gpt-4o-mini" \| "gpt-4o" \| "gpt-4-turbo" \| "gpt-3.5-turbo" \| "claude-opus-4-1-20250805" \| "claude-opus-4-20250514" \| "claude-sonnet-4-20250514" \| "claude-opus-4-5-20251101" \| "claude-sonnet-4-5-20250929" \| "claude-haiku-4-5-20251001" \| "claude-3-haiku-20240307" \| "Qwen/Qwen2.5-72B-Instruct-Turbo" \| "nvidia/llama-3.1-nemotron-70b-instruct" \| "meta-llama/Llama-3.3-70B-Instruct-Turbo" \| "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" \| "meta-llama/Llama-3.2-3B-Instruct-Turbo" \| "llama-3.3-70b-versatile" \| "llama-3.1-8b-instant" \| "llama3.3" \| "llama3.2" \| "llama3" \| "llama3.1:405b" \| "dolphin-mistral:latest" \| "openai/gpt-oss-120b" \| "openai/gpt-oss-20b" \| "google/gemini-2.5-pro-preview-03-25" \| "google/gemini-3-pro-preview" \| "google/gemini-2.5-flash" \| "google/gemini-2.0-flash-001" \| "google/gemini-2.5-flash-lite-preview-06-17" \| "google/gemini-2.0-flash-lite-001" \| "mistralai/mistral-nemo" \| "cohere/command-r-08-2024" \| "cohere/command-r-plus-08-2024" \| "deepseek/deepseek-chat" \| "deepseek/deepseek-r1-0528" \| "perplexity/sonar" \| "perplexity/sonar-pro" \| "perplexity/sonar-deep-research" \| "nousresearch/hermes-3-llama-3.1-405b" \| "nousresearch/hermes-3-llama-3.1-70b" \| "amazon/nova-lite-v1" \| "amazon/nova-micro-v1" \| "amazon/nova-pro-v1" \| "microsoft/wizardlm-2-8x22b" \| "gryphe/mythomax-l2-13b" \| "meta-llama/llama-4-scout" \| "meta-llama/llama-4-maverick" \| "x-ai/grok-4" \| "x-ai/grok-4-fast" \| "x-ai/grok-4.1-fast" \| "x-ai/grok-code-fast-1" \| "moonshotai/kimi-k2" \| "qwen/qwen3-235b-a22b-thinking-2507" \| "qwen/qwen3-coder" \| "Llama-4-Scout-17B-16E-Instruct-FP8" \| "Llama-4-Maverick-17B-128E-Instruct-FP8" \| "Llama-3.3-8B-Instruct" \| "Llama-3.3-70B-Instruct" \| "v0-1.5-md" \| "v0-1.5-lg" \| "v0-1.0-md" | No | | focus | The topic to focus on in the summary | str | No | | style | The style of the summary to generate. | "concise" \| "detailed" \| "bullet points" \| "numbered list" | No | | max_tokens | The maximum number of tokens to generate in the chat completion. | int | No | @@ -763,7 +763,7 @@ Configure agent_mode_max_iterations to control loop behavior: 0 for single decis | Input | Description | Type | Required | |-------|-------------|------|----------| | prompt | The prompt to send to the language model. | str | Yes | -| model | The language model to use for answering the prompt. | "o3-mini" \| "o3-2025-04-16" \| "o1" \| "o1-mini" \| "gpt-5.2-2025-12-11" \| "gpt-5.1-2025-11-13" \| "gpt-5-2025-08-07" \| "gpt-5-mini-2025-08-07" \| "gpt-5-nano-2025-08-07" \| "gpt-5-chat-latest" \| "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "gpt-4o-mini" \| "gpt-4o" \| "gpt-4-turbo" \| "gpt-3.5-turbo" \| "claude-opus-4-1-20250805" \| "claude-opus-4-20250514" \| "claude-sonnet-4-20250514" \| "claude-opus-4-5-20251101" \| "claude-sonnet-4-5-20250929" \| "claude-haiku-4-5-20251001" \| "claude-3-7-sonnet-20250219" \| "claude-3-haiku-20240307" \| "Qwen/Qwen2.5-72B-Instruct-Turbo" \| "nvidia/llama-3.1-nemotron-70b-instruct" \| "meta-llama/Llama-3.3-70B-Instruct-Turbo" \| "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" \| "meta-llama/Llama-3.2-3B-Instruct-Turbo" \| "llama-3.3-70b-versatile" \| "llama-3.1-8b-instant" \| "llama3.3" \| "llama3.2" \| "llama3" \| "llama3.1:405b" \| "dolphin-mistral:latest" \| "openai/gpt-oss-120b" \| "openai/gpt-oss-20b" \| "google/gemini-2.5-pro-preview-03-25" \| "google/gemini-3-pro-preview" \| "google/gemini-2.5-flash" \| "google/gemini-2.0-flash-001" \| "google/gemini-2.5-flash-lite-preview-06-17" \| "google/gemini-2.0-flash-lite-001" \| "mistralai/mistral-nemo" \| "cohere/command-r-08-2024" \| "cohere/command-r-plus-08-2024" \| "deepseek/deepseek-chat" \| "deepseek/deepseek-r1-0528" \| "perplexity/sonar" \| "perplexity/sonar-pro" \| "perplexity/sonar-deep-research" \| "nousresearch/hermes-3-llama-3.1-405b" \| "nousresearch/hermes-3-llama-3.1-70b" \| "amazon/nova-lite-v1" \| "amazon/nova-micro-v1" \| "amazon/nova-pro-v1" \| "microsoft/wizardlm-2-8x22b" \| "gryphe/mythomax-l2-13b" \| "meta-llama/llama-4-scout" \| "meta-llama/llama-4-maverick" \| "x-ai/grok-4" \| "x-ai/grok-4-fast" \| "x-ai/grok-4.1-fast" \| "x-ai/grok-code-fast-1" \| "moonshotai/kimi-k2" \| "qwen/qwen3-235b-a22b-thinking-2507" \| "qwen/qwen3-coder" \| "Llama-4-Scout-17B-16E-Instruct-FP8" \| "Llama-4-Maverick-17B-128E-Instruct-FP8" \| "Llama-3.3-8B-Instruct" \| "Llama-3.3-70B-Instruct" \| "v0-1.5-md" \| "v0-1.5-lg" \| "v0-1.0-md" | No | +| model | The language model to use for answering the prompt. | "o3-mini" \| "o3-2025-04-16" \| "o1" \| "o1-mini" \| "gpt-5.2-2025-12-11" \| "gpt-5.1-2025-11-13" \| "gpt-5-2025-08-07" \| "gpt-5-mini-2025-08-07" \| "gpt-5-nano-2025-08-07" \| "gpt-5-chat-latest" \| "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "gpt-4o-mini" \| "gpt-4o" \| "gpt-4-turbo" \| "gpt-3.5-turbo" \| "claude-opus-4-1-20250805" \| "claude-opus-4-20250514" \| "claude-sonnet-4-20250514" \| "claude-opus-4-5-20251101" \| "claude-sonnet-4-5-20250929" \| "claude-haiku-4-5-20251001" \| "claude-3-haiku-20240307" \| "Qwen/Qwen2.5-72B-Instruct-Turbo" \| "nvidia/llama-3.1-nemotron-70b-instruct" \| "meta-llama/Llama-3.3-70B-Instruct-Turbo" \| "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo" \| "meta-llama/Llama-3.2-3B-Instruct-Turbo" \| "llama-3.3-70b-versatile" \| "llama-3.1-8b-instant" \| "llama3.3" \| "llama3.2" \| "llama3" \| "llama3.1:405b" \| "dolphin-mistral:latest" \| "openai/gpt-oss-120b" \| "openai/gpt-oss-20b" \| "google/gemini-2.5-pro-preview-03-25" \| "google/gemini-3-pro-preview" \| "google/gemini-2.5-flash" \| "google/gemini-2.0-flash-001" \| "google/gemini-2.5-flash-lite-preview-06-17" \| "google/gemini-2.0-flash-lite-001" \| "mistralai/mistral-nemo" \| "cohere/command-r-08-2024" \| "cohere/command-r-plus-08-2024" \| "deepseek/deepseek-chat" \| "deepseek/deepseek-r1-0528" \| "perplexity/sonar" \| "perplexity/sonar-pro" \| "perplexity/sonar-deep-research" \| "nousresearch/hermes-3-llama-3.1-405b" \| "nousresearch/hermes-3-llama-3.1-70b" \| "amazon/nova-lite-v1" \| "amazon/nova-micro-v1" \| "amazon/nova-pro-v1" \| "microsoft/wizardlm-2-8x22b" \| "gryphe/mythomax-l2-13b" \| "meta-llama/llama-4-scout" \| "meta-llama/llama-4-maverick" \| "x-ai/grok-4" \| "x-ai/grok-4-fast" \| "x-ai/grok-4.1-fast" \| "x-ai/grok-code-fast-1" \| "moonshotai/kimi-k2" \| "qwen/qwen3-235b-a22b-thinking-2507" \| "qwen/qwen3-coder" \| "Llama-4-Scout-17B-16E-Instruct-FP8" \| "Llama-4-Maverick-17B-128E-Instruct-FP8" \| "Llama-3.3-8B-Instruct" \| "Llama-3.3-70B-Instruct" \| "v0-1.5-md" \| "v0-1.5-lg" \| "v0-1.0-md" | No | | multiple_tool_calls | Whether to allow multiple tool calls in a single response. | bool | No | | sys_prompt | The system prompt to provide additional context to the model. | str | No | | conversation_history | The conversation history to provide context for the prompt. | List[Dict[str, Any]] | No | diff --git a/docs/integrations/block-integrations/stagehand/blocks.md b/docs/integrations/block-integrations/stagehand/blocks.md index dac0586fa2..cc201d092b 100644 --- a/docs/integrations/block-integrations/stagehand/blocks.md +++ b/docs/integrations/block-integrations/stagehand/blocks.md @@ -20,7 +20,7 @@ Configure timeouts for DOM settlement and page loading. Variables can be passed | Input | Description | Type | Required | |-------|-------------|------|----------| | browserbase_project_id | Browserbase project ID (required if using Browserbase) | str | Yes | -| model | LLM to use for Stagehand (provider is inferred) | "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "claude-3-7-sonnet-20250219" | No | +| model | LLM to use for Stagehand (provider is inferred) | "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "claude-sonnet-4-5-20250929" | No | | url | URL to navigate to. | str | Yes | | action | Action to perform. Suggested actions are: click, fill, type, press, scroll, select from dropdown. For multi-step actions, add an entry for each step. | List[str] | Yes | | variables | Variables to use in the action. Variables contains data you want the action to use. | Dict[str, str] | No | @@ -65,7 +65,7 @@ Supports searching within iframes and configurable timeouts for dynamic content | Input | Description | Type | Required | |-------|-------------|------|----------| | browserbase_project_id | Browserbase project ID (required if using Browserbase) | str | Yes | -| model | LLM to use for Stagehand (provider is inferred) | "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "claude-3-7-sonnet-20250219" | No | +| model | LLM to use for Stagehand (provider is inferred) | "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "claude-sonnet-4-5-20250929" | No | | url | URL to navigate to. | str | Yes | | instruction | Natural language description of elements or actions to discover. | str | Yes | | iframes | Whether to search within iframes. If True, Stagehand will search for actions within iframes. | bool | No | @@ -106,7 +106,7 @@ Use this to explore a page's interactive elements before building automated work | Input | Description | Type | Required | |-------|-------------|------|----------| | browserbase_project_id | Browserbase project ID (required if using Browserbase) | str | Yes | -| model | LLM to use for Stagehand (provider is inferred) | "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "claude-3-7-sonnet-20250219" | No | +| model | LLM to use for Stagehand (provider is inferred) | "gpt-4.1-2025-04-14" \| "gpt-4.1-mini-2025-04-14" \| "claude-sonnet-4-5-20250929" | No | | url | URL to navigate to. | str | Yes | | instruction | Natural language description of elements or actions to discover. | str | Yes | | iframes | Whether to search within iframes. If True, Stagehand will search for actions within iframes. | bool | No | diff --git a/docs/platform/contributing/oauth-integration-flow.md b/docs/platform/contributing/oauth-integration-flow.md index dbc7a54be5..f6c3f7fd17 100644 --- a/docs/platform/contributing/oauth-integration-flow.md +++ b/docs/platform/contributing/oauth-integration-flow.md @@ -25,7 +25,7 @@ This document focuses on the **API Integration OAuth flow** used for connecting ### 2. Backend API Trust Boundary - **Location**: Server-side FastAPI application - **Components**: - - Integration router (`/backend/backend/server/integrations/router.py`) + - Integration router (`/backend/backend/api/features/integrations/router.py`) - OAuth handlers (`/backend/backend/integrations/oauth/`) - Credentials store (`/backend/backend/integrations/credentials_store.py`) - **Trust Level**: Trusted - server-controlled environment diff --git a/docs/platform/create-basic-agent.md b/docs/platform/create-basic-agent.md index 7721fb9b9c..ffe654ba99 100644 --- a/docs/platform/create-basic-agent.md +++ b/docs/platform/create-basic-agent.md @@ -4,6 +4,28 @@ This guide walks through creating a simple question-answer AI agent using AutoGPT's visual builder. This is a basic example that can be expanded into more complex agents. +## **Prerequisites** + +### **Cloud-Hosted AutoGPT** +If you're using the cloud-hosted version at [agpt.co](https://agpt.co), you're ready to go! AI blocks come with **built-in credits** — no API keys required to get started. If you'd prefer to use your own API keys, you can add them via **Profile → Integrations**. + +### **Self-Hosted (Docker)** +If you're running AutoGPT locally with Docker, you'll need to add your own API keys to `autogpt_platform/backend/.env`: + +```bash +# Create or edit backend/.env +OPENAI_API_KEY=sk-your-key-here +ANTHROPIC_API_KEY=sk-ant-your-key-here +# Add other provider keys as needed +``` + +After adding keys, restart the services: +```bash +docker compose down && docker compose up -d +``` + +**Note:** The Calculator example below doesn't require any API credentials — it's a good way to test your setup before adding AI blocks. + ## **Example Agent: Q&A (with AI)** A step-by-step guide to creating a simple Q&A agent using input and output blocks. diff --git a/docs/platform/ollama.md b/docs/platform/ollama.md index 392bfabfe8..ecab9b8ae1 100644 --- a/docs/platform/ollama.md +++ b/docs/platform/ollama.md @@ -246,7 +246,7 @@ If you encounter any issues, verify that: ```bash ollama pull llama3.2 ``` -- If using a custom model, ensure it's added to the model list in `backend/server/model.py` +- If using a custom model, ensure it's added to the model list in `backend/api/model.py` #### Docker Issues - Ensure Docker daemon is running: