Merge branch 'dev' into otto/secrt-1887-content-based-detection

fix(binary-processor): add virus scanning + tests for offset loop
Security fix: Add scan_content_safe() call before writing files to prevent bypass of ClamAV scanning infrastructure. Malicious base64-encoded content matching a supported magic number would previously be saved without scanning. Test coverage: - Add tests for 4-byte aligned offset loop (handles marker bleed-in) - Add tests for PDF_BASE64_START/END marker expansion - Add tests verifying virus scanner integration - Mock virus scanner in existing process_binary_outputs tests Addresses review feedback from Claude sub-agents and ntindle.
2026-02-17 18:21:46 -05:00 · 2026-02-09 08:04:01 +00:00 · 2026-02-05 23:29:57 +00:00 · 2026-02-05 23:50:58 +01:00 · 2026-02-05 22:46:15 +00:00 · 2026-02-05 20:26:36 +00:00
563 changed files with 15863 additions and 36263 deletions
--- a/.dockerignore
+++ b/.dockerignore
@@ -5,13 +5,42 @@
 !docs/

 # Platform - Libs
-!autogpt_platform/autogpt_libs/
+!autogpt_platform/autogpt_libs/autogpt_libs/
+!autogpt_platform/autogpt_libs/pyproject.toml
+!autogpt_platform/autogpt_libs/poetry.lock
+!autogpt_platform/autogpt_libs/README.md

 # Platform - Backend
-!autogpt_platform/backend/
+!autogpt_platform/backend/backend/
+!autogpt_platform/backend/test/e2e_test_data.py
+!autogpt_platform/backend/migrations/
+!autogpt_platform/backend/schema.prisma
+!autogpt_platform/backend/pyproject.toml
+!autogpt_platform/backend/poetry.lock
+!autogpt_platform/backend/README.md
+!autogpt_platform/backend/.env
+!autogpt_platform/backend/gen_prisma_types_stub.py
+
+# Platform - Market
+!autogpt_platform/market/market/
+!autogpt_platform/market/scripts.py
+!autogpt_platform/market/schema.prisma
+!autogpt_platform/market/pyproject.toml
+!autogpt_platform/market/poetry.lock
+!autogpt_platform/market/README.md

 # Platform - Frontend
-!autogpt_platform/frontend/
+!autogpt_platform/frontend/src/
+!autogpt_platform/frontend/public/
+!autogpt_platform/frontend/scripts/
+!autogpt_platform/frontend/package.json
+!autogpt_platform/frontend/pnpm-lock.yaml
+!autogpt_platform/frontend/tsconfig.json
+!autogpt_platform/frontend/README.md
+## config
+!autogpt_platform/frontend/*.config.*
+!autogpt_platform/frontend/.env.*
+!autogpt_platform/frontend/.env

 # Classic - AutoGPT
 !classic/original_autogpt/autogpt/
@@ -35,38 +64,6 @@
 # Classic - Frontend
 !classic/frontend/build/web/

-# Explicitly re-ignore unwanted files from whitelisted directories
-# Note: These patterns MUST come after the whitelist rules to take effect
-
-# Hidden files and directories (but keep frontend .env files needed for build)
-**/.*
-!autogpt_platform/frontend/.env
-!autogpt_platform/frontend/.env.default
-!autogpt_platform/frontend/.env.production
-
-# Python artifacts
-**/__pycache__/
-**/*.pyc
-**/*.pyo
-**/.venv/
-**/.ruff_cache/
-**/.pytest_cache/
-**/.coverage
-**/htmlcov/
-
-# Node artifacts
-**/node_modules/
-**/.next/
-**/storybook-static/
-**/playwright-report/
-**/test-results/
-
-# Build artifacts
-**/dist/
-**/build/
-!autogpt_platform/frontend/src/**/build/
-**/target/
-
-# Logs and temp files
-**/*.log
-**/*.tmp
+# Explicitly re-ignore some folders
+.*
+**/__pycache__
--- a/.github/scripts/detect_overlaps.py
+++ b/.github/scripts/detect_overlaps.py
--- a/.github/workflows/claude-ci-failure-auto-fix.yml
+++ b/.github/workflows/claude-ci-failure-auto-fix.yml
@@ -22,7 +22,7 @@ jobs:
    runs-on: ubuntu-latest
    steps:
      - name: Checkout code
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
        with:
          ref: ${{ github.event.workflow_run.head_branch }}
          fetch-depth: 0
@@ -40,48 +40,6 @@ jobs:
          git checkout -b "$BRANCH_NAME"
          echo "branch_name=$BRANCH_NAME" >> $GITHUB_OUTPUT

-      # Backend Python/Poetry setup (so Claude can run linting/tests)
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: "3.11"
-
-      - name: Set up Python dependency cache
-        uses: actions/cache@v5
-        with:
-          path: ~/.cache/pypoetry
-          key: poetry-${{ runner.os }}-${{ hashFiles('autogpt_platform/backend/poetry.lock') }}
-
-      - name: Install Poetry
-        run: |
-          cd autogpt_platform/backend
-          HEAD_POETRY_VERSION=$(python3 ../../.github/workflows/scripts/get_package_version_from_lockfile.py poetry)
-          curl -sSL https://install.python-poetry.org | POETRY_VERSION=$HEAD_POETRY_VERSION python3 -
-          echo "$HOME/.local/bin" >> $GITHUB_PATH
-
-      - name: Install Python dependencies
-        working-directory: autogpt_platform/backend
-        run: poetry install
-
-      - name: Generate Prisma Client
-        working-directory: autogpt_platform/backend
-        run: poetry run prisma generate && poetry run gen-prisma-stub
-
-      # Frontend Node.js/pnpm setup (so Claude can run linting/tests)
-      - name: Enable corepack
-        run: corepack enable
-
-      - name: Set up Node.js
-        uses: actions/setup-node@v6
-        with:
-          node-version: "22"
-          cache: "pnpm"
-          cache-dependency-path: autogpt_platform/frontend/pnpm-lock.yaml
-
-      - name: Install JavaScript dependencies
-        working-directory: autogpt_platform/frontend
-        run: pnpm install --frozen-lockfile
-
      - name: Get CI failure details
        id: failure_details
        uses: actions/github-script@v8
--- a/.github/workflows/claude-dependabot.yml
+++ b/.github/workflows/claude-dependabot.yml
@@ -30,7 +30,7 @@ jobs:
      actions: read # Required for CI access
    steps:
      - name: Checkout code
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
        with:
          fetch-depth: 1

@@ -77,15 +77,27 @@ jobs:
        run: poetry run prisma generate && poetry run gen-prisma-stub

      # Frontend Node.js/pnpm setup (mirrors platform-frontend-ci.yml)
-      - name: Enable corepack
-        run: corepack enable
-
      - name: Set up Node.js
        uses: actions/setup-node@v6
        with:
          node-version: "22"
-          cache: "pnpm"
-          cache-dependency-path: autogpt_platform/frontend/pnpm-lock.yaml
+
+      - name: Enable corepack
+        run: corepack enable
+
+      - name: Set pnpm store directory
+        run: |
+          pnpm config set store-dir ~/.pnpm-store
+          echo "PNPM_HOME=$HOME/.pnpm-store" >> $GITHUB_ENV
+
+      - name: Cache frontend dependencies
+        uses: actions/cache@v5
+        with:
+          path: ~/.pnpm-store
+          key: ${{ runner.os }}-pnpm-${{ hashFiles('autogpt_platform/frontend/pnpm-lock.yaml', 'autogpt_platform/frontend/package.json') }}
+          restore-keys: |
+            ${{ runner.os }}-pnpm-${{ hashFiles('autogpt_platform/frontend/pnpm-lock.yaml') }}
+            ${{ runner.os }}-pnpm-

      - name: Install JavaScript dependencies
        working-directory: autogpt_platform/frontend
--- a/.github/workflows/claude.yml
+++ b/.github/workflows/claude.yml
@@ -40,7 +40,7 @@ jobs:
      actions: read # Required for CI access
    steps:
      - name: Checkout code
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
        with:
          fetch-depth: 1

@@ -93,15 +93,27 @@ jobs:
        run: poetry run prisma generate && poetry run gen-prisma-stub

      # Frontend Node.js/pnpm setup (mirrors platform-frontend-ci.yml)
-      - name: Enable corepack
-        run: corepack enable
-
      - name: Set up Node.js
        uses: actions/setup-node@v6
        with:
          node-version: "22"
-          cache: "pnpm"
-          cache-dependency-path: autogpt_platform/frontend/pnpm-lock.yaml
+
+      - name: Enable corepack
+        run: corepack enable
+
+      - name: Set pnpm store directory
+        run: |
+          pnpm config set store-dir ~/.pnpm-store
+          echo "PNPM_HOME=$HOME/.pnpm-store" >> $GITHUB_ENV
+
+      - name: Cache frontend dependencies
+        uses: actions/cache@v5
+        with:
+          path: ~/.pnpm-store
+          key: ${{ runner.os }}-pnpm-${{ hashFiles('autogpt_platform/frontend/pnpm-lock.yaml', 'autogpt_platform/frontend/package.json') }}
+          restore-keys: |
+            ${{ runner.os }}-pnpm-${{ hashFiles('autogpt_platform/frontend/pnpm-lock.yaml') }}
+            ${{ runner.os }}-pnpm-

      - name: Install JavaScript dependencies
        working-directory: autogpt_platform/frontend
--- a/.github/workflows/codeql.yml
+++ b/.github/workflows/codeql.yml
@@ -58,11 +58,11 @@ jobs:
        # your codebase is analyzed, see https://docs.github.com/en/code-security/code-scanning/creating-an-advanced-setup-for-code-scanning/codeql-code-scanning-for-compiled-languages
    steps:
    - name: Checkout repository
-      uses: actions/checkout@v6
+      uses: actions/checkout@v4

    # Initializes the CodeQL tools for scanning.
    - name: Initialize CodeQL
-      uses: github/codeql-action/init@v4
+      uses: github/codeql-action/init@v3
      with:
        languages: ${{ matrix.language }}
        build-mode: ${{ matrix.build-mode }}
@@ -93,6 +93,6 @@ jobs:
        exit 1

    - name: Perform CodeQL Analysis
-      uses: github/codeql-action/analyze@v4
+      uses: github/codeql-action/analyze@v3
      with:
        category: "/language:${{matrix.language}}"
--- a/.github/workflows/copilot-setup-steps.yml
+++ b/.github/workflows/copilot-setup-steps.yml
@@ -27,7 +27,7 @@ jobs:
    # If you do not check out your code, Copilot will do this for you.
    steps:
      - name: Checkout code
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
        with:
          fetch-depth: 0
          submodules: true
--- a/.github/workflows/docs-block-sync.yml
+++ b/.github/workflows/docs-block-sync.yml
@@ -23,7 +23,7 @@ jobs:

    steps:
      - name: Checkout code
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
        with:
          fetch-depth: 1

--- a/.github/workflows/docs-claude-review.yml
+++ b/.github/workflows/docs-claude-review.yml
@@ -7,10 +7,6 @@ on:
      - "docs/integrations/**"
      - "autogpt_platform/backend/backend/blocks/**"

-concurrency:
-  group: claude-docs-review-${{ github.event.pull_request.number }}
-  cancel-in-progress: true
-
 jobs:
  claude-review:
    # Only run for PRs from members/collaborators
@@ -27,7 +23,7 @@ jobs:

    steps:
      - name: Checkout code
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
        with:
          fetch-depth: 0

@@ -95,35 +91,5 @@ jobs:
            3. Read corresponding documentation files to verify accuracy
            4. Provide your feedback as a PR comment

-            ## IMPORTANT: Comment Marker
-            Start your PR comment with exactly this HTML comment marker on its own line:
-            <!-- CLAUDE_DOCS_REVIEW -->
-
-            This marker is used to identify and replace your comment on subsequent runs.
-
            Be constructive and specific. If everything looks good, say so!
            If there are issues, explain what's wrong and suggest how to fix it.
-
-      - name: Delete old Claude review comments
-        env:
-          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-        run: |
-          # Get all comment IDs with our marker, sorted by creation date (oldest first)
-          COMMENT_IDS=$(gh api \
-            repos/${{ github.repository }}/issues/${{ github.event.pull_request.number }}/comments \
-            --jq '[.[] | select(.body | contains("<!-- CLAUDE_DOCS_REVIEW -->"))] | sort_by(.created_at) | .[].id')
-
-          # Count comments
-          COMMENT_COUNT=$(echo "$COMMENT_IDS" | grep -c . || true)
-
-          if [ "$COMMENT_COUNT" -gt 1 ]; then
-            # Delete all but the last (newest) comment
-            echo "$COMMENT_IDS" | head -n -1 | while read -r COMMENT_ID; do
-              if [ -n "$COMMENT_ID" ]; then
-                echo "Deleting old review comment: $COMMENT_ID"
-                gh api -X DELETE repos/${{ github.repository }}/issues/comments/$COMMENT_ID
-              fi
-            done
-          else
-            echo "No old review comments to clean up"
-          fi
--- a/.github/workflows/docs-enhance.yml
+++ b/.github/workflows/docs-enhance.yml
@@ -28,7 +28,7 @@ jobs:

    steps:
      - name: Checkout code
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
        with:
          fetch-depth: 1

--- a/.github/workflows/platform-autogpt-deploy-dev.yaml
+++ b/.github/workflows/platform-autogpt-deploy-dev.yaml
@@ -25,7 +25,7 @@ jobs:

    steps:
      - name: Checkout code
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
        with:
          ref: ${{ github.event.inputs.git_ref || github.ref_name }}

@@ -52,7 +52,7 @@ jobs:
    runs-on: ubuntu-latest
    steps:
      - name: Trigger deploy workflow
-        uses: peter-evans/repository-dispatch@v4
+        uses: peter-evans/repository-dispatch@v3
        with:
          token: ${{ secrets.DEPLOY_TOKEN }}
          repository: Significant-Gravitas/AutoGPT_cloud_infrastructure
--- a/.github/workflows/platform-autogpt-deploy-prod.yml
+++ b/.github/workflows/platform-autogpt-deploy-prod.yml
@@ -17,7 +17,7 @@ jobs:

    steps:
      - name: Checkout code
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
        with:
          ref: ${{ github.ref_name || 'master' }}

@@ -45,7 +45,7 @@ jobs:
    runs-on: ubuntu-latest
    steps:
      - name: Trigger deploy workflow
-        uses: peter-evans/repository-dispatch@v4
+        uses: peter-evans/repository-dispatch@v3
        with:
          token: ${{ secrets.DEPLOY_TOKEN }}
          repository: Significant-Gravitas/AutoGPT_cloud_infrastructure
--- a/.github/workflows/platform-backend-ci.yml
+++ b/.github/workflows/platform-backend-ci.yml
@@ -41,18 +41,13 @@ jobs:
        ports:
          - 6379:6379
      rabbitmq:
-        image: rabbitmq:4.1.4
+        image: rabbitmq:3.12-management
        ports:
          - 5672:5672
+          - 15672:15672
        env:
          RABBITMQ_DEFAULT_USER: ${{ env.RABBITMQ_DEFAULT_USER }}
          RABBITMQ_DEFAULT_PASS: ${{ env.RABBITMQ_DEFAULT_PASS }}
-        options: >-
-          --health-cmd "rabbitmq-diagnostics -q ping"
-          --health-interval 30s
-          --health-timeout 10s
-          --health-retries 5
-          --health-start-period 10s
      clamav:
        image: clamav/clamav-debian:latest
        ports:
@@ -73,7 +68,7 @@ jobs:

    steps:
      - name: Checkout repository
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
        with:
          fetch-depth: 0
          submodules: true
--- a/.github/workflows/platform-dev-deploy-event-dispatcher.yml
+++ b/.github/workflows/platform-dev-deploy-event-dispatcher.yml
@@ -82,7 +82,7 @@ jobs:
          
      - name: Dispatch Deploy Event
        if: steps.check_status.outputs.should_deploy == 'true'
-        uses: peter-evans/repository-dispatch@v4
+        uses: peter-evans/repository-dispatch@v3
        with:
          token: ${{ secrets.DISPATCH_TOKEN }}
          repository: Significant-Gravitas/AutoGPT_cloud_infrastructure
@@ -110,7 +110,7 @@ jobs:

      - name: Dispatch Undeploy Event (from comment)
        if: steps.check_status.outputs.should_undeploy == 'true'
-        uses: peter-evans/repository-dispatch@v4
+        uses: peter-evans/repository-dispatch@v3
        with:
          token: ${{ secrets.DISPATCH_TOKEN }}
          repository: Significant-Gravitas/AutoGPT_cloud_infrastructure
@@ -168,7 +168,7 @@ jobs:
          github.event_name == 'pull_request' &&
          github.event.action == 'closed' &&
          steps.check_pr_close.outputs.should_undeploy == 'true'
-        uses: peter-evans/repository-dispatch@v4
+        uses: peter-evans/repository-dispatch@v3
        with:
          token: ${{ secrets.DISPATCH_TOKEN }}
          repository: Significant-Gravitas/AutoGPT_cloud_infrastructure
--- a/.github/workflows/platform-frontend-ci.yml
+++ b/.github/workflows/platform-frontend-ci.yml
@@ -6,16 +6,10 @@ on:
    paths:
      - ".github/workflows/platform-frontend-ci.yml"
      - "autogpt_platform/frontend/**"
-      - "autogpt_platform/backend/Dockerfile"
-      - "autogpt_platform/docker-compose.yml"
-      - "autogpt_platform/docker-compose.platform.yml"
  pull_request:
    paths:
      - ".github/workflows/platform-frontend-ci.yml"
      - "autogpt_platform/frontend/**"
-      - "autogpt_platform/backend/Dockerfile"
-      - "autogpt_platform/docker-compose.yml"
-      - "autogpt_platform/docker-compose.platform.yml"
  merge_group:
  workflow_dispatch:

@@ -32,11 +26,12 @@ jobs:
  setup:
    runs-on: ubuntu-latest
    outputs:
+      cache-key: ${{ steps.cache-key.outputs.key }}
      components-changed: ${{ steps.filter.outputs.components }}

    steps:
      - name: Checkout repository
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4

      - name: Check for component changes
        uses: dorny/paths-filter@v3
@@ -46,17 +41,28 @@ jobs:
            components:
              - 'autogpt_platform/frontend/src/components/**'

-      - name: Enable corepack
-        run: corepack enable
-
-      - name: Set up Node
+      - name: Set up Node.js
        uses: actions/setup-node@v6
        with:
          node-version: "22.18.0"
-          cache: "pnpm"
-          cache-dependency-path: autogpt_platform/frontend/pnpm-lock.yaml

-      - name: Install dependencies to populate cache
+      - name: Enable corepack
+        run: corepack enable
+
+      - name: Generate cache key
+        id: cache-key
+        run: echo "key=${{ runner.os }}-pnpm-${{ hashFiles('autogpt_platform/frontend/pnpm-lock.yaml', 'autogpt_platform/frontend/package.json') }}" >> $GITHUB_OUTPUT
+
+      - name: Cache dependencies
+        uses: actions/cache@v5
+        with:
+          path: ~/.pnpm-store
+          key: ${{ steps.cache-key.outputs.key }}
+          restore-keys: |
+            ${{ runner.os }}-pnpm-${{ hashFiles('autogpt_platform/frontend/pnpm-lock.yaml') }}
+            ${{ runner.os }}-pnpm-
+
+      - name: Install dependencies
        run: pnpm install --frozen-lockfile

  lint:
@@ -65,17 +71,24 @@ jobs:

    steps:
      - name: Checkout repository
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
+
+      - name: Set up Node.js
+        uses: actions/setup-node@v6
+        with:
+          node-version: "22.18.0"

      - name: Enable corepack
        run: corepack enable

-      - name: Set up Node
-        uses: actions/setup-node@v6
+      - name: Restore dependencies cache
+        uses: actions/cache@v5
        with:
-          node-version: "22.18.0"
-          cache: "pnpm"
-          cache-dependency-path: autogpt_platform/frontend/pnpm-lock.yaml
+          path: ~/.pnpm-store
+          key: ${{ needs.setup.outputs.cache-key }}
+          restore-keys: |
+            ${{ runner.os }}-pnpm-${{ hashFiles('autogpt_platform/frontend/pnpm-lock.yaml') }}
+            ${{ runner.os }}-pnpm-

      - name: Install dependencies
        run: pnpm install --frozen-lockfile
@@ -94,19 +107,26 @@ jobs:

    steps:
      - name: Checkout repository
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
        with:
          fetch-depth: 0

+      - name: Set up Node.js
+        uses: actions/setup-node@v6
+        with:
+          node-version: "22.18.0"
+
      - name: Enable corepack
        run: corepack enable

-      - name: Set up Node
-        uses: actions/setup-node@v6
+      - name: Restore dependencies cache
+        uses: actions/cache@v5
        with:
-          node-version: "22.18.0"
-          cache: "pnpm"
-          cache-dependency-path: autogpt_platform/frontend/pnpm-lock.yaml
+          path: ~/.pnpm-store
+          key: ${{ needs.setup.outputs.cache-key }}
+          restore-keys: |
+            ${{ runner.os }}-pnpm-${{ hashFiles('autogpt_platform/frontend/pnpm-lock.yaml') }}
+            ${{ runner.os }}-pnpm-

      - name: Install dependencies
        run: pnpm install --frozen-lockfile
@@ -121,20 +141,30 @@ jobs:
          exitOnceUploaded: true

  e2e_test:
-    name: end-to-end tests
    runs-on: big-boi
+    needs: setup
+    strategy:
+      fail-fast: false

    steps:
      - name: Checkout repository
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
        with:
          submodules: recursive

-      - name: Set up Platform - Copy default supabase .env
+      - name: Set up Node.js
+        uses: actions/setup-node@v6
+        with:
+          node-version: "22.18.0"
+
+      - name: Enable corepack
+        run: corepack enable
+
+      - name: Copy default supabase .env
        run: |
          cp ../.env.default ../.env

-      - name: Set up Platform - Copy backend .env and set OpenAI API key
+      - name: Copy backend .env and set OpenAI API key
        run: |
          cp ../backend/.env.default ../backend/.env
          echo "OPENAI_INTERNAL_API_KEY=${{ secrets.OPENAI_API_KEY }}" >> ../backend/.env
@@ -142,125 +172,77 @@ jobs:
          # Used by E2E test data script to generate embeddings for approved store agents
          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}

-      - name: Set up Platform - Set up Docker Buildx
+      - name: Set up Docker Buildx
        uses: docker/setup-buildx-action@v3
-        with:
-          driver: docker-container
-          driver-opts: network=host

-      - name: Set up Platform - Expose GHA cache to docker buildx CLI
-        uses: crazy-max/ghaction-github-runtime@v3
-
-      - name: Set up Platform - Build Docker images (with cache)
-        working-directory: autogpt_platform
-        run: |
-          pip install pyyaml
-
-          # Resolve extends and generate a flat compose file that bake can understand
-          docker compose -f docker-compose.yml config > docker-compose.resolved.yml
-
-          # Add cache configuration to the resolved compose file
-          python ../.github/workflows/scripts/docker-ci-fix-compose-build-cache.py \
-            --source docker-compose.resolved.yml \
-            --cache-from "type=gha" \
-            --cache-to "type=gha,mode=max" \
-            --backend-hash "${{ hashFiles('autogpt_platform/backend/Dockerfile', 'autogpt_platform/backend/poetry.lock', 'autogpt_platform/backend/backend') }}" \
-            --frontend-hash "${{ hashFiles('autogpt_platform/frontend/Dockerfile', 'autogpt_platform/frontend/pnpm-lock.yaml', 'autogpt_platform/frontend/src') }}" \
-            --git-ref "${{ github.ref }}"
-
-          # Build with bake using the resolved compose file (now includes cache config)
-          docker buildx bake --allow=fs.read=.. -f docker-compose.resolved.yml --load
-        env:
-          NEXT_PUBLIC_PW_TEST: true
-
-      - name: Set up tests - Cache E2E test data
-        id: e2e-data-cache
+      - name: Cache Docker layers
        uses: actions/cache@v5
        with:
-          path: /tmp/e2e_test_data.sql
-          key: e2e-test-data-${{ hashFiles('autogpt_platform/backend/test/e2e_test_data.py', 'autogpt_platform/backend/migrations/**', '.github/workflows/platform-frontend-ci.yml') }}
+          path: /tmp/.buildx-cache
+          key: ${{ runner.os }}-buildx-frontend-test-${{ hashFiles('autogpt_platform/docker-compose.yml', 'autogpt_platform/backend/Dockerfile', 'autogpt_platform/backend/pyproject.toml', 'autogpt_platform/backend/poetry.lock') }}
+          restore-keys: |
+            ${{ runner.os }}-buildx-frontend-test-

-      - name: Set up Platform - Start Supabase DB + Auth
+      - name: Run docker compose
        run: |
-          docker compose -f ../docker-compose.resolved.yml up -d db auth --no-build
-          echo "Waiting for database to be ready..."
-          timeout 60 sh -c 'until docker compose -f ../docker-compose.resolved.yml exec -T db pg_isready -U postgres 2>/dev/null; do sleep 2; done'
-          echo "Waiting for auth service to be ready..."
-          timeout 60 sh -c 'until docker compose -f ../docker-compose.resolved.yml exec -T db psql -U postgres -d postgres -c "SELECT 1 FROM auth.users LIMIT 1" 2>/dev/null; do sleep 2; done' || echo "Auth schema check timeout, continuing..."
-
-      - name: Set up Platform - Run migrations
-        run: |
-          echo "Running migrations..."
-          docker compose -f ../docker-compose.resolved.yml run --rm migrate
-          echo "✅ Migrations completed"
+          NEXT_PUBLIC_PW_TEST=true docker compose -f ../docker-compose.yml up -d
        env:
-          NEXT_PUBLIC_PW_TEST: true
+          DOCKER_BUILDKIT: 1
+          BUILDX_CACHE_FROM: type=local,src=/tmp/.buildx-cache
+          BUILDX_CACHE_TO: type=local,dest=/tmp/.buildx-cache-new,mode=max

-      - name: Set up tests - Load cached E2E test data
-        if: steps.e2e-data-cache.outputs.cache-hit == 'true'
+      - name: Move cache
        run: |
-          echo "✅ Found cached E2E test data, restoring..."
-          {
-            echo "SET session_replication_role = 'replica';"
-            cat /tmp/e2e_test_data.sql
-            echo "SET session_replication_role = 'origin';"
-          } | docker compose -f ../docker-compose.resolved.yml exec -T db psql -U postgres -d postgres -b
-          # Refresh materialized views after restore
-          docker compose -f ../docker-compose.resolved.yml exec -T db \
-            psql -U postgres -d postgres -b -c "SET search_path TO platform; SELECT refresh_store_materialized_views();" || true
+          rm -rf /tmp/.buildx-cache
+          if [ -d "/tmp/.buildx-cache-new" ]; then
+            mv /tmp/.buildx-cache-new /tmp/.buildx-cache
+          fi

-          echo "✅ E2E test data restored from cache"
-
-      - name: Set up Platform - Start (all other services)
+      - name: Wait for services to be ready
        run: |
-          docker compose -f ../docker-compose.resolved.yml up -d --no-build
          echo "Waiting for rest_server to be ready..."
          timeout 60 sh -c 'until curl -f http://localhost:8006/health 2>/dev/null; do sleep 2; done' || echo "Rest server health check timeout, continuing..."
-        env:
-          NEXT_PUBLIC_PW_TEST: true
+          echo "Waiting for database to be ready..."
+          timeout 60 sh -c 'until docker compose -f ../docker-compose.yml exec -T db pg_isready -U postgres 2>/dev/null; do sleep 2; done' || echo "Database ready check timeout, continuing..."

-      - name: Set up tests - Create E2E test data
-        if: steps.e2e-data-cache.outputs.cache-hit != 'true'
+      - name: Create E2E test data
        run: |
          echo "Creating E2E test data..."
-          docker cp ../backend/test/e2e_test_data.py $(docker compose -f ../docker-compose.resolved.yml ps -q rest_server):/tmp/e2e_test_data.py
-          docker compose -f ../docker-compose.resolved.yml exec -T rest_server sh -c "cd /app/autogpt_platform && python /tmp/e2e_test_data.py" || {
-            echo "❌ E2E test data creation failed!"
-            docker compose -f ../docker-compose.resolved.yml logs --tail=50 rest_server
-            exit 1
-          }
+          # First try to run the script from inside the container
+          if docker compose -f ../docker-compose.yml exec -T rest_server test -f /app/autogpt_platform/backend/test/e2e_test_data.py; then
+            echo "✅ Found e2e_test_data.py in container, running it..."
+            docker compose -f ../docker-compose.yml exec -T rest_server sh -c "cd /app/autogpt_platform && python backend/test/e2e_test_data.py" || {
+              echo "❌ E2E test data creation failed!"
+              docker compose -f ../docker-compose.yml logs --tail=50 rest_server
+              exit 1
+            }
+          else
+            echo "⚠️ e2e_test_data.py not found in container, copying and running..."
+            # Copy the script into the container and run it
+            docker cp ../backend/test/e2e_test_data.py $(docker compose -f ../docker-compose.yml ps -q rest_server):/tmp/e2e_test_data.py || {
+              echo "❌ Failed to copy script to container"
+              exit 1
+            }
+            docker compose -f ../docker-compose.yml exec -T rest_server sh -c "cd /app/autogpt_platform && python /tmp/e2e_test_data.py" || {
+              echo "❌ E2E test data creation failed!"
+              docker compose -f ../docker-compose.yml logs --tail=50 rest_server
+              exit 1
+            }
+          fi

-          # Dump auth.users + platform schema for cache (two separate dumps)
-          echo "Dumping database for cache..."
-          {
-            docker compose -f ../docker-compose.resolved.yml exec -T db \
-              pg_dump -U postgres --data-only --column-inserts \
-              --table='auth.users' postgres
-            docker compose -f ../docker-compose.resolved.yml exec -T db \
-              pg_dump -U postgres --data-only --column-inserts \
-              --schema=platform \
-              --exclude-table='platform._prisma_migrations' \
-              --exclude-table='platform.apscheduler_jobs' \
-              --exclude-table='platform.apscheduler_jobs_batched_notifications' \
-              postgres
-          } > /tmp/e2e_test_data.sql
-
-          echo "✅ Database dump created for caching ($(wc -l < /tmp/e2e_test_data.sql) lines)"
-
-      - name: Set up tests - Enable corepack
-        run: corepack enable
-
-      - name: Set up tests - Set up Node
-        uses: actions/setup-node@v6
+      - name: Restore dependencies cache
+        uses: actions/cache@v5
        with:
-          node-version: "22.18.0"
-          cache: "pnpm"
-          cache-dependency-path: autogpt_platform/frontend/pnpm-lock.yaml
+          path: ~/.pnpm-store
+          key: ${{ needs.setup.outputs.cache-key }}
+          restore-keys: |
+            ${{ runner.os }}-pnpm-${{ hashFiles('autogpt_platform/frontend/pnpm-lock.yaml') }}
+            ${{ runner.os }}-pnpm-

-      - name: Set up tests - Install dependencies
+      - name: Install dependencies
        run: pnpm install --frozen-lockfile

-      - name: Set up tests - Install browser 'chromium'
+      - name: Install Browser 'chromium'
        run: pnpm playwright install --with-deps chromium

      - name: Run Playwright tests
@@ -287,7 +269,7 @@ jobs:

      - name: Print Final Docker Compose logs
        if: always()
-        run: docker compose -f ../docker-compose.resolved.yml logs
+        run: docker compose -f ../docker-compose.yml logs

  integration_test:
    runs-on: ubuntu-latest
@@ -295,19 +277,26 @@ jobs:

    steps:
      - name: Checkout repository
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
        with:
          submodules: recursive

+      - name: Set up Node.js
+        uses: actions/setup-node@v6
+        with:
+          node-version: "22.18.0"
+
      - name: Enable corepack
        run: corepack enable

-      - name: Set up Node
-        uses: actions/setup-node@v6
+      - name: Restore dependencies cache
+        uses: actions/cache@v5
        with:
-          node-version: "22.18.0"
-          cache: "pnpm"
-          cache-dependency-path: autogpt_platform/frontend/pnpm-lock.yaml
+          path: ~/.pnpm-store
+          key: ${{ needs.setup.outputs.cache-key }}
+          restore-keys: |
+            ${{ runner.os }}-pnpm-${{ hashFiles('autogpt_platform/frontend/pnpm-lock.yaml') }}
+            ${{ runner.os }}-pnpm-

      - name: Install dependencies
        run: pnpm install --frozen-lockfile
--- a/.github/workflows/platform-fullstack-ci.yml
+++ b/.github/workflows/platform-fullstack-ci.yml
@@ -29,7 +29,7 @@ jobs:

    steps:
      - name: Checkout repository
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4

      - name: Set up Node.js
        uses: actions/setup-node@v6
@@ -56,14 +56,14 @@ jobs:
        run: pnpm install --frozen-lockfile

  types:
-    runs-on: big-boi
+    runs-on: ubuntu-latest
    needs: setup
    strategy:
      fail-fast: false

    steps:
      - name: Checkout repository
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
        with:
          submodules: recursive

@@ -85,7 +85,7 @@ jobs:

      - name: Run docker compose
        run: |
-          docker compose -f ../docker-compose.yml --profile local up -d deps_backend
+          docker compose -f ../docker-compose.yml --profile local --profile deps_backend up -d

      - name: Restore dependencies cache
        uses: actions/cache@v5
--- a/.github/workflows/pr-overlap-check.yml
+++ b/.github/workflows/pr-overlap-check.yml
@@ -1,39 +0,0 @@
-name: PR Overlap Detection
-
-on:
-  pull_request:
-    types: [opened, synchronize, reopened]
-    branches:
-      - dev
-      - master
-
-permissions:
-  contents: read
-  pull-requests: write
-
-jobs:
-  check-overlaps:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v4
-        with:
-          fetch-depth: 0  # Need full history for merge testing
-
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: '3.11'
-
-      - name: Configure git
-        run: |
-          git config user.email "github-actions[bot]@users.noreply.github.com"
-          git config user.name "github-actions[bot]"
-
-      - name: Run overlap detection
-        env:
-          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-        # Always succeed - this check informs contributors, it shouldn't block merging
-        continue-on-error: true
-        run: |
-          python .github/scripts/detect_overlaps.py ${{ github.event.pull_request.number }}
--- a/.github/workflows/repo-workflow-checker.yml
+++ b/.github/workflows/repo-workflow-checker.yml
@@ -11,7 +11,7 @@ jobs:
    steps:
      # - name: Wait some time for all actions to start
      #   run: sleep 30
-      - uses: actions/checkout@v6
+      - uses: actions/checkout@v4
        # with:
          # fetch-depth: 0
      - name: Set up Python
--- a/.github/workflows/scripts/docker-ci-fix-compose-build-cache.py
+++ b/.github/workflows/scripts/docker-ci-fix-compose-build-cache.py
@@ -1,195 +0,0 @@
-#!/usr/bin/env python3
-"""
-Add cache configuration to a resolved docker-compose file for all services
-that have a build key, and ensure image names match what docker compose expects.
-"""
-
-import argparse
-
-import yaml
-
-
-DEFAULT_BRANCH = "dev"
-CACHE_BUILDS_FOR_COMPONENTS = ["backend", "frontend"]
-
-
-def main():
-    parser = argparse.ArgumentParser(
-        description="Add cache config to a resolved compose file"
-    )
-    parser.add_argument(
-        "--source",
-        required=True,
-        help="Source compose file to read (should be output of `docker compose config`)",
-    )
-    parser.add_argument(
-        "--cache-from",
-        default="type=gha",
-        help="Cache source configuration",
-    )
-    parser.add_argument(
-        "--cache-to",
-        default="type=gha,mode=max",
-        help="Cache destination configuration",
-    )
-    for component in CACHE_BUILDS_FOR_COMPONENTS:
-        parser.add_argument(
-            f"--{component}-hash",
-            default="",
-            help=f"Hash for {component} cache scope (e.g., from hashFiles())",
-        )
-    parser.add_argument(
-        "--git-ref",
-        default="",
-        help="Git ref for branch-based cache scope (e.g., refs/heads/master)",
-    )
-    args = parser.parse_args()
-
-    # Normalize git ref to a safe scope name (e.g., refs/heads/master -> master)
-    git_ref_scope = ""
-    if args.git_ref:
-        git_ref_scope = args.git_ref.replace("refs/heads/", "").replace("/", "-")
-
-    with open(args.source, "r") as f:
-        compose = yaml.safe_load(f)
-
-    # Get project name from compose file or default
-    project_name = compose.get("name", "autogpt_platform")
-
-    def get_image_name(dockerfile: str, target: str) -> str:
-        """Generate image name based on Dockerfile folder and build target."""
-        dockerfile_parts = dockerfile.replace("\\", "/").split("/")
-        if len(dockerfile_parts) >= 2:
-            folder_name = dockerfile_parts[-2]  # e.g., "backend" or "frontend"
-        else:
-            folder_name = "app"
-        return f"{project_name}-{folder_name}:{target}"
-
-    def get_build_key(dockerfile: str, target: str) -> str:
-        """Generate a unique key for a Dockerfile+target combination."""
-        return f"{dockerfile}:{target}"
-
-    def get_component(dockerfile: str) -> str | None:
-        """Get component name (frontend/backend) from dockerfile path."""
-        for component in CACHE_BUILDS_FOR_COMPONENTS:
-            if component in dockerfile:
-                return component
-        return None
-
-    # First pass: collect all services with build configs and identify duplicates
-    # Track which (dockerfile, target) combinations we've seen
-    build_key_to_first_service: dict[str, str] = {}
-    services_to_build: list[str] = []
-    services_to_dedupe: list[str] = []
-
-    for service_name, service_config in compose.get("services", {}).items():
-        if "build" not in service_config:
-            continue
-
-        build_config = service_config["build"]
-        dockerfile = build_config.get("dockerfile", "Dockerfile")
-        target = build_config.get("target", "default")
-        build_key = get_build_key(dockerfile, target)
-
-        if build_key not in build_key_to_first_service:
-            # First service with this build config - it will do the actual build
-            build_key_to_first_service[build_key] = service_name
-            services_to_build.append(service_name)
-        else:
-            # Duplicate - will just use the image from the first service
-            services_to_dedupe.append(service_name)
-
-    # Second pass: configure builds and deduplicate
-    modified_services = []
-    for service_name, service_config in compose.get("services", {}).items():
-        if "build" not in service_config:
-            continue
-
-        build_config = service_config["build"]
-        dockerfile = build_config.get("dockerfile", "Dockerfile")
-        target = build_config.get("target", "latest")
-        image_name = get_image_name(dockerfile, target)
-
-        # Set image name for all services (needed for both builders and deduped)
-        service_config["image"] = image_name
-
-        if service_name in services_to_dedupe:
-            # Remove build config - this service will use the pre-built image
-            del service_config["build"]
-            continue
-
-        # This service will do the actual build - add cache config
-        cache_from_list = []
-        cache_to_list = []
-
-        component = get_component(dockerfile)
-        if not component:
-            # Skip services that don't clearly match frontend/backend
-            continue
-
-        # Get the hash for this component
-        component_hash = getattr(args, f"{component}_hash")
-
-        # Scope format: platform-{component}-{target}-{hash|ref}
-        # Example: platform-backend-server-abc123
-
-        if "type=gha" in args.cache_from:
-            # 1. Primary: exact hash match (most specific)
-            if component_hash:
-                hash_scope = f"platform-{component}-{target}-{component_hash}"
-                cache_from_list.append(f"{args.cache_from},scope={hash_scope}")
-
-            # 2. Fallback: branch-based cache
-            if git_ref_scope:
-                ref_scope = f"platform-{component}-{target}-{git_ref_scope}"
-                cache_from_list.append(f"{args.cache_from},scope={ref_scope}")
-
-            # 3. Fallback: dev branch cache (for PRs/feature branches)
-            if git_ref_scope and git_ref_scope != DEFAULT_BRANCH:
-                master_scope = f"platform-{component}-{target}-{DEFAULT_BRANCH}"
-                cache_from_list.append(f"{args.cache_from},scope={master_scope}")
-
-        if "type=gha" in args.cache_to:
-            # Write to both hash-based and branch-based scopes
-            if component_hash:
-                hash_scope = f"platform-{component}-{target}-{component_hash}"
-                cache_to_list.append(f"{args.cache_to},scope={hash_scope}")
-
-            if git_ref_scope:
-                ref_scope = f"platform-{component}-{target}-{git_ref_scope}"
-                cache_to_list.append(f"{args.cache_to},scope={ref_scope}")
-
-        # Ensure we have at least one cache source/target
-        if not cache_from_list:
-            cache_from_list.append(args.cache_from)
-        if not cache_to_list:
-            cache_to_list.append(args.cache_to)
-
-        build_config["cache_from"] = cache_from_list
-        build_config["cache_to"] = cache_to_list
-        modified_services.append(service_name)
-
-    # Write back to the same file
-    with open(args.source, "w") as f:
-        yaml.dump(compose, f, default_flow_style=False, sort_keys=False)
-
-    print(f"Added cache config to {len(modified_services)} services in {args.source}:")
-    for svc in modified_services:
-        svc_config = compose["services"][svc]
-        build_cfg = svc_config.get("build", {})
-        cache_from_list = build_cfg.get("cache_from", ["none"])
-        cache_to_list = build_cfg.get("cache_to", ["none"])
-        print(f"  - {svc}")
-        print(f"      image: {svc_config.get('image', 'N/A')}")
-        print(f"      cache_from: {cache_from_list}")
-        print(f"      cache_to: {cache_to_list}")
-    if services_to_dedupe:
-        print(
-            f"Deduplicated {len(services_to_dedupe)} services (will use pre-built images):"
-        )
-        for svc in services_to_dedupe:
-            print(f"  - {svc} -> {compose['services'][svc].get('image', 'N/A')}")
-
-
-if __name__ == "__main__":
-    main()
--- a/autogpt_platform/CLAUDE.md
+++ b/autogpt_platform/CLAUDE.md
@@ -45,11 +45,6 @@ AutoGPT Platform is a monorepo containing:
 - Backend/Frontend services use YAML anchors for consistent configuration
 - Supabase services (`db/docker/docker-compose.yml`) follow the same pattern

-### Branching Strategy
-
- **`dev`** is the main development branch. All PRs should target `dev`.
- **`master`** is the production branch. Only used for production releases.
-
 ### Creating Pull Requests

 - Create the PR against the `dev` branch of the repository.
--- a/autogpt_platform/autogpt_libs/poetry.lock
+++ b/autogpt_platform/autogpt_libs/poetry.lock
@@ -1,4 +1,4 @@
-# This file is automatically @generated by Poetry 2.1.1 and should not be changed by hand.
+# This file is automatically @generated by Poetry 2.2.1 and should not be changed by hand.

 [[package]]
 name = "annotated-doc"
@@ -67,7 +67,7 @@ description = "Backport of asyncio.Runner, a context manager that controls event
 optional = false
 python-versions = "<3.11,>=3.8"
 groups = ["dev"]
-markers = "python_version < \"3.11\""
+markers = "python_version == \"3.10\""
 files = [
    {file = "backports_asyncio_runner-1.2.0-py3-none-any.whl", hash = "sha256:0da0a936a8aeb554eccb426dc55af3ba63bcdc69fa1a600b5bb305413a4477b5"},
    {file = "backports_asyncio_runner-1.2.0.tar.gz", hash = "sha256:a5aa7b2b7d8f8bfcaa2b57313f70792df84e32a2a746f585213373f900b42162"},
@@ -326,118 +326,100 @@ files = [

 [[package]]
 name = "coverage"
-version = "7.13.4"
+version = "7.10.5"
 description = "Code coverage measurement for Python"
 optional = false
-python-versions = ">=3.10"
+python-versions = ">=3.9"
 groups = ["dev"]
 files = [
-    {file = "coverage-7.13.4-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:0fc31c787a84f8cd6027eba44010517020e0d18487064cd3d8968941856d1415"},
-    {file = "coverage-7.13.4-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:a32ebc02a1805adf637fc8dec324b5cdacd2e493515424f70ee33799573d661b"},
-    {file = "coverage-7.13.4-cp310-cp310-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:e24f9156097ff9dc286f2f913df3a7f63c0e333dcafa3c196f2c18b4175ca09a"},
-    {file = "coverage-7.13.4-cp310-cp310-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:8041b6c5bfdc03257666e9881d33b1abc88daccaf73f7b6340fb7946655cd10f"},
-    {file = "coverage-7.13.4-cp310-cp310-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:2a09cfa6a5862bc2fc6ca7c3def5b2926194a56b8ab78ffcf617d28911123012"},
-    {file = "coverage-7.13.4-cp310-cp310-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:296f8b0af861d3970c2a4d8c91d48eb4dd4771bcef9baedec6a9b515d7de3def"},
-    {file = "coverage-7.13.4-cp310-cp310-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:e101609bcbbfb04605ea1027b10dc3735c094d12d40826a60f897b98b1c30256"},
-    {file = "coverage-7.13.4-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:aa3feb8db2e87ff5e6d00d7e1480ae241876286691265657b500886c98f38bda"},
-    {file = "coverage-7.13.4-cp310-cp310-musllinux_1_2_i686.whl", hash = "sha256:4fc7fa81bbaf5a02801b65346c8b3e657f1d93763e58c0abdf7c992addd81a92"},
-    {file = "coverage-7.13.4-cp310-cp310-musllinux_1_2_ppc64le.whl", hash = "sha256:33901f604424145c6e9c2398684b92e176c0b12df77d52db81c20abd48c3794c"},
-    {file = "coverage-7.13.4-cp310-cp310-musllinux_1_2_riscv64.whl", hash = "sha256:bb28c0f2cf2782508a40cec377935829d5fcc3ad9a3681375af4e84eb34b6b58"},
-    {file = "coverage-7.13.4-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:9d107aff57a83222ddbd8d9ee705ede2af2cc926608b57abed8ef96b50b7e8f9"},
-    {file = "coverage-7.13.4-cp310-cp310-win32.whl", hash = "sha256:a6f94a7d00eb18f1b6d403c91a88fd58cfc92d4b16080dfdb774afc8294469bf"},
-    {file = "coverage-7.13.4-cp310-cp310-win_amd64.whl", hash = "sha256:2cb0f1e000ebc419632bbe04366a8990b6e32c4e0b51543a6484ffe15eaeda95"},
-    {file = "coverage-7.13.4-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:d490ba50c3f35dd7c17953c68f3270e7ccd1c6642e2d2afe2d8e720b98f5a053"},
-    {file = "coverage-7.13.4-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:19bc3c88078789f8ef36acb014d7241961dbf883fd2533d18cb1e7a5b4e28b11"},
-    {file = "coverage-7.13.4-cp311-cp311-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:3998e5a32e62fdf410c0dbd3115df86297995d6e3429af80b8798aad894ca7aa"},
-    {file = "coverage-7.13.4-cp311-cp311-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:8e264226ec98e01a8e1054314af91ee6cde0eacac4f465cc93b03dbe0bce2fd7"},
-    {file = "coverage-7.13.4-cp311-cp311-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:a3aa4e7b9e416774b21797365b358a6e827ffadaaca81b69ee02946852449f00"},
-    {file = "coverage-7.13.4-cp311-cp311-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:71ca20079dd8f27fcf808817e281e90220475cd75115162218d0e27549f95fef"},
-    {file = "coverage-7.13.4-cp311-cp311-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:e2f25215f1a359ab17320b47bcdaca3e6e6356652e8256f2441e4ef972052903"},
-    {file = "coverage-7.13.4-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:d65b2d373032411e86960604dc4edac91fdfb5dca539461cf2cbe78327d1e64f"},
-    {file = "coverage-7.13.4-cp311-cp311-musllinux_1_2_i686.whl", hash = "sha256:94eb63f9b363180aff17de3e7c8760c3ba94664ea2695c52f10111244d16a299"},
-    {file = "coverage-7.13.4-cp311-cp311-musllinux_1_2_ppc64le.whl", hash = "sha256:e856bf6616714c3a9fbc270ab54103f4e685ba236fa98c054e8f87f266c93505"},
-    {file = "coverage-7.13.4-cp311-cp311-musllinux_1_2_riscv64.whl", hash = "sha256:65dfcbe305c3dfe658492df2d85259e0d79ead4177f9ae724b6fb245198f55d6"},
-    {file = "coverage-7.13.4-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:b507778ae8a4c915436ed5c2e05b4a6cecfa70f734e19c22a005152a11c7b6a9"},
-    {file = "coverage-7.13.4-cp311-cp311-win32.whl", hash = "sha256:784fc3cf8be001197b652d51d3fd259b1e2262888693a4636e18879f613a62a9"},
-    {file = "coverage-7.13.4-cp311-cp311-win_amd64.whl", hash = "sha256:2421d591f8ca05b308cf0092807308b2facbefe54af7c02ac22548b88b95c98f"},
-    {file = "coverage-7.13.4-cp311-cp311-win_arm64.whl", hash = "sha256:79e73a76b854d9c6088fe5d8b2ebe745f8681c55f7397c3c0a016192d681045f"},
-    {file = "coverage-7.13.4-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:02231499b08dabbe2b96612993e5fc34217cdae907a51b906ac7fca8027a4459"},
-    {file = "coverage-7.13.4-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:40aa8808140e55dc022b15d8aa7f651b6b3d68b365ea0398f1441e0b04d859c3"},
-    {file = "coverage-7.13.4-cp312-cp312-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:5b856a8ccf749480024ff3bd7310adaef57bf31fd17e1bfc404b7940b6986634"},
-    {file = "coverage-7.13.4-cp312-cp312-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:2c048ea43875fbf8b45d476ad79f179809c590ec7b79e2035c662e7afa3192e3"},
-    {file = "coverage-7.13.4-cp312-cp312-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:b7b38448866e83176e28086674fe7368ab8590e4610fb662b44e345b86d63ffa"},
-    {file = "coverage-7.13.4-cp312-cp312-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:de6defc1c9badbf8b9e67ae90fd00519186d6ab64e5cc5f3d21359c2a9b2c1d3"},
-    {file = "coverage-7.13.4-cp312-cp312-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:7eda778067ad7ffccd23ecffce537dface96212576a07924cbf0d8799d2ded5a"},
-    {file = "coverage-7.13.4-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:e87f6c587c3f34356c3759f0420693e35e7eb0e2e41e4c011cb6ec6ecbbf1db7"},
-    {file = "coverage-7.13.4-cp312-cp312-musllinux_1_2_i686.whl", hash = "sha256:8248977c2e33aecb2ced42fef99f2d319e9904a36e55a8a68b69207fb7e43edc"},
-    {file = "coverage-7.13.4-cp312-cp312-musllinux_1_2_ppc64le.whl", hash = "sha256:25381386e80ae727608e662474db537d4df1ecd42379b5ba33c84633a2b36d47"},
-    {file = "coverage-7.13.4-cp312-cp312-musllinux_1_2_riscv64.whl", hash = "sha256:ee756f00726693e5ba94d6df2bdfd64d4852d23b09bb0bc700e3b30e6f333985"},
-    {file = "coverage-7.13.4-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:fdfc1e28e7c7cdce44985b3043bc13bbd9c747520f94a4d7164af8260b3d91f0"},
-    {file = "coverage-7.13.4-cp312-cp312-win32.whl", hash = "sha256:01d4cbc3c283a17fc1e42d614a119f7f438eabb593391283adca8dc86eff1246"},
-    {file = "coverage-7.13.4-cp312-cp312-win_amd64.whl", hash = "sha256:9401ebc7ef522f01d01d45532c68c5ac40fb27113019b6b7d8b208f6e9baa126"},
-    {file = "coverage-7.13.4-cp312-cp312-win_arm64.whl", hash = "sha256:b1ec7b6b6e93255f952e27ab58fbc68dcc468844b16ecbee881aeb29b6ab4d8d"},
-    {file = "coverage-7.13.4-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:b66a2da594b6068b48b2692f043f35d4d3693fb639d5ea8b39533c2ad9ac3ab9"},
-    {file = "coverage-7.13.4-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:3599eb3992d814d23b35c536c28df1a882caa950f8f507cef23d1cbf334995ac"},
-    {file = "coverage-7.13.4-cp313-cp313-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:93550784d9281e374fb5a12bf1324cc8a963fd63b2d2f223503ef0fd4aa339ea"},
-    {file = "coverage-7.13.4-cp313-cp313-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:b720ce6a88a2755f7c697c23268ddc47a571b88052e6b155224347389fdf6a3b"},
-    {file = "coverage-7.13.4-cp313-cp313-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:7b322db1284a2ed3aa28ffd8ebe3db91c929b7a333c0820abec3d838ef5b3525"},
-    {file = "coverage-7.13.4-cp313-cp313-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:f4594c67d8a7c89cf922d9df0438c7c7bb022ad506eddb0fdb2863359ff78242"},
-    {file = "coverage-7.13.4-cp313-cp313-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:53d133df809c743eb8bce33b24bcababb371f4441340578cd406e084d94a6148"},
-    {file = "coverage-7.13.4-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:76451d1978b95ba6507a039090ba076105c87cc76fc3efd5d35d72093964d49a"},
-    {file = "coverage-7.13.4-cp313-cp313-musllinux_1_2_i686.whl", hash = "sha256:7f57b33491e281e962021de110b451ab8a24182589be17e12a22c79047935e23"},
-    {file = "coverage-7.13.4-cp313-cp313-musllinux_1_2_ppc64le.whl", hash = "sha256:1731dc33dc276dafc410a885cbf5992f1ff171393e48a21453b78727d090de80"},
-    {file = "coverage-7.13.4-cp313-cp313-musllinux_1_2_riscv64.whl", hash = "sha256:bd60d4fe2f6fa7dff9223ca1bbc9f05d2b6697bc5961072e5d3b952d46e1b1ea"},
-    {file = "coverage-7.13.4-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:9181a3ccead280b828fae232df12b16652702b49d41e99d657f46cc7b1f6ec7a"},
-    {file = "coverage-7.13.4-cp313-cp313-win32.whl", hash = "sha256:f53d492307962561ac7de4cd1de3e363589b000ab69617c6156a16ba7237998d"},
-    {file = "coverage-7.13.4-cp313-cp313-win_amd64.whl", hash = "sha256:e6f70dec1cc557e52df5306d051ef56003f74d56e9c4dd7ddb07e07ef32a84dd"},
-    {file = "coverage-7.13.4-cp313-cp313-win_arm64.whl", hash = "sha256:fb07dc5da7e849e2ad31a5d74e9bece81f30ecf5a42909d0a695f8bd1874d6af"},
-    {file = "coverage-7.13.4-cp313-cp313t-macosx_10_13_x86_64.whl", hash = "sha256:40d74da8e6c4b9ac18b15331c4b5ebc35a17069410cad462ad4f40dcd2d50c0d"},
-    {file = "coverage-7.13.4-cp313-cp313t-macosx_11_0_arm64.whl", hash = "sha256:4223b4230a376138939a9173f1bdd6521994f2aff8047fae100d6d94d50c5a12"},
-    {file = "coverage-7.13.4-cp313-cp313t-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:1d4be36a5114c499f9f1f9195e95ebf979460dbe2d88e6816ea202010ba1c34b"},
-    {file = "coverage-7.13.4-cp313-cp313t-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:200dea7d1e8095cc6e98cdabe3fd1d21ab17d3cee6dab00cadbb2fe35d9c15b9"},
-    {file = "coverage-7.13.4-cp313-cp313t-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:b8eb931ee8e6d8243e253e5ed7336deea6904369d2fd8ae6e43f68abbf167092"},
-    {file = "coverage-7.13.4-cp313-cp313t-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:75eab1ebe4f2f64d9509b984f9314d4aa788540368218b858dad56dc8f3e5eb9"},
-    {file = "coverage-7.13.4-cp313-cp313t-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:c35eb28c1d085eb7d8c9b3296567a1bebe03ce72962e932431b9a61f28facf26"},
-    {file = "coverage-7.13.4-cp313-cp313t-musllinux_1_2_aarch64.whl", hash = "sha256:eb88b316ec33760714a4720feb2816a3a59180fd58c1985012054fa7aebee4c2"},
-    {file = "coverage-7.13.4-cp313-cp313t-musllinux_1_2_i686.whl", hash = "sha256:7d41eead3cc673cbd38a4417deb7fd0b4ca26954ff7dc6078e33f6ff97bed940"},
-    {file = "coverage-7.13.4-cp313-cp313t-musllinux_1_2_ppc64le.whl", hash = "sha256:fb26a934946a6afe0e326aebe0730cdff393a8bc0bbb65a2f41e30feddca399c"},
-    {file = "coverage-7.13.4-cp313-cp313t-musllinux_1_2_riscv64.whl", hash = "sha256:dae88bc0fc77edaa65c14be099bd57ee140cf507e6bfdeea7938457ab387efb0"},
-    {file = "coverage-7.13.4-cp313-cp313t-musllinux_1_2_x86_64.whl", hash = "sha256:845f352911777a8e722bfce168958214951e07e47e5d5d9744109fa5fe77f79b"},
-    {file = "coverage-7.13.4-cp313-cp313t-win32.whl", hash = "sha256:2fa8d5f8de70688a28240de9e139fa16b153cc3cbb01c5f16d88d6505ebdadf9"},
-    {file = "coverage-7.13.4-cp313-cp313t-win_amd64.whl", hash = "sha256:9351229c8c8407645840edcc277f4a2d44814d1bc34a2128c11c2a031d45a5dd"},
-    {file = "coverage-7.13.4-cp313-cp313t-win_arm64.whl", hash = "sha256:30b8d0512f2dc8c8747557e8fb459d6176a2c9e5731e2b74d311c03b78451997"},
-    {file = "coverage-7.13.4-cp314-cp314-macosx_10_15_x86_64.whl", hash = "sha256:300deaee342f90696ed186e3a00c71b5b3d27bffe9e827677954f4ee56969601"},
-    {file = "coverage-7.13.4-cp314-cp314-macosx_11_0_arm64.whl", hash = "sha256:29e3220258d682b6226a9b0925bc563ed9a1ebcff3cad30f043eceea7eaf2689"},
-    {file = "coverage-7.13.4-cp314-cp314-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:391ee8f19bef69210978363ca930f7328081c6a0152f1166c91f0b5fdd2a773c"},
-    {file = "coverage-7.13.4-cp314-cp314-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:0dd7ab8278f0d58a0128ba2fca25824321f05d059c1441800e934ff2efa52129"},
-    {file = "coverage-7.13.4-cp314-cp314-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:78cdf0d578b15148b009ccf18c686aa4f719d887e76e6b40c38ffb61d264a552"},
-    {file = "coverage-7.13.4-cp314-cp314-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:48685fee12c2eb3b27c62f2658e7ea21e9c3239cba5a8a242801a0a3f6a8c62a"},
-    {file = "coverage-7.13.4-cp314-cp314-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:4e83efc079eb39480e6346a15a1bcb3e9b04759c5202d157e1dd4303cd619356"},
-    {file = "coverage-7.13.4-cp314-cp314-musllinux_1_2_aarch64.whl", hash = "sha256:ecae9737b72408d6a950f7e525f30aca12d4bd8dd95e37342e5beb3a2a8c4f71"},
-    {file = "coverage-7.13.4-cp314-cp314-musllinux_1_2_i686.whl", hash = "sha256:ae4578f8528569d3cf303fef2ea569c7f4c4059a38c8667ccef15c6e1f118aa5"},
-    {file = "coverage-7.13.4-cp314-cp314-musllinux_1_2_ppc64le.whl", hash = "sha256:6fdef321fdfbb30a197efa02d48fcd9981f0d8ad2ae8903ac318adc653f5df98"},
-    {file = "coverage-7.13.4-cp314-cp314-musllinux_1_2_riscv64.whl", hash = "sha256:2b0f6ccf3dbe577170bebfce1318707d0e8c3650003cb4b3a9dd744575daa8b5"},
-    {file = "coverage-7.13.4-cp314-cp314-musllinux_1_2_x86_64.whl", hash = "sha256:75fcd519f2a5765db3f0e391eb3b7d150cce1a771bf4c9f861aeab86c767a3c0"},
-    {file = "coverage-7.13.4-cp314-cp314-win32.whl", hash = "sha256:8e798c266c378da2bd819b0677df41ab46d78065fb2a399558f3f6cae78b2fbb"},
-    {file = "coverage-7.13.4-cp314-cp314-win_amd64.whl", hash = "sha256:245e37f664d89861cf2329c9afa2c1fe9e6d4e1a09d872c947e70718aeeac505"},
-    {file = "coverage-7.13.4-cp314-cp314-win_arm64.whl", hash = "sha256:ad27098a189e5838900ce4c2a99f2fe42a0bf0c2093c17c69b45a71579e8d4a2"},
-    {file = "coverage-7.13.4-cp314-cp314t-macosx_10_15_x86_64.whl", hash = "sha256:85480adfb35ffc32d40918aad81b89c69c9cc5661a9b8a81476d3e645321a056"},
-    {file = "coverage-7.13.4-cp314-cp314t-macosx_11_0_arm64.whl", hash = "sha256:79be69cf7f3bf9b0deeeb062eab7ac7f36cd4cc4c4dd694bd28921ba4d8596cc"},
-    {file = "coverage-7.13.4-cp314-cp314t-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:caa421e2684e382c5d8973ac55e4f36bed6821a9bad5c953494de960c74595c9"},
-    {file = "coverage-7.13.4-cp314-cp314t-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:14375934243ee05f56c45393fe2ce81fe5cc503c07cee2bdf1725fb8bef3ffaf"},
-    {file = "coverage-7.13.4-cp314-cp314t-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:25a41c3104d08edb094d9db0d905ca54d0cd41c928bb6be3c4c799a54753af55"},
-    {file = "coverage-7.13.4-cp314-cp314t-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:6f01afcff62bf9a08fb32b2c1d6e924236c0383c02c790732b6537269e466a72"},
-    {file = "coverage-7.13.4-cp314-cp314t-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:eb9078108fbf0bcdde37c3f4779303673c2fa1fe8f7956e68d447d0dd426d38a"},
-    {file = "coverage-7.13.4-cp314-cp314t-musllinux_1_2_aarch64.whl", hash = "sha256:0e086334e8537ddd17e5f16a344777c1ab8194986ec533711cbe6c41cde841b6"},
-    {file = "coverage-7.13.4-cp314-cp314t-musllinux_1_2_i686.whl", hash = "sha256:725d985c5ab621268b2edb8e50dfe57633dc69bda071abc470fed55a14935fd3"},
-    {file = "coverage-7.13.4-cp314-cp314t-musllinux_1_2_ppc64le.whl", hash = "sha256:3c06f0f1337c667b971ca2f975523347e63ec5e500b9aa5882d91931cd3ef750"},
-    {file = "coverage-7.13.4-cp314-cp314t-musllinux_1_2_riscv64.whl", hash = "sha256:590c0ed4bf8e85f745e6b805b2e1c457b2e33d5255dd9729743165253bc9ad39"},
-    {file = "coverage-7.13.4-cp314-cp314t-musllinux_1_2_x86_64.whl", hash = "sha256:eb30bf180de3f632cd043322dad5751390e5385108b2807368997d1a92a509d0"},
-    {file = "coverage-7.13.4-cp314-cp314t-win32.whl", hash = "sha256:c4240e7eded42d131a2d2c4dec70374b781b043ddc79a9de4d55ca71f8e98aea"},
-    {file = "coverage-7.13.4-cp314-cp314t-win_amd64.whl", hash = "sha256:4c7d3cc01e7350f2f0f6f7036caaf5673fb56b6998889ccfe9e1c1fe75a9c932"},
-    {file = "coverage-7.13.4-cp314-cp314t-win_arm64.whl", hash = "sha256:23e3f687cf945070d1c90f85db66d11e3025665d8dafa831301a0e0038f3db9b"},
-    {file = "coverage-7.13.4-py3-none-any.whl", hash = "sha256:1af1641e57cf7ba1bd67d677c9abdbcd6cc2ab7da3bca7fa1e2b7e50e65f2ad0"},
-    {file = "coverage-7.13.4.tar.gz", hash = "sha256:e5c8f6ed1e61a8b2dcdf31eb0b9bbf0130750ca79c1c49eb898e2ad86f5ccc91"},
+    {file = "coverage-7.10.5-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:c6a5c3414bfc7451b879141ce772c546985163cf553f08e0f135f0699a911801"},
+    {file = "coverage-7.10.5-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:bc8e4d99ce82f1710cc3c125adc30fd1487d3cf6c2cd4994d78d68a47b16989a"},
+    {file = "coverage-7.10.5-cp310-cp310-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:02252dc1216e512a9311f596b3169fad54abcb13827a8d76d5630c798a50a754"},
+    {file = "coverage-7.10.5-cp310-cp310-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:73269df37883e02d460bee0cc16be90509faea1e3bd105d77360b512d5bb9c33"},
+    {file = "coverage-7.10.5-cp310-cp310-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:1f8a81b0614642f91c9effd53eec284f965577591f51f547a1cbeb32035b4c2f"},
+    {file = "coverage-7.10.5-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:6a29f8e0adb7f8c2b95fa2d4566a1d6e6722e0a637634c6563cb1ab844427dd9"},
+    {file = "coverage-7.10.5-cp310-cp310-musllinux_1_2_i686.whl", hash = "sha256:fcf6ab569436b4a647d4e91accba12509ad9f2554bc93d3aee23cc596e7f99c3"},
+    {file = "coverage-7.10.5-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:90dc3d6fb222b194a5de60af8d190bedeeddcbc7add317e4a3cd333ee6b7c879"},
+    {file = "coverage-7.10.5-cp310-cp310-win32.whl", hash = "sha256:414a568cd545f9dc75f0686a0049393de8098414b58ea071e03395505b73d7a8"},
+    {file = "coverage-7.10.5-cp310-cp310-win_amd64.whl", hash = "sha256:e551f9d03347196271935fd3c0c165f0e8c049220280c1120de0084d65e9c7ff"},
+    {file = "coverage-7.10.5-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:c177e6ffe2ebc7c410785307758ee21258aa8e8092b44d09a2da767834f075f2"},
+    {file = "coverage-7.10.5-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:14d6071c51ad0f703d6440827eaa46386169b5fdced42631d5a5ac419616046f"},
+    {file = "coverage-7.10.5-cp311-cp311-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:61f78c7c3bc272a410c5ae3fde7792b4ffb4acc03d35a7df73ca8978826bb7ab"},
+    {file = "coverage-7.10.5-cp311-cp311-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:f39071caa126f69d63f99b324fb08c7b1da2ec28cbb1fe7b5b1799926492f65c"},
+    {file = "coverage-7.10.5-cp311-cp311-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:343a023193f04d46edc46b2616cdbee68c94dd10208ecd3adc56fcc54ef2baa1"},
+    {file = "coverage-7.10.5-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:585ffe93ae5894d1ebdee69fc0b0d4b7c75d8007983692fb300ac98eed146f78"},
+    {file = "coverage-7.10.5-cp311-cp311-musllinux_1_2_i686.whl", hash = "sha256:b0ef4e66f006ed181df29b59921bd8fc7ed7cd6a9289295cd8b2824b49b570df"},
+    {file = "coverage-7.10.5-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:eb7b0bbf7cc1d0453b843eca7b5fa017874735bef9bfdfa4121373d2cc885ed6"},
+    {file = "coverage-7.10.5-cp311-cp311-win32.whl", hash = "sha256:1d043a8a06987cc0c98516e57c4d3fc2c1591364831e9deb59c9e1b4937e8caf"},
+    {file = "coverage-7.10.5-cp311-cp311-win_amd64.whl", hash = "sha256:fefafcca09c3ac56372ef64a40f5fe17c5592fab906e0fdffd09543f3012ba50"},
+    {file = "coverage-7.10.5-cp311-cp311-win_arm64.whl", hash = "sha256:7e78b767da8b5fc5b2faa69bb001edafcd6f3995b42a331c53ef9572c55ceb82"},
+    {file = "coverage-7.10.5-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:c2d05c7e73c60a4cecc7d9b60dbfd603b4ebc0adafaef371445b47d0f805c8a9"},
+    {file = "coverage-7.10.5-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:32ddaa3b2c509778ed5373b177eb2bf5662405493baeff52278a0b4f9415188b"},
+    {file = "coverage-7.10.5-cp312-cp312-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:dd382410039fe062097aa0292ab6335a3f1e7af7bba2ef8d27dcda484918f20c"},
+    {file = "coverage-7.10.5-cp312-cp312-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:7fa22800f3908df31cea6fb230f20ac49e343515d968cc3a42b30d5c3ebf9b5a"},
+    {file = "coverage-7.10.5-cp312-cp312-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:f366a57ac81f5e12797136552f5b7502fa053c861a009b91b80ed51f2ce651c6"},
+    {file = "coverage-7.10.5-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:5f1dc8f1980a272ad4a6c84cba7981792344dad33bf5869361576b7aef42733a"},
+    {file = "coverage-7.10.5-cp312-cp312-musllinux_1_2_i686.whl", hash = "sha256:2285c04ee8676f7938b02b4936d9b9b672064daab3187c20f73a55f3d70e6b4a"},
+    {file = "coverage-7.10.5-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:c2492e4dd9daab63f5f56286f8a04c51323d237631eb98505d87e4c4ff19ec34"},
+    {file = "coverage-7.10.5-cp312-cp312-win32.whl", hash = "sha256:38a9109c4ee8135d5df5505384fc2f20287a47ccbe0b3f04c53c9a1989c2bbaf"},
+    {file = "coverage-7.10.5-cp312-cp312-win_amd64.whl", hash = "sha256:6b87f1ad60b30bc3c43c66afa7db6b22a3109902e28c5094957626a0143a001f"},
+    {file = "coverage-7.10.5-cp312-cp312-win_arm64.whl", hash = "sha256:672a6c1da5aea6c629819a0e1461e89d244f78d7b60c424ecf4f1f2556c041d8"},
+    {file = "coverage-7.10.5-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:ef3b83594d933020f54cf65ea1f4405d1f4e41a009c46df629dd964fcb6e907c"},
+    {file = "coverage-7.10.5-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:2b96bfdf7c0ea9faebce088a3ecb2382819da4fbc05c7b80040dbc428df6af44"},
+    {file = "coverage-7.10.5-cp313-cp313-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:63df1fdaffa42d914d5c4d293e838937638bf75c794cf20bee12978fc8c4e3bc"},
+    {file = "coverage-7.10.5-cp313-cp313-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:8002dc6a049aac0e81ecec97abfb08c01ef0c1fbf962d0c98da3950ace89b869"},
+    {file = "coverage-7.10.5-cp313-cp313-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:63d4bb2966d6f5f705a6b0c6784c8969c468dbc4bcf9d9ded8bff1c7e092451f"},
+    {file = "coverage-7.10.5-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:1f672efc0731a6846b157389b6e6d5d5e9e59d1d1a23a5c66a99fd58339914d5"},
+    {file = "coverage-7.10.5-cp313-cp313-musllinux_1_2_i686.whl", hash = "sha256:3f39cef43d08049e8afc1fde4a5da8510fc6be843f8dea350ee46e2a26b2f54c"},
+    {file = "coverage-7.10.5-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:2968647e3ed5a6c019a419264386b013979ff1fb67dd11f5c9886c43d6a31fc2"},
+    {file = "coverage-7.10.5-cp313-cp313-win32.whl", hash = "sha256:0d511dda38595b2b6934c2b730a1fd57a3635c6aa2a04cb74714cdfdd53846f4"},
+    {file = "coverage-7.10.5-cp313-cp313-win_amd64.whl", hash = "sha256:9a86281794a393513cf117177fd39c796b3f8e3759bb2764259a2abba5cce54b"},
+    {file = "coverage-7.10.5-cp313-cp313-win_arm64.whl", hash = "sha256:cebd8e906eb98bb09c10d1feed16096700b1198d482267f8bf0474e63a7b8d84"},
+    {file = "coverage-7.10.5-cp313-cp313t-macosx_10_13_x86_64.whl", hash = "sha256:0520dff502da5e09d0d20781df74d8189ab334a1e40d5bafe2efaa4158e2d9e7"},
+    {file = "coverage-7.10.5-cp313-cp313t-macosx_11_0_arm64.whl", hash = "sha256:d9cd64aca68f503ed3f1f18c7c9174cbb797baba02ca8ab5112f9d1c0328cd4b"},
+    {file = "coverage-7.10.5-cp313-cp313t-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:0913dd1613a33b13c4f84aa6e3f4198c1a21ee28ccb4f674985c1f22109f0aae"},
+    {file = "coverage-7.10.5-cp313-cp313t-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:1b7181c0feeb06ed8a02da02792f42f829a7b29990fef52eff257fef0885d760"},
+    {file = "coverage-7.10.5-cp313-cp313t-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:36d42b7396b605f774d4372dd9c49bed71cbabce4ae1ccd074d155709dd8f235"},
+    {file = "coverage-7.10.5-cp313-cp313t-musllinux_1_2_aarch64.whl", hash = "sha256:b4fdc777e05c4940b297bf47bf7eedd56a39a61dc23ba798e4b830d585486ca5"},
+    {file = "coverage-7.10.5-cp313-cp313t-musllinux_1_2_i686.whl", hash = "sha256:42144e8e346de44a6f1dbd0a56575dd8ab8dfa7e9007da02ea5b1c30ab33a7db"},
+    {file = "coverage-7.10.5-cp313-cp313t-musllinux_1_2_x86_64.whl", hash = "sha256:66c644cbd7aed8fe266d5917e2c9f65458a51cfe5eeff9c05f15b335f697066e"},
+    {file = "coverage-7.10.5-cp313-cp313t-win32.whl", hash = "sha256:2d1b73023854068c44b0c554578a4e1ef1b050ed07cf8b431549e624a29a66ee"},
+    {file = "coverage-7.10.5-cp313-cp313t-win_amd64.whl", hash = "sha256:54a1532c8a642d8cc0bd5a9a51f5a9dcc440294fd06e9dda55e743c5ec1a8f14"},
+    {file = "coverage-7.10.5-cp313-cp313t-win_arm64.whl", hash = "sha256:74d5b63fe3f5f5d372253a4ef92492c11a4305f3550631beaa432fc9df16fcff"},
+    {file = "coverage-7.10.5-cp314-cp314-macosx_10_13_x86_64.whl", hash = "sha256:68c5e0bc5f44f68053369fa0d94459c84548a77660a5f2561c5e5f1e3bed7031"},
+    {file = "coverage-7.10.5-cp314-cp314-macosx_11_0_arm64.whl", hash = "sha256:cf33134ffae93865e32e1e37df043bef15a5e857d8caebc0099d225c579b0fa3"},
+    {file = "coverage-7.10.5-cp314-cp314-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:ad8fa9d5193bafcf668231294241302b5e683a0518bf1e33a9a0dfb142ec3031"},
+    {file = "coverage-7.10.5-cp314-cp314-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:146fa1531973d38ab4b689bc764592fe6c2f913e7e80a39e7eeafd11f0ef6db2"},
+    {file = "coverage-7.10.5-cp314-cp314-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:6013a37b8a4854c478d3219ee8bc2392dea51602dd0803a12d6f6182a0061762"},
+    {file = "coverage-7.10.5-cp314-cp314-musllinux_1_2_aarch64.whl", hash = "sha256:eb90fe20db9c3d930fa2ad7a308207ab5b86bf6a76f54ab6a40be4012d88fcae"},
+    {file = "coverage-7.10.5-cp314-cp314-musllinux_1_2_i686.whl", hash = "sha256:384b34482272e960c438703cafe63316dfbea124ac62006a455c8410bf2a2262"},
+    {file = "coverage-7.10.5-cp314-cp314-musllinux_1_2_x86_64.whl", hash = "sha256:467dc74bd0a1a7de2bedf8deaf6811f43602cb532bd34d81ffd6038d6d8abe99"},
+    {file = "coverage-7.10.5-cp314-cp314-win32.whl", hash = "sha256:556d23d4e6393ca898b2e63a5bca91e9ac2d5fb13299ec286cd69a09a7187fde"},
+    {file = "coverage-7.10.5-cp314-cp314-win_amd64.whl", hash = "sha256:f4446a9547681533c8fa3e3c6cf62121eeee616e6a92bd9201c6edd91beffe13"},
+    {file = "coverage-7.10.5-cp314-cp314-win_arm64.whl", hash = "sha256:5e78bd9cf65da4c303bf663de0d73bf69f81e878bf72a94e9af67137c69b9fe9"},
+    {file = "coverage-7.10.5-cp314-cp314t-macosx_10_13_x86_64.whl", hash = "sha256:5661bf987d91ec756a47c7e5df4fbcb949f39e32f9334ccd3f43233bbb65e508"},
+    {file = "coverage-7.10.5-cp314-cp314t-macosx_11_0_arm64.whl", hash = "sha256:a46473129244db42a720439a26984f8c6f834762fc4573616c1f37f13994b357"},
+    {file = "coverage-7.10.5-cp314-cp314t-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:1f64b8d3415d60f24b058b58d859e9512624bdfa57a2d1f8aff93c1ec45c429b"},
+    {file = "coverage-7.10.5-cp314-cp314t-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:44d43de99a9d90b20e0163f9770542357f58860a26e24dc1d924643bd6aa7cb4"},
+    {file = "coverage-7.10.5-cp314-cp314t-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:a931a87e5ddb6b6404e65443b742cb1c14959622777f2a4efd81fba84f5d91ba"},
+    {file = "coverage-7.10.5-cp314-cp314t-musllinux_1_2_aarch64.whl", hash = "sha256:f9559b906a100029274448f4c8b8b0a127daa4dade5661dfd821b8c188058842"},
+    {file = "coverage-7.10.5-cp314-cp314t-musllinux_1_2_i686.whl", hash = "sha256:b08801e25e3b4526ef9ced1aa29344131a8f5213c60c03c18fe4c6170ffa2874"},
+    {file = "coverage-7.10.5-cp314-cp314t-musllinux_1_2_x86_64.whl", hash = "sha256:ed9749bb8eda35f8b636fb7632f1c62f735a236a5d4edadd8bbcc5ea0542e732"},
+    {file = "coverage-7.10.5-cp314-cp314t-win32.whl", hash = "sha256:609b60d123fc2cc63ccee6d17e4676699075db72d14ac3c107cc4976d516f2df"},
+    {file = "coverage-7.10.5-cp314-cp314t-win_amd64.whl", hash = "sha256:0666cf3d2c1626b5a3463fd5b05f5e21f99e6aec40a3192eee4d07a15970b07f"},
+    {file = "coverage-7.10.5-cp314-cp314t-win_arm64.whl", hash = "sha256:bc85eb2d35e760120540afddd3044a5bf69118a91a296a8b3940dfc4fdcfe1e2"},
+    {file = "coverage-7.10.5-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:62835c1b00c4a4ace24c1a88561a5a59b612fbb83a525d1c70ff5720c97c0610"},
+    {file = "coverage-7.10.5-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:5255b3bbcc1d32a4069d6403820ac8e6dbcc1d68cb28a60a1ebf17e47028e898"},
+    {file = "coverage-7.10.5-cp39-cp39-manylinux1_i686.manylinux_2_28_i686.manylinux_2_5_i686.whl", hash = "sha256:3876385722e335d6e991c430302c24251ef9c2a9701b2b390f5473199b1b8ebf"},
+    {file = "coverage-7.10.5-cp39-cp39-manylinux1_x86_64.manylinux_2_28_x86_64.manylinux_2_5_x86_64.whl", hash = "sha256:8048ce4b149c93447a55d279078c8ae98b08a6951a3c4d2d7e87f4efc7bfe100"},
+    {file = "coverage-7.10.5-cp39-cp39-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:4028e7558e268dd8bcf4d9484aad393cafa654c24b4885f6f9474bf53183a82a"},
+    {file = "coverage-7.10.5-cp39-cp39-musllinux_1_2_aarch64.whl", hash = "sha256:03f47dc870eec0367fcdd603ca6a01517d2504e83dc18dbfafae37faec66129a"},
+    {file = "coverage-7.10.5-cp39-cp39-musllinux_1_2_i686.whl", hash = "sha256:2d488d7d42b6ded7ea0704884f89dcabd2619505457de8fc9a6011c62106f6e5"},
+    {file = "coverage-7.10.5-cp39-cp39-musllinux_1_2_x86_64.whl", hash = "sha256:b3dcf2ead47fa8be14224ee817dfc1df98043af568fe120a22f81c0eb3c34ad2"},
+    {file = "coverage-7.10.5-cp39-cp39-win32.whl", hash = "sha256:02650a11324b80057b8c9c29487020073d5e98a498f1857f37e3f9b6ea1b2426"},
+    {file = "coverage-7.10.5-cp39-cp39-win_amd64.whl", hash = "sha256:b45264dd450a10f9e03237b41a9a24e85cbb1e278e5a32adb1a303f58f0017f3"},
+    {file = "coverage-7.10.5-py3-none-any.whl", hash = "sha256:0be24d35e4db1d23d0db5c0f6a74a962e2ec83c426b5cac09f4234aadef38e4a"},
+    {file = "coverage-7.10.5.tar.gz", hash = "sha256:f2e57716a78bc3ae80b2207be0709a3b2b63b9f2dcf9740ee6ac03588a2015b6"},
 ]

 [package.dependencies]
@@ -448,61 +430,61 @@ toml = ["tomli ; python_full_version <= \"3.11.0a6\""]

 [[package]]
 name = "cryptography"
-version = "46.0.5"
+version = "46.0.4"
 description = "cryptography is a package which provides cryptographic recipes and primitives to Python developers."
 optional = false
 python-versions = "!=3.9.0,!=3.9.1,>=3.8"
 groups = ["main"]
 files = [
-    {file = "cryptography-46.0.5-cp311-abi3-macosx_10_9_universal2.whl", hash = "sha256:351695ada9ea9618b3500b490ad54c739860883df6c1f555e088eaf25b1bbaad"},
-    {file = "cryptography-46.0.5-cp311-abi3-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:c18ff11e86df2e28854939acde2d003f7984f721eba450b56a200ad90eeb0e6b"},
-    {file = "cryptography-46.0.5-cp311-abi3-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:4d7e3d356b8cd4ea5aff04f129d5f66ebdc7b6f8eae802b93739ed520c47c79b"},
-    {file = "cryptography-46.0.5-cp311-abi3-manylinux_2_28_aarch64.whl", hash = "sha256:50bfb6925eff619c9c023b967d5b77a54e04256c4281b0e21336a130cd7fc263"},
-    {file = "cryptography-46.0.5-cp311-abi3-manylinux_2_28_ppc64le.whl", hash = "sha256:803812e111e75d1aa73690d2facc295eaefd4439be1023fefc4995eaea2af90d"},
-    {file = "cryptography-46.0.5-cp311-abi3-manylinux_2_28_x86_64.whl", hash = "sha256:3ee190460e2fbe447175cda91b88b84ae8322a104fc27766ad09428754a618ed"},
-    {file = "cryptography-46.0.5-cp311-abi3-manylinux_2_31_armv7l.whl", hash = "sha256:f145bba11b878005c496e93e257c1e88f154d278d2638e6450d17e0f31e558d2"},
-    {file = "cryptography-46.0.5-cp311-abi3-manylinux_2_34_aarch64.whl", hash = "sha256:e9251e3be159d1020c4030bd2e5f84d6a43fe54b6c19c12f51cde9542a2817b2"},
-    {file = "cryptography-46.0.5-cp311-abi3-manylinux_2_34_ppc64le.whl", hash = "sha256:47fb8a66058b80e509c47118ef8a75d14c455e81ac369050f20ba0d23e77fee0"},
-    {file = "cryptography-46.0.5-cp311-abi3-manylinux_2_34_x86_64.whl", hash = "sha256:4c3341037c136030cb46e4b1e17b7418ea4cbd9dd207e4a6f3b2b24e0d4ac731"},
-    {file = "cryptography-46.0.5-cp311-abi3-musllinux_1_2_aarch64.whl", hash = "sha256:890bcb4abd5a2d3f852196437129eb3667d62630333aacc13dfd470fad3aaa82"},
-    {file = "cryptography-46.0.5-cp311-abi3-musllinux_1_2_x86_64.whl", hash = "sha256:80a8d7bfdf38f87ca30a5391c0c9ce4ed2926918e017c29ddf643d0ed2778ea1"},
-    {file = "cryptography-46.0.5-cp311-abi3-win32.whl", hash = "sha256:60ee7e19e95104d4c03871d7d7dfb3d22ef8a9b9c6778c94e1c8fcc8365afd48"},
-    {file = "cryptography-46.0.5-cp311-abi3-win_amd64.whl", hash = "sha256:38946c54b16c885c72c4f59846be9743d699eee2b69b6988e0a00a01f46a61a4"},
-    {file = "cryptography-46.0.5-cp314-cp314t-macosx_10_9_universal2.whl", hash = "sha256:94a76daa32eb78d61339aff7952ea819b1734b46f73646a07decb40e5b3448e2"},
-    {file = "cryptography-46.0.5-cp314-cp314t-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:5be7bf2fb40769e05739dd0046e7b26f9d4670badc7b032d6ce4db64dddc0678"},
-    {file = "cryptography-46.0.5-cp314-cp314t-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:fe346b143ff9685e40192a4960938545c699054ba11d4f9029f94751e3f71d87"},
-    {file = "cryptography-46.0.5-cp314-cp314t-manylinux_2_28_aarch64.whl", hash = "sha256:c69fd885df7d089548a42d5ec05be26050ebcd2283d89b3d30676eb32ff87dee"},
-    {file = "cryptography-46.0.5-cp314-cp314t-manylinux_2_28_ppc64le.whl", hash = "sha256:8293f3dea7fc929ef7240796ba231413afa7b68ce38fd21da2995549f5961981"},
-    {file = "cryptography-46.0.5-cp314-cp314t-manylinux_2_28_x86_64.whl", hash = "sha256:1abfdb89b41c3be0365328a410baa9df3ff8a9110fb75e7b52e66803ddabc9a9"},
-    {file = "cryptography-46.0.5-cp314-cp314t-manylinux_2_31_armv7l.whl", hash = "sha256:d66e421495fdb797610a08f43b05269e0a5ea7f5e652a89bfd5a7d3c1dee3648"},
-    {file = "cryptography-46.0.5-cp314-cp314t-manylinux_2_34_aarch64.whl", hash = "sha256:4e817a8920bfbcff8940ecfd60f23d01836408242b30f1a708d93198393a80b4"},
-    {file = "cryptography-46.0.5-cp314-cp314t-manylinux_2_34_ppc64le.whl", hash = "sha256:68f68d13f2e1cb95163fa3b4db4bf9a159a418f5f6e7242564fc75fcae667fd0"},
-    {file = "cryptography-46.0.5-cp314-cp314t-manylinux_2_34_x86_64.whl", hash = "sha256:a3d1fae9863299076f05cb8a778c467578262fae09f9dc0ee9b12eb4268ce663"},
-    {file = "cryptography-46.0.5-cp314-cp314t-musllinux_1_2_aarch64.whl", hash = "sha256:c4143987a42a2397f2fc3b4d7e3a7d313fbe684f67ff443999e803dd75a76826"},
-    {file = "cryptography-46.0.5-cp314-cp314t-musllinux_1_2_x86_64.whl", hash = "sha256:7d731d4b107030987fd61a7f8ab512b25b53cef8f233a97379ede116f30eb67d"},
-    {file = "cryptography-46.0.5-cp314-cp314t-win32.whl", hash = "sha256:c3bcce8521d785d510b2aad26ae2c966092b7daa8f45dd8f44734a104dc0bc1a"},
-    {file = "cryptography-46.0.5-cp314-cp314t-win_amd64.whl", hash = "sha256:4d8ae8659ab18c65ced284993c2265910f6c9e650189d4e3f68445ef82a810e4"},
-    {file = "cryptography-46.0.5-cp38-abi3-macosx_10_9_universal2.whl", hash = "sha256:4108d4c09fbbf2789d0c926eb4152ae1760d5a2d97612b92d508d96c861e4d31"},
-    {file = "cryptography-46.0.5-cp38-abi3-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:7d1f30a86d2757199cb2d56e48cce14deddf1f9c95f1ef1b64ee91ea43fe2e18"},
-    {file = "cryptography-46.0.5-cp38-abi3-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:039917b0dc418bb9f6edce8a906572d69e74bd330b0b3fea4f79dab7f8ddd235"},
-    {file = "cryptography-46.0.5-cp38-abi3-manylinux_2_28_aarch64.whl", hash = "sha256:ba2a27ff02f48193fc4daeadf8ad2590516fa3d0adeeb34336b96f7fa64c1e3a"},
-    {file = "cryptography-46.0.5-cp38-abi3-manylinux_2_28_ppc64le.whl", hash = "sha256:61aa400dce22cb001a98014f647dc21cda08f7915ceb95df0c9eaf84b4b6af76"},
-    {file = "cryptography-46.0.5-cp38-abi3-manylinux_2_28_x86_64.whl", hash = "sha256:3ce58ba46e1bc2aac4f7d9290223cead56743fa6ab94a5d53292ffaac6a91614"},
-    {file = "cryptography-46.0.5-cp38-abi3-manylinux_2_31_armv7l.whl", hash = "sha256:420d0e909050490d04359e7fdb5ed7e667ca5c3c402b809ae2563d7e66a92229"},
-    {file = "cryptography-46.0.5-cp38-abi3-manylinux_2_34_aarch64.whl", hash = "sha256:582f5fcd2afa31622f317f80426a027f30dc792e9c80ffee87b993200ea115f1"},
-    {file = "cryptography-46.0.5-cp38-abi3-manylinux_2_34_ppc64le.whl", hash = "sha256:bfd56bb4b37ed4f330b82402f6f435845a5f5648edf1ad497da51a8452d5d62d"},
-    {file = "cryptography-46.0.5-cp38-abi3-manylinux_2_34_x86_64.whl", hash = "sha256:a3d507bb6a513ca96ba84443226af944b0f7f47dcc9a399d110cd6146481d24c"},
-    {file = "cryptography-46.0.5-cp38-abi3-musllinux_1_2_aarch64.whl", hash = "sha256:9f16fbdf4da055efb21c22d81b89f155f02ba420558db21288b3d0035bafd5f4"},
-    {file = "cryptography-46.0.5-cp38-abi3-musllinux_1_2_x86_64.whl", hash = "sha256:ced80795227d70549a411a4ab66e8ce307899fad2220ce5ab2f296e687eacde9"},
-    {file = "cryptography-46.0.5-cp38-abi3-win32.whl", hash = "sha256:02f547fce831f5096c9a567fd41bc12ca8f11df260959ecc7c3202555cc47a72"},
-    {file = "cryptography-46.0.5-cp38-abi3-win_amd64.whl", hash = "sha256:556e106ee01aa13484ce9b0239bca667be5004efb0aabbed28d353df86445595"},
-    {file = "cryptography-46.0.5-pp311-pypy311_pp73-macosx_11_0_arm64.whl", hash = "sha256:3b4995dc971c9fb83c25aa44cf45f02ba86f71ee600d81091c2f0cbae116b06c"},
-    {file = "cryptography-46.0.5-pp311-pypy311_pp73-manylinux_2_28_aarch64.whl", hash = "sha256:bc84e875994c3b445871ea7181d424588171efec3e185dced958dad9e001950a"},
-    {file = "cryptography-46.0.5-pp311-pypy311_pp73-manylinux_2_28_x86_64.whl", hash = "sha256:2ae6971afd6246710480e3f15824ed3029a60fc16991db250034efd0b9fb4356"},
-    {file = "cryptography-46.0.5-pp311-pypy311_pp73-manylinux_2_34_aarch64.whl", hash = "sha256:d861ee9e76ace6cf36a6a89b959ec08e7bc2493ee39d07ffe5acb23ef46d27da"},
-    {file = "cryptography-46.0.5-pp311-pypy311_pp73-manylinux_2_34_x86_64.whl", hash = "sha256:2b7a67c9cd56372f3249b39699f2ad479f6991e62ea15800973b956f4b73e257"},
-    {file = "cryptography-46.0.5-pp311-pypy311_pp73-win_amd64.whl", hash = "sha256:8456928655f856c6e1533ff59d5be76578a7157224dbd9ce6872f25055ab9ab7"},
-    {file = "cryptography-46.0.5.tar.gz", hash = "sha256:abace499247268e3757271b2f1e244b36b06f8515cf27c4d49468fc9eb16e93d"},
+    {file = "cryptography-46.0.4-cp311-abi3-macosx_10_9_universal2.whl", hash = "sha256:281526e865ed4166009e235afadf3a4c4cba6056f99336a99efba65336fd5485"},
+    {file = "cryptography-46.0.4-cp311-abi3-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:5f14fba5bf6f4390d7ff8f086c566454bff0411f6d8aa7af79c88b6f9267aecc"},
+    {file = "cryptography-46.0.4-cp311-abi3-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:47bcd19517e6389132f76e2d5303ded6cf3f78903da2158a671be8de024f4cd0"},
+    {file = "cryptography-46.0.4-cp311-abi3-manylinux_2_28_aarch64.whl", hash = "sha256:01df4f50f314fbe7009f54046e908d1754f19d0c6d3070df1e6268c5a4af09fa"},
+    {file = "cryptography-46.0.4-cp311-abi3-manylinux_2_28_ppc64le.whl", hash = "sha256:5aa3e463596b0087b3da0dbe2b2487e9fc261d25da85754e30e3b40637d61f81"},
+    {file = "cryptography-46.0.4-cp311-abi3-manylinux_2_28_x86_64.whl", hash = "sha256:0a9ad24359fee86f131836a9ac3bffc9329e956624a2d379b613f8f8abaf5255"},
+    {file = "cryptography-46.0.4-cp311-abi3-manylinux_2_31_armv7l.whl", hash = "sha256:dc1272e25ef673efe72f2096e92ae39dea1a1a450dd44918b15351f72c5a168e"},
+    {file = "cryptography-46.0.4-cp311-abi3-manylinux_2_34_aarch64.whl", hash = "sha256:de0f5f4ec8711ebc555f54735d4c673fc34b65c44283895f1a08c2b49d2fd99c"},
+    {file = "cryptography-46.0.4-cp311-abi3-manylinux_2_34_ppc64le.whl", hash = "sha256:eeeb2e33d8dbcccc34d64651f00a98cb41b2dc69cef866771a5717e6734dfa32"},
+    {file = "cryptography-46.0.4-cp311-abi3-manylinux_2_34_x86_64.whl", hash = "sha256:3d425eacbc9aceafd2cb429e42f4e5d5633c6f873f5e567077043ef1b9bbf616"},
+    {file = "cryptography-46.0.4-cp311-abi3-musllinux_1_2_aarch64.whl", hash = "sha256:91627ebf691d1ea3976a031b61fb7bac1ccd745afa03602275dda443e11c8de0"},
+    {file = "cryptography-46.0.4-cp311-abi3-musllinux_1_2_x86_64.whl", hash = "sha256:2d08bc22efd73e8854b0b7caff402d735b354862f1145d7be3b9c0f740fef6a0"},
+    {file = "cryptography-46.0.4-cp311-abi3-win32.whl", hash = "sha256:82a62483daf20b8134f6e92898da70d04d0ef9a75829d732ea1018678185f4f5"},
+    {file = "cryptography-46.0.4-cp311-abi3-win_amd64.whl", hash = "sha256:6225d3ebe26a55dbc8ead5ad1265c0403552a63336499564675b29eb3184c09b"},
+    {file = "cryptography-46.0.4-cp314-cp314t-macosx_10_9_universal2.whl", hash = "sha256:485e2b65d25ec0d901bca7bcae0f53b00133bf3173916d8e421f6fddde103908"},
+    {file = "cryptography-46.0.4-cp314-cp314t-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:078e5f06bd2fa5aea5a324f2a09f914b1484f1d0c2a4d6a8a28c74e72f65f2da"},
+    {file = "cryptography-46.0.4-cp314-cp314t-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:dce1e4f068f03008da7fa51cc7abc6ddc5e5de3e3d1550334eaf8393982a5829"},
+    {file = "cryptography-46.0.4-cp314-cp314t-manylinux_2_28_aarch64.whl", hash = "sha256:2067461c80271f422ee7bdbe79b9b4be54a5162e90345f86a23445a0cf3fd8a2"},
+    {file = "cryptography-46.0.4-cp314-cp314t-manylinux_2_28_ppc64le.whl", hash = "sha256:c92010b58a51196a5f41c3795190203ac52edfd5dc3ff99149b4659eba9d2085"},
+    {file = "cryptography-46.0.4-cp314-cp314t-manylinux_2_28_x86_64.whl", hash = "sha256:829c2b12bbc5428ab02d6b7f7e9bbfd53e33efd6672d21341f2177470171ad8b"},
+    {file = "cryptography-46.0.4-cp314-cp314t-manylinux_2_31_armv7l.whl", hash = "sha256:62217ba44bf81b30abaeda1488686a04a702a261e26f87db51ff61d9d3510abd"},
+    {file = "cryptography-46.0.4-cp314-cp314t-manylinux_2_34_aarch64.whl", hash = "sha256:9c2da296c8d3415b93e6053f5a728649a87a48ce084a9aaf51d6e46c87c7f2d2"},
+    {file = "cryptography-46.0.4-cp314-cp314t-manylinux_2_34_ppc64le.whl", hash = "sha256:9b34d8ba84454641a6bf4d6762d15847ecbd85c1316c0a7984e6e4e9f748ec2e"},
+    {file = "cryptography-46.0.4-cp314-cp314t-manylinux_2_34_x86_64.whl", hash = "sha256:df4a817fa7138dd0c96c8c8c20f04b8aaa1fac3bbf610913dcad8ea82e1bfd3f"},
+    {file = "cryptography-46.0.4-cp314-cp314t-musllinux_1_2_aarch64.whl", hash = "sha256:b1de0ebf7587f28f9190b9cb526e901bf448c9e6a99655d2b07fff60e8212a82"},
+    {file = "cryptography-46.0.4-cp314-cp314t-musllinux_1_2_x86_64.whl", hash = "sha256:9b4d17bc7bd7cdd98e3af40b441feaea4c68225e2eb2341026c84511ad246c0c"},
+    {file = "cryptography-46.0.4-cp314-cp314t-win32.whl", hash = "sha256:c411f16275b0dea722d76544a61d6421e2cc829ad76eec79280dbdc9ddf50061"},
+    {file = "cryptography-46.0.4-cp314-cp314t-win_amd64.whl", hash = "sha256:728fedc529efc1439eb6107b677f7f7558adab4553ef8669f0d02d42d7b959a7"},
+    {file = "cryptography-46.0.4-cp38-abi3-macosx_10_9_universal2.whl", hash = "sha256:a9556ba711f7c23f77b151d5798f3ac44a13455cc68db7697a1096e6d0563cab"},
+    {file = "cryptography-46.0.4-cp38-abi3-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:8bf75b0259e87fa70bddc0b8b4078b76e7fd512fd9afae6c1193bcf440a4dbef"},
+    {file = "cryptography-46.0.4-cp38-abi3-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:3c268a3490df22270955966ba236d6bc4a8f9b6e4ffddb78aac535f1a5ea471d"},
+    {file = "cryptography-46.0.4-cp38-abi3-manylinux_2_28_aarch64.whl", hash = "sha256:812815182f6a0c1d49a37893a303b44eaac827d7f0d582cecfc81b6427f22973"},
+    {file = "cryptography-46.0.4-cp38-abi3-manylinux_2_28_ppc64le.whl", hash = "sha256:a90e43e3ef65e6dcf969dfe3bb40cbf5aef0d523dff95bfa24256be172a845f4"},
+    {file = "cryptography-46.0.4-cp38-abi3-manylinux_2_28_x86_64.whl", hash = "sha256:a05177ff6296644ef2876fce50518dffb5bcdf903c85250974fc8bc85d54c0af"},
+    {file = "cryptography-46.0.4-cp38-abi3-manylinux_2_31_armv7l.whl", hash = "sha256:daa392191f626d50f1b136c9b4cf08af69ca8279d110ea24f5c2700054d2e263"},
+    {file = "cryptography-46.0.4-cp38-abi3-manylinux_2_34_aarch64.whl", hash = "sha256:e07ea39c5b048e085f15923511d8121e4a9dc45cee4e3b970ca4f0d338f23095"},
+    {file = "cryptography-46.0.4-cp38-abi3-manylinux_2_34_ppc64le.whl", hash = "sha256:d5a45ddc256f492ce42a4e35879c5e5528c09cd9ad12420828c972951d8e016b"},
+    {file = "cryptography-46.0.4-cp38-abi3-manylinux_2_34_x86_64.whl", hash = "sha256:6bb5157bf6a350e5b28aee23beb2d84ae6f5be390b2f8ee7ea179cda077e1019"},
+    {file = "cryptography-46.0.4-cp38-abi3-musllinux_1_2_aarch64.whl", hash = "sha256:dd5aba870a2c40f87a3af043e0dee7d9eb02d4aff88a797b48f2b43eff8c3ab4"},
+    {file = "cryptography-46.0.4-cp38-abi3-musllinux_1_2_x86_64.whl", hash = "sha256:93d8291da8d71024379ab2cb0b5c57915300155ad42e07f76bea6ad838d7e59b"},
+    {file = "cryptography-46.0.4-cp38-abi3-win32.whl", hash = "sha256:0563655cb3c6d05fb2afe693340bc050c30f9f34e15763361cf08e94749401fc"},
+    {file = "cryptography-46.0.4-cp38-abi3-win_amd64.whl", hash = "sha256:fa0900b9ef9c49728887d1576fd8d9e7e3ea872fa9b25ef9b64888adc434e976"},
+    {file = "cryptography-46.0.4-pp311-pypy311_pp73-macosx_11_0_arm64.whl", hash = "sha256:766330cce7416c92b5e90c3bb71b1b79521760cdcfc3a6a1a182d4c9fab23d2b"},
+    {file = "cryptography-46.0.4-pp311-pypy311_pp73-manylinux_2_28_aarch64.whl", hash = "sha256:c236a44acfb610e70f6b3e1c3ca20ff24459659231ef2f8c48e879e2d32b73da"},
+    {file = "cryptography-46.0.4-pp311-pypy311_pp73-manylinux_2_28_x86_64.whl", hash = "sha256:8a15fb869670efa8f83cbffbc8753c1abf236883225aed74cd179b720ac9ec80"},
+    {file = "cryptography-46.0.4-pp311-pypy311_pp73-manylinux_2_34_aarch64.whl", hash = "sha256:fdc3daab53b212472f1524d070735b2f0c214239df131903bae1d598016fa822"},
+    {file = "cryptography-46.0.4-pp311-pypy311_pp73-manylinux_2_34_x86_64.whl", hash = "sha256:44cc0675b27cadb71bdbb96099cca1fa051cd11d2ade09e5cd3a2edb929ed947"},
+    {file = "cryptography-46.0.4-pp311-pypy311_pp73-win_amd64.whl", hash = "sha256:be8c01a7d5a55f9a47d1888162b76c8f49d62b234d88f0ff91a9fbebe32ffbc3"},
+    {file = "cryptography-46.0.4.tar.gz", hash = "sha256:bfd019f60f8abc2ed1b9be4ddc21cfef059c841d86d710bb69909a688cbb8f59"},
 ]

 [package.dependencies]
@@ -516,7 +498,7 @@ nox = ["nox[uv] (>=2024.4.15)"]
 pep8test = ["check-sdist", "click (>=8.0.1)", "mypy (>=1.14)", "ruff (>=0.11.11)"]
 sdist = ["build (>=1.0.0)"]
 ssh = ["bcrypt (>=3.1.5)"]
-test = ["certifi (>=2024)", "cryptography-vectors (==46.0.5)", "pretend (>=0.7)", "pytest (>=7.4.0)", "pytest-benchmark (>=4.0)", "pytest-cov (>=2.10.1)", "pytest-xdist (>=3.5.0)"]
+test = ["certifi (>=2024)", "cryptography-vectors (==46.0.4)", "pretend (>=0.7)", "pytest (>=7.4.0)", "pytest-benchmark (>=4.0)", "pytest-cov (>=2.10.1)", "pytest-xdist (>=3.5.0)"]
 test-randomorder = ["pytest-randomly"]

 [[package]]
@@ -541,7 +523,7 @@ description = "Backport of PEP 654 (exception groups)"
 optional = false
 python-versions = ">=3.7"
 groups = ["main", "dev"]
-markers = "python_version < \"3.11\""
+markers = "python_version == \"3.10\""
 files = [
    {file = "exceptiongroup-1.3.0-py3-none-any.whl", hash = "sha256:4d111e6e0c13d0644cad6ddaa7ed0261a0b36971f6d23e7ec9b4b9097da78a10"},
    {file = "exceptiongroup-1.3.0.tar.gz", hash = "sha256:b241f5885f560bc56a59ee63ca4c6a8bfa46ae4ad651af316d4e81817bb9fd88"},
@@ -570,25 +552,24 @@ tests = ["coverage", "coveralls", "dill", "mock", "nose"]

 [[package]]
 name = "fastapi"
-version = "0.128.7"
+version = "0.128.0"
 description = "FastAPI framework, high performance, easy to learn, fast to code, ready for production"
 optional = false
 python-versions = ">=3.9"
 groups = ["main"]
 files = [
-    {file = "fastapi-0.128.7-py3-none-any.whl", hash = "sha256:6bd9bd31cb7047465f2d3fa3ba3f33b0870b17d4eaf7cdb36d1576ab060ad662"},
-    {file = "fastapi-0.128.7.tar.gz", hash = "sha256:783c273416995486c155ad2c0e2b45905dedfaf20b9ef8d9f6a9124670639a24"},
+    {file = "fastapi-0.128.0-py3-none-any.whl", hash = "sha256:aebd93f9716ee3b4f4fcfe13ffb7cf308d99c9f3ab5622d8877441072561582d"},
+    {file = "fastapi-0.128.0.tar.gz", hash = "sha256:1cc179e1cef10a6be60ffe429f79b829dce99d8de32d7acb7e6c8dfdf7f2645a"},
 ]

 [package.dependencies]
 annotated-doc = ">=0.0.2"
 pydantic = ">=2.7.0"
-starlette = ">=0.40.0,<1.0.0"
+starlette = ">=0.40.0,<0.51.0"
 typing-extensions = ">=4.8.0"
-typing-inspection = ">=0.4.2"

 [package.extras]
-all = ["email-validator (>=2.0.0)", "fastapi-cli[standard] (>=0.0.8)", "httpx (>=0.23.0,<1.0.0)", "itsdangerous (>=1.1.0)", "jinja2 (>=3.1.5)", "orjson (>=3.9.3)", "pydantic-extra-types (>=2.0.0)", "pydantic-settings (>=2.0.0)", "python-multipart (>=0.0.18)", "pyyaml (>=5.3.1)", "ujson (>=5.8.0)", "uvicorn[standard] (>=0.12.0)"]
+all = ["email-validator (>=2.0.0)", "fastapi-cli[standard] (>=0.0.8)", "httpx (>=0.23.0,<1.0.0)", "itsdangerous (>=1.1.0)", "jinja2 (>=3.1.5)", "orjson (>=3.2.1)", "pydantic-extra-types (>=2.0.0)", "pydantic-settings (>=2.0.0)", "python-multipart (>=0.0.18)", "pyyaml (>=5.3.1)", "ujson (>=4.0.1,!=4.0.2,!=4.1.0,!=4.2.0,!=4.3.0,!=5.0.0,!=5.1.0)", "uvicorn[standard] (>=0.12.0)"]
 standard = ["email-validator (>=2.0.0)", "fastapi-cli[standard] (>=0.0.8)", "httpx (>=0.23.0,<1.0.0)", "jinja2 (>=3.1.5)", "pydantic-extra-types (>=2.0.0)", "pydantic-settings (>=2.0.0)", "python-multipart (>=0.0.18)", "uvicorn[standard] (>=0.12.0)"]
 standard-no-fastapi-cloud-cli = ["email-validator (>=2.0.0)", "fastapi-cli[standard-no-fastapi-cloud-cli] (>=0.0.8)", "httpx (>=0.23.0,<1.0.0)", "jinja2 (>=3.1.5)", "pydantic-extra-types (>=2.0.0)", "pydantic-settings (>=2.0.0)", "python-multipart (>=0.0.18)", "uvicorn[standard] (>=0.12.0)"]

@@ -1063,14 +1044,14 @@ urllib3 = ">=1.26.0,<3"

 [[package]]
 name = "launchdarkly-server-sdk"
-version = "9.15.0"
+version = "9.14.1"
 description = "LaunchDarkly SDK for Python"
 optional = false
-python-versions = ">=3.10"
+python-versions = ">=3.9"
 groups = ["main"]
 files = [
-    {file = "launchdarkly_server_sdk-9.15.0-py3-none-any.whl", hash = "sha256:c267e29bfa3fb5e2a06a208448ada6ed5557a2924979b8d79c970b45d227c668"},
-    {file = "launchdarkly_server_sdk-9.15.0.tar.gz", hash = "sha256:f31441b74bc1a69c381db57c33116509e407a2612628ad6dff0a7dbb39d5020b"},
+    {file = "launchdarkly_server_sdk-9.14.1-py3-none-any.whl", hash = "sha256:a9e2bd9ecdef845cd631ae0d4334a1115e5b44257c42eb2349492be4bac7815c"},
+    {file = "launchdarkly_server_sdk-9.14.1.tar.gz", hash = "sha256:1df44baf0a0efa74d8c1dad7a00592b98bce7d19edded7f770da8dbc49922213"},
 ]

 [package.dependencies]
@@ -1479,14 +1460,14 @@ testing = ["coverage", "pytest", "pytest-benchmark"]

 [[package]]
 name = "postgrest"
-version = "2.28.0"
+version = "2.27.2"
 description = "PostgREST client for Python. This library provides an ORM interface to PostgREST."
 optional = false
 python-versions = ">=3.9"
 groups = ["main"]
 files = [
-    {file = "postgrest-2.28.0-py3-none-any.whl", hash = "sha256:7bca2f24dd1a1bf8a3d586c7482aba6cd41662da6733045fad585b63b7f7df75"},
-    {file = "postgrest-2.28.0.tar.gz", hash = "sha256:c36b38646d25ea4255321d3d924ce70f8d20ec7799cb42c1221d6a818d4f6515"},
+    {file = "postgrest-2.27.2-py3-none-any.whl", hash = "sha256:1666fef3de05ca097a314433dd5ae2f2d71c613cb7b233d0f468c4ffe37277da"},
+    {file = "postgrest-2.27.2.tar.gz", hash = "sha256:55407d530b5af3d64e883a71fec1f345d369958f723ce4a8ab0b7d169e313242"},
 ]

 [package.dependencies]
@@ -2181,23 +2162,23 @@ testing = ["coverage (>=6.2)", "hypothesis (>=5.7.1)"]

 [[package]]
 name = "pytest-cov"
-version = "7.0.0"
+version = "6.2.1"
 description = "Pytest plugin for measuring coverage."
 optional = false
 python-versions = ">=3.9"
 groups = ["dev"]
 files = [
-    {file = "pytest_cov-7.0.0-py3-none-any.whl", hash = "sha256:3b8e9558b16cc1479da72058bdecf8073661c7f57f7d3c5f22a1c23507f2d861"},
-    {file = "pytest_cov-7.0.0.tar.gz", hash = "sha256:33c97eda2e049a0c5298e91f519302a1334c26ac65c1a483d6206fd458361af1"},
+    {file = "pytest_cov-6.2.1-py3-none-any.whl", hash = "sha256:f5bc4c23f42f1cdd23c70b1dab1bbaef4fc505ba950d53e0081d0730dd7e86d5"},
+    {file = "pytest_cov-6.2.1.tar.gz", hash = "sha256:25cc6cc0a5358204b8108ecedc51a9b57b34cc6b8c967cc2c01a4e00d8a67da2"},
 ]

 [package.dependencies]
-coverage = {version = ">=7.10.6", extras = ["toml"]}
+coverage = {version = ">=7.5", extras = ["toml"]}
 pluggy = ">=1.2"
-pytest = ">=7"
+pytest = ">=6.2.5"

 [package.extras]
-testing = ["process-tests", "pytest-xdist", "virtualenv"]
+testing = ["fields", "hunter", "process-tests", "pytest-xdist", "virtualenv"]

 [[package]]
 name = "pytest-mock"
@@ -2249,14 +2230,14 @@ cli = ["click (>=5.0)"]

 [[package]]
 name = "realtime"
-version = "2.28.0"
+version = "2.27.2"
 description = ""
 optional = false
 python-versions = ">=3.9"
 groups = ["main"]
 files = [
-    {file = "realtime-2.28.0-py3-none-any.whl", hash = "sha256:db1bd59bab9b1fcc9f9d3b1a073bed35bf4994d720e6751f10031a58d57a3836"},
-    {file = "realtime-2.28.0.tar.gz", hash = "sha256:d18cedcebd6a8f22fcd509bc767f639761eb218b7b2b6f14fc4205b6259b50fc"},
+    {file = "realtime-2.27.2-py3-none-any.whl", hash = "sha256:34a9cbb26a274e707e8fc9e3ee0a66de944beac0fe604dc336d1e985db2c830f"},
+    {file = "realtime-2.27.2.tar.gz", hash = "sha256:b960a90294d2cea1b3f1275ecb89204304728e08fff1c393cc1b3150739556b3"},
 ]

 [package.dependencies]
@@ -2437,14 +2418,14 @@ full = ["httpx (>=0.27.0,<0.29.0)", "itsdangerous", "jinja2", "python-multipart

 [[package]]
 name = "storage3"
-version = "2.28.0"
+version = "2.27.2"
 description = "Supabase Storage client for Python."
 optional = false
 python-versions = ">=3.9"
 groups = ["main"]
 files = [
-    {file = "storage3-2.28.0-py3-none-any.whl", hash = "sha256:ecb50efd2ac71dabbdf97e99ad346eafa630c4c627a8e5a138ceb5fbbadae716"},
-    {file = "storage3-2.28.0.tar.gz", hash = "sha256:bc1d008aff67de7a0f2bd867baee7aadbcdb6f78f5a310b4f7a38e8c13c19865"},
+    {file = "storage3-2.27.2-py3-none-any.whl", hash = "sha256:e6f16e7a260729e7b1f46e9bf61746805a02e30f5e419ee1291007c432e3ec63"},
+    {file = "storage3-2.27.2.tar.gz", hash = "sha256:cb4807b7f86b4bb1272ac6fdd2f3cfd8ba577297046fa5f88557425200275af5"},
 ]

 [package.dependencies]
@@ -2488,35 +2469,35 @@ python-dateutil = ">=2.6.0"

 [[package]]
 name = "supabase"
-version = "2.28.0"
+version = "2.27.2"
 description = "Supabase client for Python."
 optional = false
 python-versions = ">=3.9"
 groups = ["main"]
 files = [
-    {file = "supabase-2.28.0-py3-none-any.whl", hash = "sha256:42776971c7d0ccca16034df1ab96a31c50228eb1eb19da4249ad2f756fc20272"},
-    {file = "supabase-2.28.0.tar.gz", hash = "sha256:aea299aaab2a2eed3c57e0be7fc035c6807214194cce795a3575add20268ece1"},
+    {file = "supabase-2.27.2-py3-none-any.whl", hash = "sha256:d4dce00b3a418ee578017ec577c0e5be47a9a636355009c76f20ed2faa15bc54"},
+    {file = "supabase-2.27.2.tar.gz", hash = "sha256:2aed40e4f3454438822442a1e94a47be6694c2c70392e7ae99b51a226d4293f7"},
 ]

 [package.dependencies]
 httpx = ">=0.26,<0.29"
-postgrest = "2.28.0"
-realtime = "2.28.0"
-storage3 = "2.28.0"
-supabase-auth = "2.28.0"
-supabase-functions = "2.28.0"
+postgrest = "2.27.2"
+realtime = "2.27.2"
+storage3 = "2.27.2"
+supabase-auth = "2.27.2"
+supabase-functions = "2.27.2"
 yarl = ">=1.22.0"

 [[package]]
 name = "supabase-auth"
-version = "2.28.0"
+version = "2.27.2"
 description = "Python Client Library for Supabase Auth"
 optional = false
 python-versions = ">=3.9"
 groups = ["main"]
 files = [
-    {file = "supabase_auth-2.28.0-py3-none-any.whl", hash = "sha256:2ac85026cc285054c7fa6d41924f3a333e9ec298c013e5b5e1754039ba7caec9"},
-    {file = "supabase_auth-2.28.0.tar.gz", hash = "sha256:2bb8f18ff39934e44b28f10918db965659f3735cd6fbfcc022fe0b82dbf8233e"},
+    {file = "supabase_auth-2.27.2-py3-none-any.whl", hash = "sha256:78ec25b11314d0a9527a7205f3b1c72560dccdc11b38392f80297ef98664ee91"},
+    {file = "supabase_auth-2.27.2.tar.gz", hash = "sha256:0f5bcc79b3677cb42e9d321f3c559070cfa40d6a29a67672cc8382fb7dc2fe97"},
 ]

 [package.dependencies]
@@ -2526,14 +2507,14 @@ pyjwt = {version = ">=2.10.1", extras = ["crypto"]}

 [[package]]
 name = "supabase-functions"
-version = "2.28.0"
+version = "2.27.2"
 description = "Library for Supabase Functions"
 optional = false
 python-versions = ">=3.9"
 groups = ["main"]
 files = [
-    {file = "supabase_functions-2.28.0-py3-none-any.whl", hash = "sha256:30bf2d586f8df285faf0621bb5d5bb3ec3157234fc820553ca156f009475e4ae"},
-    {file = "supabase_functions-2.28.0.tar.gz", hash = "sha256:db3dddfc37aca5858819eb461130968473bd8c75bd284581013958526dac718b"},
+    {file = "supabase_functions-2.27.2-py3-none-any.whl", hash = "sha256:db480efc669d0bca07605b9b6f167312af43121adcc842a111f79bea416ef754"},
+    {file = "supabase_functions-2.27.2.tar.gz", hash = "sha256:d0c8266207a94371cb3fd35ad3c7f025b78a97cf026861e04ccd35ac1775f80b"},
 ]

 [package.dependencies]
@@ -2564,7 +2545,7 @@ description = "A lil' TOML parser"
 optional = false
 python-versions = ">=3.8"
 groups = ["dev"]
-markers = "python_version < \"3.11\""
+markers = "python_version == \"3.10\""
 files = [
    {file = "tomli-2.2.1-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:678e4fa69e4575eb77d103de3df8a895e1591b48e740211bd1067378c69e8249"},
    {file = "tomli-2.2.1-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:023aa114dd824ade0100497eb2318602af309e5a55595f76b626d6d9f3b7b0a6"},
@@ -2912,4 +2893,4 @@ type = ["pytest-mypy"]
 [metadata]
 lock-version = "2.1"
 python-versions = ">=3.10,<4.0"
-content-hash = "9619cae908ad38fa2c48016a58bcf4241f6f5793aa0e6cc140276e91c433cbbb"
+content-hash = "b7ac335a86aa44c3d7d2802298818b389a6f1286e3e9b7b0edb2ff06377cecaf"
--- a/autogpt_platform/autogpt_libs/pyproject.toml
+++ b/autogpt_platform/autogpt_libs/pyproject.toml
@@ -11,14 +11,14 @@ python = ">=3.10,<4.0"
 colorama = "^0.4.6"
 cryptography = "^46.0"
 expiringdict = "^1.2.2"
-fastapi = "^0.128.7"
+fastapi = "^0.128.0"
 google-cloud-logging = "^3.13.0"
-launchdarkly-server-sdk = "^9.15.0"
+launchdarkly-server-sdk = "^9.14.1"
 pydantic = "^2.12.5"
 pydantic-settings = "^2.12.0"
 pyjwt = { version = "^2.11.0", extras = ["crypto"] }
 redis = "^6.2.0"
-supabase = "^2.28.0"
+supabase = "^2.27.2"
 uvicorn = "^0.40.0"

 [tool.poetry.group.dev.dependencies]
@@ -26,7 +26,7 @@ pyright = "^1.1.408"
 pytest = "^8.4.1"
 pytest-asyncio = "^1.3.0"
 pytest-mock = "^3.15.1"
-pytest-cov = "^7.0.0"
+pytest-cov = "^6.2.1"
 ruff = "^0.15.0"

 [build-system]
--- a/autogpt_platform/backend/.env.default
+++ b/autogpt_platform/backend/.env.default
@@ -104,12 +104,6 @@ TWITTER_CLIENT_SECRET=
 # Make a new workspace for your OAuth APP -- trust me
 # https://linear.app/settings/api/applications/new
 # Callback URL: http://localhost:3000/auth/integrations/oauth_callback
-LINEAR_API_KEY=
-# Linear project and team IDs for the feature request tracker.
-# Find these in your Linear workspace URL: linear.app/<workspace>/project/<project-id>
-# and in team settings. Used by the chat copilot to file and search feature requests.
-LINEAR_FEATURE_REQUEST_PROJECT_ID=
-LINEAR_FEATURE_REQUEST_TEAM_ID=
 LINEAR_CLIENT_ID=
 LINEAR_CLIENT_SECRET=

--- a/autogpt_platform/backend/Dockerfile
+++ b/autogpt_platform/backend/Dockerfile
@@ -1,5 +1,3 @@
-# ============================ DEPENDENCY BUILDER ============================ #
-
 FROM debian:13-slim AS builder

 # Set environment variables
@@ -53,62 +51,27 @@ COPY autogpt_platform/backend/backend/data/partial_types.py ./backend/data/parti
 COPY autogpt_platform/backend/gen_prisma_types_stub.py ./
 RUN poetry run prisma generate && poetry run gen-prisma-stub

-# =============================== DB MIGRATOR =============================== #
-
-# Lightweight migrate stage - only needs Prisma CLI, not full Python environment
-FROM debian:13-slim AS migrate
-
-WORKDIR /app/autogpt_platform/backend
-
-ENV DEBIAN_FRONTEND=noninteractive
-
-# Install only what's needed for prisma migrate: Node.js and minimal Python for prisma-python
-RUN apt-get update && apt-get install -y --no-install-recommends \
-    python3.13 \
-    python3-pip \
-    ca-certificates \
-    && rm -rf /var/lib/apt/lists/*
-
-# Copy Node.js from builder (needed for Prisma CLI)
-COPY --from=builder /usr/bin/node /usr/bin/node
-COPY --from=builder /usr/lib/node_modules /usr/lib/node_modules
-COPY --from=builder /usr/bin/npm /usr/bin/npm
-
-# Copy Prisma binaries
-COPY --from=builder /root/.cache/prisma-python/binaries /root/.cache/prisma-python/binaries
-
-# Install prisma-client-py directly (much smaller than copying full venv)
-RUN pip3 install prisma>=0.15.0 --break-system-packages
-
-COPY autogpt_platform/backend/schema.prisma ./
-COPY autogpt_platform/backend/backend/data/partial_types.py ./backend/data/partial_types.py
-COPY autogpt_platform/backend/gen_prisma_types_stub.py ./
-COPY autogpt_platform/backend/migrations ./migrations
-
-# ============================== BACKEND SERVER ============================== #
-
-FROM debian:13-slim AS server
+FROM debian:13-slim AS server_dependencies

 WORKDIR /app

-ENV DEBIAN_FRONTEND=noninteractive
+ENV POETRY_HOME=/opt/poetry \
+    POETRY_NO_INTERACTION=1 \
+    POETRY_VIRTUALENVS_CREATE=true \
+    POETRY_VIRTUALENVS_IN_PROJECT=true \
+    DEBIAN_FRONTEND=noninteractive
+ENV PATH=/opt/poetry/bin:$PATH

-# Install Python, FFmpeg, ImageMagick, and CLI tools for agent use.
-# bubblewrap provides OS-level sandbox (whitelist-only FS + no network)
-# for the bash_exec MCP tool.
-# Using --no-install-recommends saves ~650MB by skipping unnecessary deps like llvm, mesa, etc.
-RUN apt-get update && apt-get install -y --no-install-recommends \
+# Install Python, FFmpeg, and ImageMagick (required for video processing blocks)
+RUN apt-get update && apt-get install -y \
    python3.13 \
    python3-pip \
    ffmpeg \
    imagemagick \
-    jq \
-    ripgrep \
-    tree \
-    bubblewrap \
    && rm -rf /var/lib/apt/lists/*

-# Copy poetry (build-time only, for `poetry install --only-root` to create entry points)
+# Copy only necessary files from builder
+COPY --from=builder /app /app
 COPY --from=builder /usr/local/lib/python3* /usr/local/lib/python3*
 COPY --from=builder /usr/local/bin/poetry /usr/local/bin/poetry
 # Copy Node.js installation for Prisma
@@ -118,25 +81,30 @@ COPY --from=builder /usr/bin/npm /usr/bin/npm
 COPY --from=builder /usr/bin/npx /usr/bin/npx
 COPY --from=builder /root/.cache/prisma-python/binaries /root/.cache/prisma-python/binaries

-WORKDIR /app/autogpt_platform/backend
-
-# Copy only the .venv from builder (not the entire /app directory)
-# The .venv includes the generated Prisma client
-COPY --from=builder /app/autogpt_platform/backend/.venv ./.venv
 ENV PATH="/app/autogpt_platform/backend/.venv/bin:$PATH"

-# Copy dependency files + autogpt_libs (path dependency)
-COPY autogpt_platform/autogpt_libs /app/autogpt_platform/autogpt_libs
-COPY autogpt_platform/backend/poetry.lock autogpt_platform/backend/pyproject.toml ./
+RUN mkdir -p /app/autogpt_platform/autogpt_libs
+RUN mkdir -p /app/autogpt_platform/backend

-# Copy backend code + docs (for Copilot docs search)
-COPY autogpt_platform/backend ./
+COPY autogpt_platform/autogpt_libs /app/autogpt_platform/autogpt_libs
+
+COPY autogpt_platform/backend/poetry.lock autogpt_platform/backend/pyproject.toml /app/autogpt_platform/backend/
+
+WORKDIR /app/autogpt_platform/backend
+
+FROM server_dependencies AS migrate
+
+# Migration stage only needs schema and migrations - much lighter than full backend
+COPY autogpt_platform/backend/schema.prisma /app/autogpt_platform/backend/
+COPY autogpt_platform/backend/backend/data/partial_types.py /app/autogpt_platform/backend/backend/data/partial_types.py
+COPY autogpt_platform/backend/migrations /app/autogpt_platform/backend/migrations
+
+FROM server_dependencies AS server
+
+COPY autogpt_platform/backend /app/autogpt_platform/backend
 COPY docs /app/docs
-# Install the project package to create entry point scripts in .venv/bin/
-# (e.g., rest, executor, ws, db, scheduler, notification - see [tool.poetry.scripts])
-RUN POETRY_VIRTUALENVS_CREATE=true POETRY_VIRTUALENVS_IN_PROJECT=true \
-    poetry install --no-ansi --only-root
+RUN poetry install --no-ansi --only-root

 ENV PORT=8000

-CMD ["rest"]
+CMD ["poetry", "run", "rest"]
--- a/autogpt_platform/backend/backend/api/conftest.py
+++ b/autogpt_platform/backend/backend/api/conftest.py
@@ -1,9 +1,4 @@
-"""Common test fixtures for server tests.
-
-Note: Common fixtures like test_user_id, admin_user_id, target_user_id,
-setup_test_user, and setup_admin_user are defined in the parent conftest.py
-(backend/conftest.py) and are available here automatically.
-"""
+"""Common test fixtures for server tests."""

 import pytest
 from pytest_snapshot.plugin import Snapshot
@@ -16,6 +11,54 @@ def configured_snapshot(snapshot: Snapshot) -> Snapshot:
    return snapshot


+@pytest.fixture
+def test_user_id() -> str:
+    """Test user ID fixture."""
+    return "3e53486c-cf57-477e-ba2a-cb02dc828e1a"
+
+
+@pytest.fixture
+def admin_user_id() -> str:
+    """Admin user ID fixture."""
+    return "4e53486c-cf57-477e-ba2a-cb02dc828e1b"
+
+
+@pytest.fixture
+def target_user_id() -> str:
+    """Target user ID fixture."""
+    return "5e53486c-cf57-477e-ba2a-cb02dc828e1c"
+
+
+@pytest.fixture
+async def setup_test_user(test_user_id):
+    """Create test user in database before tests."""
+    from backend.data.user import get_or_create_user
+
+    # Create the test user in the database using JWT token format
+    user_data = {
+        "sub": test_user_id,
+        "email": "test@example.com",
+        "user_metadata": {"name": "Test User"},
+    }
+    await get_or_create_user(user_data)
+    return test_user_id
+
+
+@pytest.fixture
+async def setup_admin_user(admin_user_id):
+    """Create admin user in database before tests."""
+    from backend.data.user import get_or_create_user
+
+    # Create the admin user in the database using JWT token format
+    user_data = {
+        "sub": admin_user_id,
+        "email": "test-admin@example.com",
+        "user_metadata": {"name": "Test Admin"},
+    }
+    await get_or_create_user(user_data)
+    return admin_user_id
+
+
@pytest.fixture
 def mock_jwt_user(test_user_id):
    """Provide mock JWT payload for regular user testing."""
--- a/autogpt_platform/backend/backend/api/external/v1/routes.py
+++ b/autogpt_platform/backend/backend/api/external/v1/routes.py
@@ -10,7 +10,7 @@ from typing_extensions import TypedDict

 import backend.api.features.store.cache as store_cache
 import backend.api.features.store.model as store_model
-import backend.blocks
+import backend.data.block
 from backend.api.external.middleware import require_permission
 from backend.data import execution as execution_db
 from backend.data import graph as graph_db
@@ -67,7 +67,7 @@ async def get_user_info(
    dependencies=[Security(require_permission(APIKeyPermission.READ_BLOCK))],
 )
 async def get_graph_blocks() -> Sequence[dict[Any, Any]]:
-    blocks = [block() for block in backend.blocks.get_blocks().values()]
+    blocks = [block() for block in backend.data.block.get_blocks().values()]
    return [b.to_dict() for b in blocks if not b.disabled]


@@ -83,7 +83,7 @@ async def execute_graph_block(
        require_permission(APIKeyPermission.EXECUTE_BLOCK)
    ),
 ) -> CompletedBlockOutput:
-    obj = backend.blocks.get_block(block_id)
+    obj = backend.data.block.get_block(block_id)
    if not obj:
        raise HTTPException(status_code=404, detail=f"Block #{block_id} not found.")
    if obj.disabled:
--- a/autogpt_platform/backend/backend/api/external/v1/tools.py
+++ b/autogpt_platform/backend/backend/api/external/v1/tools.py
@@ -15,9 +15,9 @@ from prisma.enums import APIKeyPermission
 from pydantic import BaseModel, Field

 from backend.api.external.middleware import require_permission
-from backend.copilot.model import ChatSession
-from backend.copilot.tools import find_agent_tool, run_agent_tool
-from backend.copilot.tools.models import ToolResponseBase
+from backend.api.features.chat.model import ChatSession
+from backend.api.features.chat.tools import find_agent_tool, run_agent_tool
+from backend.api.features.chat.tools.models import ToolResponseBase
 from backend.data.auth.base import APIAuthorizationInfo

 logger = logging.getLogger(__name__)
--- a/autogpt_platform/backend/backend/api/features/builder/db.py
+++ b/autogpt_platform/backend/backend/api/features/builder/db.py
@@ -10,15 +10,10 @@ import backend.api.features.library.db as library_db
 import backend.api.features.library.model as library_model
 import backend.api.features.store.db as store_db
 import backend.api.features.store.model as store_model
+import backend.data.block
 from backend.blocks import load_all_blocks
-from backend.blocks._base import (
-    AnyBlockSchema,
-    BlockCategory,
-    BlockInfo,
-    BlockSchema,
-    BlockType,
-)
 from backend.blocks.llm import LlmModel
+from backend.data.block import AnyBlockSchema, BlockCategory, BlockInfo, BlockSchema
 from backend.data.db import query_raw_with_schema
 from backend.integrations.providers import ProviderName
 from backend.util.cache import cached
@@ -27,7 +22,7 @@ from backend.util.models import Pagination
 from .model import (
    BlockCategoryResponse,
    BlockResponse,
-    BlockTypeFilter,
+    BlockType,
    CountResponse,
    FilterType,
    Provider,
@@ -93,7 +88,7 @@ def get_block_categories(category_blocks: int = 3) -> list[BlockCategoryResponse
 def get_blocks(
    *,
    category: str | None = None,
-    type: BlockTypeFilter | None = None,
+    type: BlockType | None = None,
    provider: ProviderName | None = None,
    page: int = 1,
    page_size: int = 50,
@@ -674,9 +669,9 @@ async def get_suggested_blocks(count: int = 5) -> list[BlockInfo]:
    for block_type in load_all_blocks().values():
        block: AnyBlockSchema = block_type()
        if block.disabled or block.block_type in (
-            BlockType.INPUT,
-            BlockType.OUTPUT,
-            BlockType.AGENT,
+            backend.data.block.BlockType.INPUT,
+            backend.data.block.BlockType.OUTPUT,
+            backend.data.block.BlockType.AGENT,
        ):
            continue
        # Find the execution count for this block
--- a/autogpt_platform/backend/backend/api/features/builder/model.py
+++ b/autogpt_platform/backend/backend/api/features/builder/model.py
@@ -4,7 +4,7 @@ from pydantic import BaseModel

 import backend.api.features.library.model as library_model
 import backend.api.features.store.model as store_model
-from backend.blocks._base import BlockInfo
+from backend.data.block import BlockInfo
 from backend.integrations.providers import ProviderName
 from backend.util.models import Pagination

@@ -15,7 +15,7 @@ FilterType = Literal[
    "my_agents",
 ]

-BlockTypeFilter = Literal["all", "input", "action", "output"]
+BlockType = Literal["all", "input", "action", "output"]


 class SearchEntry(BaseModel):
--- a/autogpt_platform/backend/backend/api/features/builder/routes.py
+++ b/autogpt_platform/backend/backend/api/features/builder/routes.py
@@ -88,7 +88,7 @@ async def get_block_categories(
 )
 async def get_blocks(
    category: Annotated[str | None, fastapi.Query()] = None,
-    type: Annotated[builder_model.BlockTypeFilter | None, fastapi.Query()] = None,
+    type: Annotated[builder_model.BlockType | None, fastapi.Query()] = None,
    provider: Annotated[ProviderName | None, fastapi.Query()] = None,
    page: Annotated[int, fastapi.Query()] = 1,
    page_size: Annotated[int, fastapi.Query()] = 50,
--- a/autogpt_platform/backend/backend/api/features/chat/completion_consumer.py
+++ b/autogpt_platform/backend/backend/api/features/chat/completion_consumer.py
@@ -37,10 +37,12 @@ stale pending messages from dead consumers.

 import asyncio
 import logging
+import os
 import uuid
 from typing import Any

 import orjson
+from prisma import Prisma
 from pydantic import BaseModel
 from redis.exceptions import ResponseError

@@ -67,8 +69,8 @@ class OperationCompleteMessage(BaseModel):
 class ChatCompletionConsumer:
    """Consumer for chat operation completion messages from Redis Streams.

-    Database operations are handled through the chat_db() accessor, which
-    routes through DatabaseManager RPC when Prisma is not directly connected.
+    This consumer initializes its own Prisma client in start() to ensure
+    database operations work correctly within this async context.

    Uses Redis consumer groups to allow multiple platform pods to consume
    messages reliably with automatic redelivery on failure.
@@ -77,6 +79,7 @@ class ChatCompletionConsumer:
    def __init__(self):
        self._consumer_task: asyncio.Task | None = None
        self._running = False
+        self._prisma: Prisma | None = None
        self._consumer_name = f"consumer-{uuid.uuid4().hex[:8]}"

    async def start(self) -> None:
@@ -112,6 +115,15 @@ class ChatCompletionConsumer:
            f"Chat completion consumer started (consumer: {self._consumer_name})"
        )

+    async def _ensure_prisma(self) -> Prisma:
+        """Lazily initialize Prisma client on first use."""
+        if self._prisma is None:
+            database_url = os.getenv("DATABASE_URL", "postgresql://localhost:5432")
+            self._prisma = Prisma(datasource={"url": database_url})
+            await self._prisma.connect()
+            logger.info("[COMPLETION] Consumer Prisma client connected (lazy init)")
+        return self._prisma
+
    async def stop(self) -> None:
        """Stop the completion consumer."""
        self._running = False
@@ -124,6 +136,11 @@ class ChatCompletionConsumer:
                pass
            self._consumer_task = None

+        if self._prisma:
+            await self._prisma.disconnect()
+            self._prisma = None
+            logger.info("[COMPLETION] Consumer Prisma client disconnected")
+
        logger.info("Chat completion consumer stopped")

    async def _consume_messages(self) -> None:
@@ -235,7 +252,7 @@ class ChatCompletionConsumer:
            # XAUTOCLAIM after min_idle_time expires

    async def _handle_message(self, body: bytes) -> None:
-        """Handle a completion message."""
+        """Handle a completion message using our own Prisma client."""
        try:
            data = orjson.loads(body)
            message = OperationCompleteMessage(**data)
@@ -285,7 +302,8 @@ class ChatCompletionConsumer:
        message: OperationCompleteMessage,
    ) -> None:
        """Handle successful operation completion."""
-        await process_operation_success(task, message.result)
+        prisma = await self._ensure_prisma()
+        await process_operation_success(task, message.result, prisma)

    async def _handle_failure(
        self,
@@ -293,7 +311,8 @@ class ChatCompletionConsumer:
        message: OperationCompleteMessage,
    ) -> None:
        """Handle failed operation completion."""
-        await process_operation_failure(task, message.error)
+        prisma = await self._ensure_prisma()
+        await process_operation_failure(task, message.error, prisma)


 # Module-level consumer instance
--- a/autogpt_platform/backend/backend/api/features/chat/completion_handler.py
+++ b/autogpt_platform/backend/backend/api/features/chat/completion_handler.py
@@ -9,8 +9,7 @@ import logging
 from typing import Any

 import orjson
-
-from backend.data.db_accessors import chat_db
+from prisma import Prisma

 from . import service as chat_service
 from . import stream_registry
@@ -73,40 +72,48 @@ async def _update_tool_message(
    session_id: str,
    tool_call_id: str,
    content: str,
+    prisma_client: Prisma | None,
 ) -> None:
-    """Update tool message in database using the chat_db accessor.
-
-    Routes through DatabaseManager RPC when Prisma is not directly
-    connected (e.g. in the CoPilot Executor microservice).
+    """Update tool message in database.

    Args:
        session_id: The session ID
        tool_call_id: The tool call ID to update
        content: The new content for the message
+        prisma_client: Optional Prisma client. If None, uses chat_service.

    Raises:
-        ToolMessageUpdateError: If the database update fails.
+        ToolMessageUpdateError: If the database update fails. The caller should
+            handle this to avoid marking the task as completed with inconsistent state.
    """
    try:
-        updated = await chat_db().update_tool_message_content(
-            session_id=session_id,
-            tool_call_id=tool_call_id,
-            new_content=content,
-        )
-        if not updated:
-            raise ToolMessageUpdateError(
-                f"No message found with tool_call_id="
-                f"{tool_call_id} in session {session_id}"
+        if prisma_client:
+            # Use provided Prisma client (for consumer with its own connection)
+            updated_count = await prisma_client.chatmessage.update_many(
+                where={
+                    "sessionId": session_id,
+                    "toolCallId": tool_call_id,
+                },
+                data={"content": content},
+            )
+            # Check if any rows were updated - 0 means message not found
+            if updated_count == 0:
+                raise ToolMessageUpdateError(
+                    f"No message found with tool_call_id={tool_call_id} in session {session_id}"
+                )
+        else:
+            # Use service function (for webhook endpoint)
+            await chat_service._update_pending_operation(
+                session_id=session_id,
+                tool_call_id=tool_call_id,
+                result=content,
            )
    except ToolMessageUpdateError:
        raise
    except Exception as e:
-        logger.error(
-            f"[COMPLETION] Failed to update tool message: {e}",
-            exc_info=True,
-        )
+        logger.error(f"[COMPLETION] Failed to update tool message: {e}", exc_info=True)
        raise ToolMessageUpdateError(
-            f"Failed to update tool message for tool call #{tool_call_id}: {e}"
+            f"Failed to update tool message for tool_call_id={tool_call_id}: {e}"
        ) from e


@@ -195,6 +202,7 @@ async def _save_agent_from_result(
 async def process_operation_success(
    task: stream_registry.ActiveTask,
    result: dict | str | None,
+    prisma_client: Prisma | None = None,
 ) -> None:
    """Handle successful operation completion.

@@ -204,10 +212,12 @@ async def process_operation_success(
    Args:
        task: The active task that completed
        result: The result data from the operation
+        prisma_client: Optional Prisma client for database operations.
+            If None, uses chat_service._update_pending_operation instead.

    Raises:
-        ToolMessageUpdateError: If the database update fails. The task
-            will be marked as failed instead of completed.
+        ToolMessageUpdateError: If the database update fails. The task will be
+            marked as failed instead of completed to avoid inconsistent state.
    """
    # For agent generation tools, save the agent to library
    if task.tool_name in AGENT_GENERATION_TOOLS and isinstance(result, dict):
@@ -240,6 +250,7 @@ async def process_operation_success(
            session_id=task.session_id,
            tool_call_id=task.tool_call_id,
            content=result_str,
+            prisma_client=prisma_client,
        )
    except ToolMessageUpdateError:
        # DB update failed - mark task as failed to avoid inconsistent state
@@ -282,15 +293,18 @@ async def process_operation_success(
 async def process_operation_failure(
    task: stream_registry.ActiveTask,
    error: str | None,
+    prisma_client: Prisma | None = None,
 ) -> None:
    """Handle failed operation completion.

-    Publishes the error to the stream registry, updates the database
-    with the error response, and marks the task as failed.
+    Publishes the error to the stream registry, updates the database with
+    the error response, and marks the task as failed.

    Args:
        task: The active task that failed
        error: The error message from the operation
+        prisma_client: Optional Prisma client for database operations.
+            If None, uses chat_service._update_pending_operation instead.
    """
    error_msg = error or "Operation failed"

@@ -311,6 +325,7 @@ async def process_operation_failure(
            session_id=task.session_id,
            tool_call_id=task.tool_call_id,
            content=error_response.model_dump_json(),
+            prisma_client=prisma_client,
        )
    except ToolMessageUpdateError:
        # DB update failed - log but continue with cleanup
--- a/autogpt_platform/backend/backend/api/features/chat/config.py
+++ b/autogpt_platform/backend/backend/api/features/chat/config.py
@@ -27,11 +27,12 @@ class ChatConfig(BaseSettings):
    session_ttl: int = Field(default=43200, description="Session TTL in seconds")

    # Streaming Configuration
-    stream_timeout: int = Field(default=300, description="Stream timeout in seconds")
-    max_retries: int = Field(
-        default=3,
-        description="Max retries for fallback path (SDK handles retries internally)",
+    max_context_messages: int = Field(
+        default=50, ge=1, le=200, description="Maximum context messages"
    )
+
+    stream_timeout: int = Field(default=300, description="Stream timeout in seconds")
+    max_retries: int = Field(default=3, description="Maximum number of retries")
    max_agent_runs: int = Field(default=30, description="Maximum number of agent runs")
    max_agent_schedules: int = Field(
        default=30, description="Maximum number of agent schedules"
@@ -92,37 +93,6 @@ class ChatConfig(BaseSettings):
        description="Name of the prompt in Langfuse to fetch",
    )

-    # Claude Agent SDK Configuration
-    use_claude_agent_sdk: bool = Field(
-        default=True,
-        description="Use Claude Agent SDK for chat completions",
-    )
-    claude_agent_model: str | None = Field(
-        default=None,
-        description="Model for the Claude Agent SDK path. If None, derives from "
-        "the `model` field by stripping the OpenRouter provider prefix.",
-    )
-    claude_agent_max_buffer_size: int = Field(
-        default=10 * 1024 * 1024,  # 10MB (default SDK is 1MB)
-        description="Max buffer size in bytes for Claude Agent SDK JSON message parsing. "
-        "Increase if tool outputs exceed the limit.",
-    )
-    claude_agent_max_subtasks: int = Field(
-        default=10,
-        description="Max number of sub-agent Tasks the SDK can spawn per session.",
-    )
-    claude_agent_use_resume: bool = Field(
-        default=True,
-        description="Use --resume for multi-turn conversations instead of "
-        "history compression. Falls back to compression when unavailable.",
-    )
-
-    # Extended thinking configuration for Claude models
-    thinking_enabled: bool = Field(
-        default=True,
-        description="Enable adaptive thinking for Claude models via OpenRouter",
-    )
-
    @field_validator("api_key", mode="before")
    @classmethod
    def get_api_key(cls, v):
@@ -162,17 +132,6 @@ class ChatConfig(BaseSettings):
            v = os.getenv("CHAT_INTERNAL_API_KEY")
        return v

-    @field_validator("use_claude_agent_sdk", mode="before")
-    @classmethod
-    def get_use_claude_agent_sdk(cls, v):
-        """Get use_claude_agent_sdk from environment if not provided."""
-        # Check environment variable - default to True if not set
-        env_val = os.getenv("CHAT_USE_CLAUDE_AGENT_SDK", "").lower()
-        if env_val:
-            return env_val in ("true", "1", "yes", "on")
-        # Default to True (SDK enabled by default)
-        return True if v is None else v
-
    # Prompt paths for different contexts
    PROMPT_PATHS: dict[str, str] = {
        "default": "prompts/chat_system.md",
--- a/autogpt_platform/backend/backend/api/features/chat/db.py
+++ b/autogpt_platform/backend/backend/api/features/chat/db.py
@@ -14,27 +14,29 @@ from prisma.types import (
    ChatSessionWhereInput,
 )

-from backend.data import db
+from backend.data.db import transaction
 from backend.util.json import SafeJson

-from .model import ChatMessage, ChatSession, ChatSessionInfo
-
 logger = logging.getLogger(__name__)


-async def get_chat_session(session_id: str) -> ChatSession | None:
+async def get_chat_session(session_id: str) -> PrismaChatSession | None:
    """Get a chat session by ID from the database."""
    session = await PrismaChatSession.prisma().find_unique(
        where={"id": session_id},
-        include={"Messages": {"order_by": {"sequence": "asc"}}},
+        include={"Messages": True},
    )
-    return ChatSession.from_db(session) if session else None
+    if session and session.Messages:
+        # Sort messages by sequence in Python - Prisma Python client doesn't support
+        # order_by in include clauses (unlike Prisma JS), so we sort after fetching
+        session.Messages.sort(key=lambda m: m.sequence)
+    return session


 async def create_chat_session(
    session_id: str,
    user_id: str,
-) -> ChatSessionInfo:
+) -> PrismaChatSession:
    """Create a new chat session in the database."""
    data = ChatSessionCreateInput(
        id=session_id,
@@ -43,8 +45,10 @@ async def create_chat_session(
        successfulAgentRuns=SafeJson({}),
        successfulAgentSchedules=SafeJson({}),
    )
-    prisma_session = await PrismaChatSession.prisma().create(data=data)
-    return ChatSessionInfo.from_db(prisma_session)
+    return await PrismaChatSession.prisma().create(
+        data=data,
+        include={"Messages": True},
+    )


 async def update_chat_session(
@@ -55,7 +59,7 @@ async def update_chat_session(
    total_prompt_tokens: int | None = None,
    total_completion_tokens: int | None = None,
    title: str | None = None,
-) -> ChatSession | None:
+) -> PrismaChatSession | None:
    """Update a chat session's metadata."""
    data: ChatSessionUpdateInput = {"updatedAt": datetime.now(UTC)}

@@ -75,9 +79,12 @@ async def update_chat_session(
    session = await PrismaChatSession.prisma().update(
        where={"id": session_id},
        data=data,
-        include={"Messages": {"order_by": {"sequence": "asc"}}},
+        include={"Messages": True},
    )
-    return ChatSession.from_db(session) if session else None
+    if session and session.Messages:
+        # Sort in Python - Prisma Python doesn't support order_by in include clauses
+        session.Messages.sort(key=lambda m: m.sequence)
+    return session


 async def add_chat_message(
@@ -90,7 +97,7 @@ async def add_chat_message(
    refusal: str | None = None,
    tool_calls: list[dict[str, Any]] | None = None,
    function_call: dict[str, Any] | None = None,
-) -> ChatMessage:
+) -> PrismaChatMessage:
    """Add a message to a chat session."""
    # Build input dict dynamically rather than using ChatMessageCreateInput directly
    # because Prisma's TypedDict validation rejects optional fields set to None.
@@ -125,14 +132,14 @@ async def add_chat_message(
        ),
        PrismaChatMessage.prisma().create(data=cast(ChatMessageCreateInput, data)),
    )
-    return ChatMessage.from_db(message)
+    return message


 async def add_chat_messages_batch(
    session_id: str,
    messages: list[dict[str, Any]],
    start_sequence: int,
-) -> list[ChatMessage]:
+) -> list[PrismaChatMessage]:
    """Add multiple messages to a chat session in a batch.

    Uses a transaction for atomicity - if any message creation fails,
@@ -143,7 +150,7 @@ async def add_chat_messages_batch(

    created_messages = []

-    async with db.transaction() as tx:
+    async with transaction() as tx:
        for i, msg in enumerate(messages):
            # Build input dict dynamically rather than using ChatMessageCreateInput
            # directly because Prisma's TypedDict validation rejects optional fields
@@ -183,22 +190,21 @@ async def add_chat_messages_batch(
            data={"updatedAt": datetime.now(UTC)},
        )

-    return [ChatMessage.from_db(m) for m in created_messages]
+    return created_messages


 async def get_user_chat_sessions(
    user_id: str,
    limit: int = 50,
    offset: int = 0,
-) -> list[ChatSessionInfo]:
+) -> list[PrismaChatSession]:
    """Get chat sessions for a user, ordered by most recent."""
-    prisma_sessions = await PrismaChatSession.prisma().find_many(
+    return await PrismaChatSession.prisma().find_many(
        where={"userId": user_id},
        order={"updatedAt": "desc"},
        take=limit,
        skip=offset,
    )
-    return [ChatSessionInfo.from_db(s) for s in prisma_sessions]


 async def get_user_session_count(user_id: str) -> int:
--- a/autogpt_platform/backend/backend/api/features/chat/model.py
+++ b/autogpt_platform/backend/backend/api/features/chat/model.py
@@ -2,7 +2,7 @@ import asyncio
 import logging
 import uuid
 from datetime import UTC, datetime
-from typing import Any, Self, cast
+from typing import Any
 from weakref import WeakValueDictionary

 from openai.types.chat import (
@@ -23,17 +23,26 @@ from prisma.models import ChatMessage as PrismaChatMessage
 from prisma.models import ChatSession as PrismaChatSession
 from pydantic import BaseModel

-from backend.data.db_accessors import chat_db
 from backend.data.redis_client import get_redis_async
 from backend.util import json
 from backend.util.exceptions import DatabaseError, RedisError

+from . import db as chat_db
 from .config import ChatConfig

 logger = logging.getLogger(__name__)
 config = ChatConfig()


+def _parse_json_field(value: str | dict | list | None, default: Any = None) -> Any:
+    """Parse a JSON field that may be stored as string or already parsed."""
+    if value is None:
+        return default
+    if isinstance(value, str):
+        return json.loads(value)
+    return value
+
+
 # Redis cache key prefix for chat sessions
 CHAT_SESSION_CACHE_PREFIX = "chat:session:"

@@ -43,7 +52,28 @@ def _get_session_cache_key(session_id: str) -> str:
    return f"{CHAT_SESSION_CACHE_PREFIX}{session_id}"


-# ===================== Chat data models ===================== #
+# Session-level locks to prevent race conditions during concurrent upserts.
+# Uses WeakValueDictionary to automatically garbage collect locks when no longer referenced,
+# preventing unbounded memory growth while maintaining lock semantics for active sessions.
+# Invalidation: Locks are auto-removed by GC when no coroutine holds a reference (after
+# async with lock: completes). Explicit cleanup also occurs in delete_chat_session().
+_session_locks: WeakValueDictionary[str, asyncio.Lock] = WeakValueDictionary()
+_session_locks_mutex = asyncio.Lock()
+
+
+async def _get_session_lock(session_id: str) -> asyncio.Lock:
+    """Get or create a lock for a specific session to prevent concurrent upserts.
+
+    Uses WeakValueDictionary for automatic cleanup: locks are garbage collected
+    when no coroutine holds a reference to them, preventing memory leaks from
+    unbounded growth of session locks.
+    """
+    async with _session_locks_mutex:
+        lock = _session_locks.get(session_id)
+        if lock is None:
+            lock = asyncio.Lock()
+            _session_locks[session_id] = lock
+        return lock


 class ChatMessage(BaseModel):
@@ -55,19 +85,6 @@ class ChatMessage(BaseModel):
    tool_calls: list[dict] | None = None
    function_call: dict | None = None

-    @staticmethod
-    def from_db(prisma_message: PrismaChatMessage) -> "ChatMessage":
-        """Convert a Prisma ChatMessage to a Pydantic ChatMessage."""
-        return ChatMessage(
-            role=prisma_message.role,
-            content=prisma_message.content,
-            name=prisma_message.name,
-            tool_call_id=prisma_message.toolCallId,
-            refusal=prisma_message.refusal,
-            tool_calls=_parse_json_field(prisma_message.toolCalls),
-            function_call=_parse_json_field(prisma_message.functionCall),
-        )
-

 class Usage(BaseModel):
    prompt_tokens: int
@@ -75,10 +92,11 @@ class Usage(BaseModel):
    total_tokens: int


-class ChatSessionInfo(BaseModel):
+class ChatSession(BaseModel):
    session_id: str
    user_id: str
    title: str | None = None
+    messages: list[ChatMessage]
    usage: list[Usage]
    credentials: dict[str, dict] = {}  # Map of provider -> credential metadata
    started_at: datetime
@@ -86,9 +104,40 @@ class ChatSessionInfo(BaseModel):
    successful_agent_runs: dict[str, int] = {}
    successful_agent_schedules: dict[str, int] = {}

-    @classmethod
-    def from_db(cls, prisma_session: PrismaChatSession) -> Self:
-        """Convert Prisma ChatSession to Pydantic ChatSession."""
+    @staticmethod
+    def new(user_id: str) -> "ChatSession":
+        return ChatSession(
+            session_id=str(uuid.uuid4()),
+            user_id=user_id,
+            title=None,
+            messages=[],
+            usage=[],
+            credentials={},
+            started_at=datetime.now(UTC),
+            updated_at=datetime.now(UTC),
+        )
+
+    @staticmethod
+    def from_db(
+        prisma_session: PrismaChatSession,
+        prisma_messages: list[PrismaChatMessage] | None = None,
+    ) -> "ChatSession":
+        """Convert Prisma models to Pydantic ChatSession."""
+        messages = []
+        if prisma_messages:
+            for msg in prisma_messages:
+                messages.append(
+                    ChatMessage(
+                        role=msg.role,
+                        content=msg.content,
+                        name=msg.name,
+                        tool_call_id=msg.toolCallId,
+                        refusal=msg.refusal,
+                        tool_calls=_parse_json_field(msg.toolCalls),
+                        function_call=_parse_json_field(msg.functionCall),
+                    )
+                )
+
        # Parse JSON fields from Prisma
        credentials = _parse_json_field(prisma_session.credentials, default={})
        successful_agent_runs = _parse_json_field(
@@ -110,10 +159,11 @@ class ChatSessionInfo(BaseModel):
                )
            )

-        return cls(
+        return ChatSession(
            session_id=prisma_session.id,
            user_id=prisma_session.userId,
            title=prisma_session.title,
+            messages=messages,
            usage=usage,
            credentials=credentials,
            started_at=prisma_session.createdAt,
@@ -122,56 +172,6 @@ class ChatSessionInfo(BaseModel):
            successful_agent_schedules=successful_agent_schedules,
        )

-
-class ChatSession(ChatSessionInfo):
-    messages: list[ChatMessage]
-
-    @classmethod
-    def new(cls, user_id: str) -> Self:
-        return cls(
-            session_id=str(uuid.uuid4()),
-            user_id=user_id,
-            title=None,
-            messages=[],
-            usage=[],
-            credentials={},
-            started_at=datetime.now(UTC),
-            updated_at=datetime.now(UTC),
-        )
-
-    @classmethod
-    def from_db(cls, prisma_session: PrismaChatSession) -> Self:
-        """Convert Prisma ChatSession to Pydantic ChatSession."""
-        if prisma_session.Messages is None:
-            raise ValueError(
-                f"Prisma session {prisma_session.id} is missing Messages relation"
-            )
-
-        return cls(
-            **ChatSessionInfo.from_db(prisma_session).model_dump(),
-            messages=[ChatMessage.from_db(m) for m in prisma_session.Messages],
-        )
-
-    def add_tool_call_to_current_turn(self, tool_call: dict) -> None:
-        """Attach a tool_call to the current turn's assistant message.
-
-        Searches backwards for the most recent assistant message (stopping at
-        any user message boundary). If found, appends the tool_call to it.
-        Otherwise creates a new assistant message with the tool_call.
-        """
-        for msg in reversed(self.messages):
-            if msg.role == "user":
-                break
-            if msg.role == "assistant":
-                if not msg.tool_calls:
-                    msg.tool_calls = []
-                msg.tool_calls.append(tool_call)
-                return
-
-        self.messages.append(
-            ChatMessage(role="assistant", content="", tool_calls=[tool_call])
-        )
-
    def to_openai_messages(self) -> list[ChatCompletionMessageParam]:
        messages = []
        for message in self.messages:
@@ -258,72 +258,43 @@ class ChatSession(ChatSessionInfo):
                        name=message.name or "",
                    )
                )
-        return self._merge_consecutive_assistant_messages(messages)
-
-    @staticmethod
-    def _merge_consecutive_assistant_messages(
-        messages: list[ChatCompletionMessageParam],
-    ) -> list[ChatCompletionMessageParam]:
-        """Merge consecutive assistant messages into single messages.
-
-        Long-running tool flows can create split assistant messages: one with
-        text content and another with tool_calls. Anthropic's API requires
-        tool_result blocks to reference a tool_use in the immediately preceding
-        assistant message, so these splits cause 400 errors via OpenRouter.
-        """
-        if len(messages) < 2:
-            return messages
-
-        result: list[ChatCompletionMessageParam] = [messages[0]]
-        for msg in messages[1:]:
-            prev = result[-1]
-            if prev.get("role") != "assistant" or msg.get("role") != "assistant":
-                result.append(msg)
-                continue
-
-            prev = cast(ChatCompletionAssistantMessageParam, prev)
-            curr = cast(ChatCompletionAssistantMessageParam, msg)
-
-            curr_content = curr.get("content") or ""
-            if curr_content:
-                prev_content = prev.get("content") or ""
-                prev["content"] = (
-                    f"{prev_content}\n{curr_content}" if prev_content else curr_content
-                )
-
-            curr_tool_calls = curr.get("tool_calls")
-            if curr_tool_calls:
-                prev_tool_calls = prev.get("tool_calls")
-                prev["tool_calls"] = (
-                    list(prev_tool_calls) + list(curr_tool_calls)
-                    if prev_tool_calls
-                    else list(curr_tool_calls)
-                )
-        return result
+        return messages


-def _parse_json_field(value: str | dict | list | None, default: Any = None) -> Any:
-    """Parse a JSON field that may be stored as string or already parsed."""
-    if value is None:
-        return default
-    if isinstance(value, str):
-        return json.loads(value)
-    return value
+async def _get_session_from_cache(session_id: str) -> ChatSession | None:
+    """Get a chat session from Redis cache."""
+    redis_key = _get_session_cache_key(session_id)
+    async_redis = await get_redis_async()
+    raw_session: bytes | None = await async_redis.get(redis_key)
+
+    if raw_session is None:
+        return None
+
+    try:
+        session = ChatSession.model_validate_json(raw_session)
+        logger.info(
+            f"Loading session {session_id} from cache: "
+            f"message_count={len(session.messages)}, "
+            f"roles={[m.role for m in session.messages]}"
+        )
+        return session
+    except Exception as e:
+        logger.error(f"Failed to deserialize session {session_id}: {e}", exc_info=True)
+        raise RedisError(f"Corrupted session data for {session_id}") from e


-# ================ Chat cache + DB operations ================ #
-
-# NOTE: Database calls are automatically routed through DatabaseManager if Prisma is not
-#       connected directly.
-
-
-async def cache_chat_session(session: ChatSession) -> None:
-    """Cache a chat session in Redis (without persisting to the database)."""
+async def _cache_session(session: ChatSession) -> None:
+    """Cache a chat session in Redis."""
    redis_key = _get_session_cache_key(session.session_id)
    async_redis = await get_redis_async()
    await async_redis.setex(redis_key, config.session_ttl, session.model_dump_json())


+async def cache_chat_session(session: ChatSession) -> None:
+    """Cache a chat session without persisting to the database."""
+    await _cache_session(session)
+
+
 async def invalidate_session_cache(session_id: str) -> None:
    """Invalidate a chat session from Redis cache.

@@ -339,6 +310,80 @@ async def invalidate_session_cache(session_id: str) -> None:
        logger.warning(f"Failed to invalidate session cache for {session_id}: {e}")


+async def _get_session_from_db(session_id: str) -> ChatSession | None:
+    """Get a chat session from the database."""
+    prisma_session = await chat_db.get_chat_session(session_id)
+    if not prisma_session:
+        return None
+
+    messages = prisma_session.Messages
+    logger.info(
+        f"Loading session {session_id} from DB: "
+        f"has_messages={messages is not None}, "
+        f"message_count={len(messages) if messages else 0}, "
+        f"roles={[m.role for m in messages] if messages else []}"
+    )
+
+    return ChatSession.from_db(prisma_session, messages)
+
+
+async def _save_session_to_db(
+    session: ChatSession, existing_message_count: int
+) -> None:
+    """Save or update a chat session in the database."""
+    # Check if session exists in DB
+    existing = await chat_db.get_chat_session(session.session_id)
+
+    if not existing:
+        # Create new session
+        await chat_db.create_chat_session(
+            session_id=session.session_id,
+            user_id=session.user_id,
+        )
+        existing_message_count = 0
+
+    # Calculate total tokens from usage
+    total_prompt = sum(u.prompt_tokens for u in session.usage)
+    total_completion = sum(u.completion_tokens for u in session.usage)
+
+    # Update session metadata
+    await chat_db.update_chat_session(
+        session_id=session.session_id,
+        credentials=session.credentials,
+        successful_agent_runs=session.successful_agent_runs,
+        successful_agent_schedules=session.successful_agent_schedules,
+        total_prompt_tokens=total_prompt,
+        total_completion_tokens=total_completion,
+    )
+
+    # Add new messages (only those after existing count)
+    new_messages = session.messages[existing_message_count:]
+    if new_messages:
+        messages_data = []
+        for msg in new_messages:
+            messages_data.append(
+                {
+                    "role": msg.role,
+                    "content": msg.content,
+                    "name": msg.name,
+                    "tool_call_id": msg.tool_call_id,
+                    "refusal": msg.refusal,
+                    "tool_calls": msg.tool_calls,
+                    "function_call": msg.function_call,
+                }
+            )
+        logger.info(
+            f"Saving {len(new_messages)} new messages to DB for session {session.session_id}: "
+            f"roles={[m['role'] for m in messages_data]}, "
+            f"start_sequence={existing_message_count}"
+        )
+        await chat_db.add_chat_messages_batch(
+            session_id=session.session_id,
+            messages=messages_data,
+            start_sequence=existing_message_count,
+        )
+
+
 async def get_chat_session(
    session_id: str,
    user_id: str | None = None,
@@ -370,7 +415,7 @@ async def get_chat_session(
        logger.warning(f"Unexpected cache error for session {session_id}: {e}")

    # Fall back to database
-    logger.debug(f"Session {session_id} not in cache, checking database")
+    logger.info(f"Session {session_id} not in cache, checking database")
    session = await _get_session_from_db(session_id)

    if session is None:
@@ -386,7 +431,7 @@ async def get_chat_session(

    # Cache the session from DB
    try:
-        await cache_chat_session(session)
+        await _cache_session(session)
        logger.info(f"Cached session {session_id} from database")
    except Exception as e:
        logger.warning(f"Failed to cache session {session_id}: {e}")
@@ -394,45 +439,9 @@ async def get_chat_session(
    return session


-async def _get_session_from_cache(session_id: str) -> ChatSession | None:
-    """Get a chat session from Redis cache."""
-    redis_key = _get_session_cache_key(session_id)
-    async_redis = await get_redis_async()
-    raw_session: bytes | None = await async_redis.get(redis_key)
-
-    if raw_session is None:
-        return None
-
-    try:
-        session = ChatSession.model_validate_json(raw_session)
-        logger.info(
-            f"Loading session {session_id} from cache: "
-            f"message_count={len(session.messages)}, "
-            f"roles={[m.role for m in session.messages]}"
-        )
-        return session
-    except Exception as e:
-        logger.error(f"Failed to deserialize session {session_id}: {e}", exc_info=True)
-        raise RedisError(f"Corrupted session data for {session_id}") from e
-
-
-async def _get_session_from_db(session_id: str) -> ChatSession | None:
-    """Get a chat session from the database."""
-    session = await chat_db().get_chat_session(session_id)
-    if not session:
-        return None
-
-    logger.info(
-        f"Loaded session {session_id} from DB: "
-        f"has_messages={bool(session.messages)}, "
-        f"message_count={len(session.messages)}, "
-        f"roles={[m.role for m in session.messages]}"
-    )
-
-    return session
-
-
-async def upsert_chat_session(session: ChatSession) -> ChatSession:
+async def upsert_chat_session(
+    session: ChatSession,
+) -> ChatSession:
    """Update a chat session in both cache and database.

    Uses session-level locking to prevent race conditions when concurrent
@@ -450,7 +459,7 @@ async def upsert_chat_session(session: ChatSession) -> ChatSession:

    async with lock:
        # Get existing message count from DB for incremental saves
-        existing_message_count = await chat_db().get_chat_session_message_count(
+        existing_message_count = await chat_db.get_chat_session_message_count(
            session.session_id
        )

@@ -467,7 +476,7 @@ async def upsert_chat_session(session: ChatSession) -> ChatSession:

        # Save to cache (best-effort, even if DB failed)
        try:
-            await cache_chat_session(session)
+            await _cache_session(session)
        except Exception as e:
            # If DB succeeded but cache failed, raise cache error
            if db_error is None:
@@ -488,99 +497,6 @@ async def upsert_chat_session(session: ChatSession) -> ChatSession:
        return session


-async def _save_session_to_db(
-    session: ChatSession, existing_message_count: int
-) -> None:
-    """Save or update a chat session in the database."""
-    db = chat_db()
-
-    # Check if session exists in DB
-    existing = await db.get_chat_session(session.session_id)
-
-    if not existing:
-        # Create new session
-        await db.create_chat_session(
-            session_id=session.session_id,
-            user_id=session.user_id,
-        )
-        existing_message_count = 0
-
-    # Calculate total tokens from usage
-    total_prompt = sum(u.prompt_tokens for u in session.usage)
-    total_completion = sum(u.completion_tokens for u in session.usage)
-
-    # Update session metadata
-    await db.update_chat_session(
-        session_id=session.session_id,
-        credentials=session.credentials,
-        successful_agent_runs=session.successful_agent_runs,
-        successful_agent_schedules=session.successful_agent_schedules,
-        total_prompt_tokens=total_prompt,
-        total_completion_tokens=total_completion,
-    )
-
-    # Add new messages (only those after existing count)
-    new_messages = session.messages[existing_message_count:]
-    if new_messages:
-        messages_data = []
-        for msg in new_messages:
-            messages_data.append(
-                {
-                    "role": msg.role,
-                    "content": msg.content,
-                    "name": msg.name,
-                    "tool_call_id": msg.tool_call_id,
-                    "refusal": msg.refusal,
-                    "tool_calls": msg.tool_calls,
-                    "function_call": msg.function_call,
-                }
-            )
-        logger.info(
-            f"Saving {len(new_messages)} new messages to DB for session {session.session_id}: "
-            f"roles={[m['role'] for m in messages_data]}, "
-            f"start_sequence={existing_message_count}"
-        )
-        await db.add_chat_messages_batch(
-            session_id=session.session_id,
-            messages=messages_data,
-            start_sequence=existing_message_count,
-        )
-
-
-async def append_and_save_message(session_id: str, message: ChatMessage) -> ChatSession:
-    """Atomically append a message to a session and persist it.
-
-    Acquires the session lock, re-fetches the latest session state,
-    appends the message, and saves — preventing message loss when
-    concurrent requests modify the same session.
-    """
-    lock = await _get_session_lock(session_id)
-
-    async with lock:
-        session = await get_chat_session(session_id)
-        if session is None:
-            raise ValueError(f"Session {session_id} not found")
-
-        session.messages.append(message)
-        existing_message_count = await chat_db().get_chat_session_message_count(
-            session_id
-        )
-
-        try:
-            await _save_session_to_db(session, existing_message_count)
-        except Exception as e:
-            raise DatabaseError(
-                f"Failed to persist message to session {session_id}"
-            ) from e
-
-        try:
-            await cache_chat_session(session)
-        except Exception as e:
-            logger.warning(f"Cache write failed for session {session_id}: {e}")
-
-        return session
-
-
 async def create_chat_session(user_id: str) -> ChatSession:
    """Create a new chat session and persist it.

@@ -593,7 +509,7 @@ async def create_chat_session(user_id: str) -> ChatSession:

    # Create in database first - fail fast if this fails
    try:
-        await chat_db().create_chat_session(
+        await chat_db.create_chat_session(
            session_id=session.session_id,
            user_id=user_id,
        )
@@ -605,7 +521,7 @@ async def create_chat_session(user_id: str) -> ChatSession:

    # Cache the session (best-effort optimization, DB is source of truth)
    try:
-        await cache_chat_session(session)
+        await _cache_session(session)
    except Exception as e:
        logger.warning(f"Failed to cache new session {session.session_id}: {e}")

@@ -616,16 +532,20 @@ async def get_user_sessions(
    user_id: str,
    limit: int = 50,
    offset: int = 0,
-) -> tuple[list[ChatSessionInfo], int]:
+) -> tuple[list[ChatSession], int]:
    """Get chat sessions for a user from the database with total count.

    Returns:
        A tuple of (sessions, total_count) where total_count is the overall
        number of sessions for the user (not just the current page).
    """
-    db = chat_db()
-    sessions = await db.get_user_chat_sessions(user_id, limit, offset)
-    total_count = await db.get_user_session_count(user_id)
+    prisma_sessions = await chat_db.get_user_chat_sessions(user_id, limit, offset)
+    total_count = await chat_db.get_user_session_count(user_id)
+
+    sessions = []
+    for prisma_session in prisma_sessions:
+        # Convert without messages for listing (lighter weight)
+        sessions.append(ChatSession.from_db(prisma_session, None))

    return sessions, total_count

@@ -643,7 +563,7 @@ async def delete_chat_session(session_id: str, user_id: str | None = None) -> bo
    """
    # Delete from database first (with optional user_id validation)
    # This confirms ownership before invalidating cache
-    deleted = await chat_db().delete_chat_session(session_id, user_id)
+    deleted = await chat_db.delete_chat_session(session_id, user_id)

    if not deleted:
        return False
@@ -678,52 +598,20 @@ async def update_session_title(session_id: str, title: str) -> bool:
        True if updated successfully, False otherwise.
    """
    try:
-        result = await chat_db().update_chat_session(session_id=session_id, title=title)
+        result = await chat_db.update_chat_session(session_id=session_id, title=title)
        if result is None:
            logger.warning(f"Session {session_id} not found for title update")
            return False

-        # Update title in cache if it exists (instead of invalidating).
-        # This prevents race conditions where cache invalidation causes
-        # the frontend to see stale DB data while streaming is still in progress.
+        # Invalidate cache so next fetch gets updated title
        try:
-            cached = await _get_session_from_cache(session_id)
-            if cached:
-                cached.title = title
-                await cache_chat_session(cached)
+            redis_key = _get_session_cache_key(session_id)
+            async_redis = await get_redis_async()
+            await async_redis.delete(redis_key)
        except Exception as e:
-            # Not critical - title will be correct on next full cache refresh
-            logger.warning(
-                f"Failed to update title in cache for session {session_id}: {e}"
-            )
+            logger.warning(f"Failed to invalidate cache for session {session_id}: {e}")

        return True
    except Exception as e:
        logger.error(f"Failed to update title for session {session_id}: {e}")
        return False
-
-
-# ==================== Chat session locks ==================== #
-
-_session_locks: WeakValueDictionary[str, asyncio.Lock] = WeakValueDictionary()
-_session_locks_mutex = asyncio.Lock()
-
-
-async def _get_session_lock(session_id: str) -> asyncio.Lock:
-    """Get or create a lock for a specific session to prevent concurrent upserts.
-
-    This was originally added to solve the specific problem of race conditions between
-    the session title thread and the conversation thread, which always occurs on the
-    same instance as we prevent rapid request sends on the frontend.
-
-    Uses WeakValueDictionary for automatic cleanup: locks are garbage collected
-    when no coroutine holds a reference to them, preventing memory leaks from
-    unbounded growth of session locks. Explicit cleanup also occurs
-    in `delete_chat_session()`.
-    """
-    async with _session_locks_mutex:
-        lock = _session_locks.get(session_id)
-        if lock is None:
-            lock = asyncio.Lock()
-            _session_locks[session_id] = lock
-        return lock
--- a/autogpt_platform/backend/backend/api/features/chat/model_test.py
+++ b/autogpt_platform/backend/backend/api/features/chat/model_test.py
@@ -0,0 +1,119 @@
+import pytest
+
+from .model import (
+    ChatMessage,
+    ChatSession,
+    Usage,
+    get_chat_session,
+    upsert_chat_session,
+)
+
+messages = [
+    ChatMessage(content="Hello, how are you?", role="user"),
+    ChatMessage(
+        content="I'm fine, thank you!",
+        role="assistant",
+        tool_calls=[
+            {
+                "id": "t123",
+                "type": "function",
+                "function": {
+                    "name": "get_weather",
+                    "arguments": '{"city": "New York"}',
+                },
+            }
+        ],
+    ),
+    ChatMessage(
+        content="I'm using the tool to get the weather",
+        role="tool",
+        tool_call_id="t123",
+    ),
+]
+
+
+@pytest.mark.asyncio(loop_scope="session")
+async def test_chatsession_serialization_deserialization():
+    s = ChatSession.new(user_id="abc123")
+    s.messages = messages
+    s.usage = [Usage(prompt_tokens=100, completion_tokens=200, total_tokens=300)]
+    serialized = s.model_dump_json()
+    s2 = ChatSession.model_validate_json(serialized)
+    assert s2.model_dump() == s.model_dump()
+
+
+@pytest.mark.asyncio(loop_scope="session")
+async def test_chatsession_redis_storage(setup_test_user, test_user_id):
+
+    s = ChatSession.new(user_id=test_user_id)
+    s.messages = messages
+
+    s = await upsert_chat_session(s)
+
+    s2 = await get_chat_session(
+        session_id=s.session_id,
+        user_id=s.user_id,
+    )
+
+    assert s2 == s
+
+
+@pytest.mark.asyncio(loop_scope="session")
+async def test_chatsession_redis_storage_user_id_mismatch(
+    setup_test_user, test_user_id
+):
+
+    s = ChatSession.new(user_id=test_user_id)
+    s.messages = messages
+    s = await upsert_chat_session(s)
+
+    s2 = await get_chat_session(s.session_id, "different_user_id")
+
+    assert s2 is None
+
+
+@pytest.mark.asyncio(loop_scope="session")
+async def test_chatsession_db_storage(setup_test_user, test_user_id):
+    """Test that messages are correctly saved to and loaded from DB (not cache)."""
+    from backend.data.redis_client import get_redis_async
+
+    # Create session with messages including assistant message
+    s = ChatSession.new(user_id=test_user_id)
+    s.messages = messages  # Contains user, assistant, and tool messages
+    assert s.session_id is not None, "Session id is not set"
+    # Upsert to save to both cache and DB
+    s = await upsert_chat_session(s)
+
+    # Clear the Redis cache to force DB load
+    redis_key = f"chat:session:{s.session_id}"
+    async_redis = await get_redis_async()
+    await async_redis.delete(redis_key)
+
+    # Load from DB (cache was cleared)
+    s2 = await get_chat_session(
+        session_id=s.session_id,
+        user_id=s.user_id,
+    )
+
+    assert s2 is not None, "Session not found after loading from DB"
+    assert len(s2.messages) == len(
+        s.messages
+    ), f"Message count mismatch: expected {len(s.messages)}, got {len(s2.messages)}"
+
+    # Verify all roles are present
+    roles = [m.role for m in s2.messages]
+    assert "user" in roles, f"User message missing. Roles found: {roles}"
+    assert "assistant" in roles, f"Assistant message missing. Roles found: {roles}"
+    assert "tool" in roles, f"Tool message missing. Roles found: {roles}"
+
+    # Verify message content
+    for orig, loaded in zip(s.messages, s2.messages):
+        assert orig.role == loaded.role, f"Role mismatch: {orig.role} != {loaded.role}"
+        assert (
+            orig.content == loaded.content
+        ), f"Content mismatch for {orig.role}: {orig.content} != {loaded.content}"
+        if orig.tool_calls:
+            assert (
+                loaded.tool_calls is not None
+            ), f"Tool calls missing for {orig.role} message"
+            assert len(orig.tool_calls) == len(loaded.tool_calls)
--- a/autogpt_platform/backend/backend/api/features/chat/response_model.py
+++ b/autogpt_platform/backend/backend/api/features/chat/response_model.py
@@ -10,8 +10,6 @@ from typing import Any

 from pydantic import BaseModel, Field

-from backend.util.json import dumps as json_dumps
-

 class ResponseType(str, Enum):
    """Types of streaming responses following AI SDK protocol."""
@@ -20,10 +18,6 @@ class ResponseType(str, Enum):
    START = "start"
    FINISH = "finish"

-    # Step lifecycle (one LLM API call within a message)
-    START_STEP = "start-step"
-    FINISH_STEP = "finish-step"
-
    # Text streaming
    TEXT_START = "text-start"
    TEXT_DELTA = "text-delta"
@@ -63,16 +57,6 @@ class StreamStart(StreamBaseResponse):
        description="Task ID for SSE reconnection. Clients can reconnect using GET /tasks/{taskId}/stream",
    )

-    def to_sse(self) -> str:
-        """Convert to SSE format, excluding non-protocol fields like taskId."""
-        import json
-
-        data: dict[str, Any] = {
-            "type": self.type.value,
-            "messageId": self.messageId,
-        }
-        return f"data: {json.dumps(data)}\n\n"
-

 class StreamFinish(StreamBaseResponse):
    """End of message/stream."""
@@ -80,26 +64,6 @@ class StreamFinish(StreamBaseResponse):
    type: ResponseType = ResponseType.FINISH


-class StreamStartStep(StreamBaseResponse):
-    """Start of a step (one LLM API call within a message).
-
-    The AI SDK uses this to add a step-start boundary to message.parts,
-    enabling visual separation between multiple LLM calls in a single message.
-    """
-
-    type: ResponseType = ResponseType.START_STEP
-
-
-class StreamFinishStep(StreamBaseResponse):
-    """End of a step (one LLM API call within a message).
-
-    The AI SDK uses this to reset activeTextParts and activeReasoningParts,
-    so the next LLM call in a tool-call continuation starts with clean state.
-    """
-
-    type: ResponseType = ResponseType.FINISH_STEP
-
-
 # ========== Text Streaming ==========


@@ -153,7 +117,7 @@ class StreamToolOutputAvailable(StreamBaseResponse):
    type: ResponseType = ResponseType.TOOL_OUTPUT_AVAILABLE
    toolCallId: str = Field(..., description="Tool call ID this responds to")
    output: str | dict[str, Any] = Field(..., description="Tool execution output")
-    # Keep these for internal backend use
+    # Additional fields for internal use (not part of AI SDK spec but useful)
    toolName: str | None = Field(
        default=None, description="Name of the tool that was executed"
    )
@@ -161,17 +125,6 @@ class StreamToolOutputAvailable(StreamBaseResponse):
        default=True, description="Whether the tool execution succeeded"
    )

-    def to_sse(self) -> str:
-        """Convert to SSE format, excluding non-spec fields."""
-        import json
-
-        data = {
-            "type": self.type.value,
-            "toolCallId": self.toolCallId,
-            "output": self.output,
-        }
-        return f"data: {json.dumps(data)}\n\n"
-

 # ========== Other ==========

@@ -195,18 +148,6 @@ class StreamError(StreamBaseResponse):
        default=None, description="Additional error details"
    )

-    def to_sse(self) -> str:
-        """Convert to SSE format, only emitting fields required by AI SDK protocol.
-
-        The AI SDK uses z.strictObject({type, errorText}) which rejects
-        any extra fields like `code` or `details`.
-        """
-        data = {
-            "type": self.type.value,
-            "errorText": self.errorText,
-        }
-        return f"data: {json_dumps(data)}\n\n"
-

 class StreamHeartbeat(StreamBaseResponse):
    """Heartbeat to keep SSE connection alive during long-running operations.
--- a/autogpt_platform/backend/backend/api/features/chat/routes.py
+++ b/autogpt_platform/backend/backend/api/features/chat/routes.py
@@ -1,60 +1,24 @@
 """Chat API routes for chat session management and streaming via SSE."""

-import asyncio
 import logging
 import uuid as uuid_module
 from collections.abc import AsyncGenerator
 from typing import Annotated

 from autogpt_libs import auth
-from fastapi import APIRouter, Depends, Header, HTTPException, Query, Response, Security
+from fastapi import APIRouter, Depends, Header, HTTPException, Query, Security
 from fastapi.responses import StreamingResponse
 from pydantic import BaseModel

-from backend.copilot import service as chat_service
-from backend.copilot import stream_registry
-from backend.copilot.completion_handler import (
-    process_operation_failure,
-    process_operation_success,
-)
-from backend.copilot.config import ChatConfig
-from backend.copilot.executor.utils import enqueue_copilot_task
-from backend.copilot.model import (
-    ChatMessage,
-    ChatSession,
-    append_and_save_message,
-    create_chat_session,
-    delete_chat_session,
-    get_chat_session,
-    get_user_sessions,
-)
-from backend.copilot.response_model import StreamError, StreamFinish, StreamHeartbeat
-from backend.copilot.tools.models import (
-    AgentDetailsResponse,
-    AgentOutputResponse,
-    AgentPreviewResponse,
-    AgentSavedResponse,
-    AgentsFoundResponse,
-    BlockDetailsResponse,
-    BlockListResponse,
-    BlockOutputResponse,
-    ClarificationNeededResponse,
-    DocPageResponse,
-    DocSearchResultsResponse,
-    ErrorResponse,
-    ExecutionStartedResponse,
-    InputValidationErrorResponse,
-    NeedLoginResponse,
-    NoResultsResponse,
-    OperationInProgressResponse,
-    OperationPendingResponse,
-    OperationStartedResponse,
-    SetupRequirementsResponse,
-    UnderstandingUpdatedResponse,
-)
-from backend.copilot.tracking import track_user_message
 from backend.util.exceptions import NotFoundError

+from . import service as chat_service
+from . import stream_registry
+from .completion_handler import process_operation_failure, process_operation_success
+from .config import ChatConfig
+from .model import ChatSession, create_chat_session, get_chat_session, get_user_sessions
+from .response_model import StreamFinish, StreamHeartbeat, StreamStart
+
 config = ChatConfig()


@@ -213,43 +177,6 @@ async def create_session(
    )


-@router.delete(
-    "/sessions/{session_id}",
-    dependencies=[Security(auth.requires_user)],
-    status_code=204,
-    responses={404: {"description": "Session not found or access denied"}},
-)
-async def delete_session(
-    session_id: str,
-    user_id: Annotated[str, Security(auth.get_user_id)],
-) -> Response:
-    """
-    Delete a chat session.
-
-    Permanently removes a chat session and all its messages.
-    Only the owner can delete their sessions.
-
-    Args:
-        session_id: The session ID to delete.
-        user_id: The authenticated user's ID.
-
-    Returns:
-        204 No Content on success.
-
-    Raises:
-        HTTPException: 404 if session not found or not owned by user.
-    """
-    deleted = await delete_chat_session(session_id, user_id)
-
-    if not deleted:
-        raise HTTPException(
-            status_code=404,
-            detail=f"Session {session_id} not found or access denied",
-        )
-
-    return Response(status_code=204)
-
-
@router.get(
    "/sessions/{session_id}",
 )
@@ -282,10 +209,6 @@ async def get_session(
    active_task, last_message_id = await stream_registry.get_active_task_for_session(
        session_id, user_id
    )
-    logger.info(
-        f"[GET_SESSION] session={session_id}, active_task={active_task is not None}, "
-        f"msg_count={len(messages)}, last_role={messages[-1].get('role') if messages else 'none'}"
-    )
    if active_task:
        # Filter out the in-progress assistant message from the session response.
        # The client will receive the complete assistant response through the SSE
@@ -343,54 +266,12 @@ async def stream_chat_post(

    """
    import asyncio
-    import time

-    stream_start_time = time.perf_counter()
-    log_meta = {"component": "ChatStream", "session_id": session_id}
-    if user_id:
-        log_meta["user_id"] = user_id
-
-    logger.info(
-        f"[TIMING] stream_chat_post STARTED, session={session_id}, "
-        f"user={user_id}, message_len={len(request.message)}",
-        extra={"json_fields": log_meta},
-    )
-    await _validate_and_get_session(session_id, user_id)
-    logger.info(
-        f"[TIMING] session validated in {(time.perf_counter() - stream_start_time) * 1000:.1f}ms",
-        extra={
-            "json_fields": {
-                **log_meta,
-                "duration_ms": (time.perf_counter() - stream_start_time) * 1000,
-            }
-        },
-    )
-
-    # Atomically append user message to session BEFORE creating task to avoid
-    # race condition where GET_SESSION sees task as "running" but message isn't
-    # saved yet.  append_and_save_message re-fetches inside a lock to prevent
-    # message loss from concurrent requests.
-    if request.message:
-        message = ChatMessage(
-            role="user" if request.is_user_message else "assistant",
-            content=request.message,
-        )
-        if request.is_user_message:
-            track_user_message(
-                user_id=user_id,
-                session_id=session_id,
-                message_length=len(request.message),
-            )
-        logger.info(f"[STREAM] Saving user message to session {session_id}")
-        await append_and_save_message(session_id, message)
-        logger.info(f"[STREAM] User message saved for session {session_id}")
+    session = await _validate_and_get_session(session_id, user_id)

    # Create a task in the stream registry for reconnection support
    task_id = str(uuid_module.uuid4())
    operation_id = str(uuid_module.uuid4())
-    log_meta["task_id"] = task_id
-
-    task_create_start = time.perf_counter()
    await stream_registry.create_task(
        task_id=task_id,
        session_id=session_id,
@@ -399,45 +280,40 @@ async def stream_chat_post(
        tool_name="chat",
        operation_id=operation_id,
    )
-    logger.info(
-        f"[TIMING] create_task completed in {(time.perf_counter() - task_create_start) * 1000:.1f}ms",
-        extra={
-            "json_fields": {
-                **log_meta,
-                "duration_ms": (time.perf_counter() - task_create_start) * 1000,
-            }
-        },
-    )

-    await enqueue_copilot_task(
-        task_id=task_id,
-        session_id=session_id,
-        user_id=user_id,
-        operation_id=operation_id,
-        message=request.message,
-        is_user_message=request.is_user_message,
-        context=request.context,
-    )
+    # Background task that runs the AI generation independently of SSE connection
+    async def run_ai_generation():
+        try:
+            # Emit a start event with task_id for reconnection
+            start_chunk = StreamStart(messageId=task_id, taskId=task_id)
+            await stream_registry.publish_chunk(task_id, start_chunk)

-    setup_time = (time.perf_counter() - stream_start_time) * 1000
-    logger.info(
-        f"[TIMING] Task enqueued to RabbitMQ, setup={setup_time:.1f}ms",
-        extra={"json_fields": {**log_meta, "setup_time_ms": setup_time}},
-    )
+            async for chunk in chat_service.stream_chat_completion(
+                session_id,
+                request.message,
+                is_user_message=request.is_user_message,
+                user_id=user_id,
+                session=session,  # Pass pre-fetched session to avoid double-fetch
+                context=request.context,
+            ):
+                # Write to Redis (subscribers will receive via XREAD)
+                await stream_registry.publish_chunk(task_id, chunk)
+
+            # Mark task as completed
+            await stream_registry.mark_task_completed(task_id, "completed")
+        except Exception as e:
+            logger.error(
+                f"Error in background AI generation for session {session_id}: {e}"
+            )
+            await stream_registry.mark_task_completed(task_id, "failed")
+
+    # Start the AI generation in a background task
+    bg_task = asyncio.create_task(run_ai_generation())
+    await stream_registry.set_task_asyncio_task(task_id, bg_task)

    # SSE endpoint that subscribes to the task's stream
    async def event_generator() -> AsyncGenerator[str, None]:
-        import time as time_module
-
-        event_gen_start = time_module.perf_counter()
-        logger.info(
-            f"[TIMING] event_generator STARTED, task={task_id}, session={session_id}, "
-            f"user={user_id}",
-            extra={"json_fields": log_meta},
-        )
        subscriber_queue = None
-        first_chunk_yielded = False
-        chunks_yielded = 0
        try:
            # Subscribe to the task stream (this replays existing messages + live updates)
            subscriber_queue = await stream_registry.subscribe_to_task(
@@ -452,78 +328,24 @@ async def stream_chat_post(
                return

            # Read from the subscriber queue and yield to SSE
-            logger.info(
-                "[TIMING] Starting to read from subscriber_queue",
-                extra={"json_fields": log_meta},
-            )
            while True:
                try:
                    chunk = await asyncio.wait_for(subscriber_queue.get(), timeout=30.0)
-                    chunks_yielded += 1
-
-                    if not first_chunk_yielded:
-                        first_chunk_yielded = True
-                        elapsed = time_module.perf_counter() - event_gen_start
-                        logger.info(
-                            f"[TIMING] FIRST CHUNK from queue at {elapsed:.2f}s, "
-                            f"type={type(chunk).__name__}",
-                            extra={
-                                "json_fields": {
-                                    **log_meta,
-                                    "chunk_type": type(chunk).__name__,
-                                    "elapsed_ms": elapsed * 1000,
-                                }
-                            },
-                        )
-
                    yield chunk.to_sse()

                    # Check for finish signal
                    if isinstance(chunk, StreamFinish):
-                        total_time = time_module.perf_counter() - event_gen_start
-                        logger.info(
-                            f"[TIMING] StreamFinish received in {total_time:.2f}s; "
-                            f"n_chunks={chunks_yielded}",
-                            extra={
-                                "json_fields": {
-                                    **log_meta,
-                                    "chunks_yielded": chunks_yielded,
-                                    "total_time_ms": total_time * 1000,
-                                }
-                            },
-                        )
                        break
                except asyncio.TimeoutError:
+                    # Send heartbeat to keep connection alive
                    yield StreamHeartbeat().to_sse()

        except GeneratorExit:
-            logger.info(
-                f"[TIMING] GeneratorExit (client disconnected), chunks={chunks_yielded}",
-                extra={
-                    "json_fields": {
-                        **log_meta,
-                        "chunks_yielded": chunks_yielded,
-                        "reason": "client_disconnect",
-                    }
-                },
-            )
            pass  # Client disconnected - background task continues
        except Exception as e:
-            elapsed = (time_module.perf_counter() - event_gen_start) * 1000
-            logger.error(
-                f"[TIMING] event_generator ERROR after {elapsed:.1f}ms: {e}",
-                extra={
-                    "json_fields": {**log_meta, "elapsed_ms": elapsed, "error": str(e)}
-                },
-            )
-            # Surface error to frontend so it doesn't appear stuck
-            yield StreamError(
-                errorText="An error occurred. Please try again.",
-                code="stream_error",
-            ).to_sse()
-            yield StreamFinish().to_sse()
+            logger.error(f"Error in SSE stream for task {task_id}: {e}")
        finally:
-            # Unsubscribe when client disconnects or stream ends
+            # Unsubscribe when client disconnects or stream ends to prevent resource leak
            if subscriber_queue is not None:
                try:
                    await stream_registry.unsubscribe_from_task(
@@ -535,18 +357,6 @@ async def stream_chat_post(
                        exc_info=True,
                    )
            # AI SDK protocol termination - always yield even if unsubscribe fails
-            total_time = time_module.perf_counter() - event_gen_start
-            logger.info(
-                f"[TIMING] event_generator FINISHED in {total_time:.2f}s; "
-                f"task={task_id}, session={session_id}, n_chunks={chunks_yielded}",
-                extra={
-                    "json_fields": {
-                        **log_meta,
-                        "total_time_ms": total_time * 1000,
-                        "chunks_yielded": chunks_yielded,
-                    }
-                },
-            )
            yield "data: [DONE]\n\n"

    return StreamingResponse(
@@ -564,90 +374,63 @@ async def stream_chat_post(
@router.get(
    "/sessions/{session_id}/stream",
 )
-async def resume_session_stream(
+async def stream_chat_get(
    session_id: str,
+    message: Annotated[str, Query(min_length=1, max_length=10000)],
    user_id: str | None = Depends(auth.get_user_id),
+    is_user_message: bool = Query(default=True),
 ):
    """
-    Resume an active stream for a session.
+    Stream chat responses for a session (GET - legacy endpoint).

-    Called by the AI SDK's ``useChat(resume: true)`` on page load.
-    Checks for an active (in-progress) task on the session and either replays
-    the full SSE stream or returns 204 No Content if nothing is running.
+    Streams the AI/completion responses in real time over Server-Sent Events (SSE), including:
+      - Text fragments as they are generated
+      - Tool call UI elements (if invoked)
+      - Tool execution results

    Args:
-        session_id: The chat session identifier.
+        session_id: The chat session identifier to associate with the streamed messages.
+        message: The user's new message to process.
        user_id: Optional authenticated user ID.
-
+        is_user_message: Whether the message is a user message.
    Returns:
-        StreamingResponse (SSE) when an active stream exists,
-        or 204 No Content when there is nothing to resume.
+        StreamingResponse: SSE-formatted response chunks.
+
    """
-    import asyncio
-
-    active_task, _last_id = await stream_registry.get_active_task_for_session(
-        session_id, user_id
-    )
-
-    if not active_task:
-        return Response(status_code=204)
-
-    subscriber_queue = await stream_registry.subscribe_to_task(
-        task_id=active_task.task_id,
-        user_id=user_id,
-        last_message_id="0-0",  # Full replay so useChat rebuilds the message
-    )
-
-    if subscriber_queue is None:
-        return Response(status_code=204)
+    session = await _validate_and_get_session(session_id, user_id)

    async def event_generator() -> AsyncGenerator[str, None]:
        chunk_count = 0
        first_chunk_type: str | None = None
-        try:
-            while True:
-                try:
-                    chunk = await asyncio.wait_for(subscriber_queue.get(), timeout=30.0)
-                    if chunk_count < 3:
-                        logger.info(
-                            "Resume stream chunk",
-                            extra={
-                                "session_id": session_id,
-                                "chunk_type": str(chunk.type),
-                            },
-                        )
-                    if not first_chunk_type:
-                        first_chunk_type = str(chunk.type)
-                    chunk_count += 1
-                    yield chunk.to_sse()
-
-                    if isinstance(chunk, StreamFinish):
-                        break
-                except asyncio.TimeoutError:
-                    yield StreamHeartbeat().to_sse()
-        except GeneratorExit:
-            pass
-        except Exception as e:
-            logger.error(f"Error in resume stream for session {session_id}: {e}")
-        finally:
-            try:
-                await stream_registry.unsubscribe_from_task(
-                    active_task.task_id, subscriber_queue
+        async for chunk in chat_service.stream_chat_completion(
+            session_id,
+            message,
+            is_user_message=is_user_message,
+            user_id=user_id,
+            session=session,  # Pass pre-fetched session to avoid double-fetch
+        ):
+            if chunk_count < 3:
+                logger.info(
+                    "Chat stream chunk",
+                    extra={
+                        "session_id": session_id,
+                        "chunk_type": str(chunk.type),
+                    },
                )
-            except Exception as unsub_err:
-                logger.error(
-                    f"Error unsubscribing from task {active_task.task_id}: {unsub_err}",
-                    exc_info=True,
-                )
-            logger.info(
-                "Resume stream completed",
-                extra={
-                    "session_id": session_id,
-                    "n_chunks": chunk_count,
-                    "first_chunk_type": first_chunk_type,
-                },
-            )
-            yield "data: [DONE]\n\n"
+            if not first_chunk_type:
+                first_chunk_type = str(chunk.type)
+            chunk_count += 1
+            yield chunk.to_sse()
+        logger.info(
+            "Chat stream completed",
+            extra={
+                "session_id": session_id,
+                "chunk_count": chunk_count,
+                "first_chunk_type": first_chunk_type,
+            },
+        )
+        # AI SDK protocol termination
+        yield "data: [DONE]\n\n"

    return StreamingResponse(
        event_generator(),
@@ -655,8 +438,8 @@ async def resume_session_stream(
        headers={
            "Cache-Control": "no-cache",
            "Connection": "keep-alive",
-            "X-Accel-Buffering": "no",
-            "x-vercel-ai-ui-message-stream": "v1",
+            "X-Accel-Buffering": "no",  # Disable nginx buffering
+            "x-vercel-ai-ui-message-stream": "v1",  # AI SDK protocol header
        },
    )

@@ -767,6 +550,8 @@ async def stream_task(
        )

    async def event_generator() -> AsyncGenerator[str, None]:
+        import asyncio
+
        heartbeat_interval = 15.0  # Send heartbeat every 15 seconds
        try:
            while True:
@@ -966,43 +751,3 @@ async def health_check() -> dict:
        "service": "chat",
        "version": "0.1.0",
    }
-
-
-# ========== Schema Export (for OpenAPI / Orval codegen) ==========
-
-ToolResponseUnion = (
-    AgentsFoundResponse
-    | NoResultsResponse
-    | AgentDetailsResponse
-    | SetupRequirementsResponse
-    | ExecutionStartedResponse
-    | NeedLoginResponse
-    | ErrorResponse
-    | InputValidationErrorResponse
-    | AgentOutputResponse
-    | UnderstandingUpdatedResponse
-    | AgentPreviewResponse
-    | AgentSavedResponse
-    | ClarificationNeededResponse
-    | BlockListResponse
-    | BlockDetailsResponse
-    | BlockOutputResponse
-    | DocSearchResultsResponse
-    | DocPageResponse
-    | OperationStartedResponse
-    | OperationPendingResponse
-    | OperationInProgressResponse
-)
-
-
-@router.get(
-    "/schema/tool-responses",
-    response_model=ToolResponseUnion,
-    include_in_schema=True,
-    summary="[Dummy] Tool response type export for codegen",
-    description="This endpoint is not meant to be called. It exists solely to "
-    "expose tool response models in the OpenAPI schema for frontend codegen.",
-)
-async def _tool_response_schema() -> ToolResponseUnion:  # type: ignore[return]
-    """Never called at runtime. Exists only so Orval generates TS types."""
-    raise HTTPException(status_code=501, detail="Schema-only endpoint")
--- a/autogpt_platform/backend/backend/api/features/chat/service.py
+++ b/autogpt_platform/backend/backend/api/features/chat/service.py
@@ -27,18 +27,20 @@ from openai.types.chat import (
    ChatCompletionToolParam,
 )

-from backend.data.db_accessors import chat_db, understanding_db
 from backend.data.redis_client import get_redis_async
-from backend.data.understanding import format_understanding_for_prompt
+from backend.data.understanding import (
+    format_understanding_for_prompt,
+    get_business_understanding,
+)
 from backend.util.exceptions import NotFoundError
 from backend.util.settings import AppEnvironment, Settings

+from . import db as chat_db
 from . import stream_registry
 from .config import ChatConfig
 from .model import (
    ChatMessage,
    ChatSession,
-    ChatSessionInfo,
    Usage,
    cache_chat_session,
    get_chat_session,
@@ -50,10 +52,8 @@ from .response_model import (
    StreamBaseResponse,
    StreamError,
    StreamFinish,
-    StreamFinishStep,
    StreamHeartbeat,
    StreamStart,
-    StreamStartStep,
    StreamTextDelta,
    StreamTextEnd,
    StreamTextStart,
@@ -243,16 +243,12 @@ async def _get_system_prompt_template(context: str) -> str:
    return DEFAULT_SYSTEM_PROMPT.format(users_information=context)


-async def _build_system_prompt(
-    user_id: str | None, has_conversation_history: bool = False
-) -> tuple[str, Any]:
+async def _build_system_prompt(user_id: str | None) -> tuple[str, Any]:
    """Build the full system prompt including business understanding if available.

    Args:
-        user_id: The user ID for fetching business understanding.
-        has_conversation_history: Whether there's existing conversation history.
-            If True, we don't tell the model to greet/introduce (since they're
-            already in a conversation).
+        user_id: The user ID for fetching business understanding
+                     If "default" and this is the user's first session, will use "onboarding" instead.

    Returns:
        Tuple of (compiled prompt string, business understanding object)
@@ -261,15 +257,13 @@ async def _build_system_prompt(
    understanding = None
    if user_id:
        try:
-            understanding = await understanding_db().get_business_understanding(user_id)
+            understanding = await get_business_understanding(user_id)
        except Exception as e:
            logger.warning(f"Failed to fetch business understanding: {e}")
            understanding = None

    if understanding:
        context = format_understanding_for_prompt(understanding)
-    elif has_conversation_history:
-        context = "No prior understanding saved yet. Continue the existing conversation naturally."
    else:
        context = "This is the first time you are meeting the user. Greet them and introduce them to the platform"

@@ -337,7 +331,7 @@ async def _generate_session_title(
 async def assign_user_to_session(
    session_id: str,
    user_id: str,
-) -> ChatSessionInfo:
+) -> ChatSession:
    """
    Assign a user to a chat session.
    """
@@ -357,10 +351,6 @@ async def stream_chat_completion(
    retry_count: int = 0,
    session: ChatSession | None = None,
    context: dict[str, str] | None = None,  # {url: str, content: str}
-    _continuation_message_id: (
-        str | None
-    ) = None,  # Internal: reuse message ID for tool call continuations
-    _task_id: str | None = None,  # Internal: task ID for SSE reconnection support
 ) -> AsyncGenerator[StreamBaseResponse, None]:
    """Main entry point for streaming chat completions with database handling.

@@ -378,47 +368,24 @@ async def stream_chat_completion(

    Raises:
        NotFoundError: If session_id is invalid
+        ValueError: If max_context_messages is exceeded

    """
-    completion_start = time.monotonic()
-
-    # Build log metadata for structured logging
-    log_meta = {"component": "ChatService", "session_id": session_id}
-    if user_id:
-        log_meta["user_id"] = user_id
-
    logger.info(
-        f"[TIMING] stream_chat_completion STARTED, session={session_id}, user={user_id}, "
-        f"message_len={len(message) if message else 0}, is_user={is_user_message}",
-        extra={
-            "json_fields": {
-                **log_meta,
-                "message_len": len(message) if message else 0,
-                "is_user_message": is_user_message,
-            }
-        },
+        f"Streaming chat completion for session {session_id} for message {message} and user id {user_id}. Message is user message: {is_user_message}"
    )

    # Only fetch from Redis if session not provided (initial call)
    if session is None:
-        fetch_start = time.monotonic()
        session = await get_chat_session(session_id, user_id)
-        fetch_time = (time.monotonic() - fetch_start) * 1000
        logger.info(
-            f"[TIMING] get_chat_session took {fetch_time:.1f}ms, "
-            f"n_messages={len(session.messages) if session else 0}",
-            extra={
-                "json_fields": {
-                    **log_meta,
-                    "duration_ms": fetch_time,
-                    "n_messages": len(session.messages) if session else 0,
-                }
-            },
+            f"Fetched session from Redis: {session.session_id if session else 'None'}, "
+            f"message_count={len(session.messages) if session else 0}"
        )
    else:
        logger.info(
-            f"[TIMING] Using provided session, messages={len(session.messages)}",
-            extra={"json_fields": {**log_meta, "n_messages": len(session.messages)}},
+            f"Using provided session object: {session.session_id}, "
+            f"message_count={len(session.messages)}"
        )

    if not session:
@@ -426,16 +393,12 @@ async def stream_chat_completion(
            f"Session {session_id} not found. Please create a new session first."
        )

-    # Append the new message to the session if it's not already there
-    new_message_role = "user" if is_user_message else "assistant"
-    if message and (
-        len(session.messages) == 0
-        or not (
-            session.messages[-1].role == new_message_role
-            and session.messages[-1].content == message
+    if message:
+        session.messages.append(
+            ChatMessage(
+                role="user" if is_user_message else "assistant", content=message
+            )
        )
-    ):
-        session.messages.append(ChatMessage(role=new_message_role, content=message))
        logger.info(
            f"Appended message (role={'user' if is_user_message else 'assistant'}), "
            f"new message_count={len(session.messages)}"
@@ -443,32 +406,23 @@ async def stream_chat_completion(

        # Track user message in PostHog
        if is_user_message:
-            posthog_start = time.monotonic()
            track_user_message(
                user_id=user_id,
                session_id=session_id,
                message_length=len(message),
            )
-            posthog_time = (time.monotonic() - posthog_start) * 1000
-            logger.info(
-                f"[TIMING] track_user_message took {posthog_time:.1f}ms",
-                extra={"json_fields": {**log_meta, "duration_ms": posthog_time}},
-            )

-    upsert_start = time.monotonic()
-    session = await upsert_chat_session(session)
-    upsert_time = (time.monotonic() - upsert_start) * 1000
    logger.info(
-        f"[TIMING] upsert_chat_session took {upsert_time:.1f}ms",
-        extra={"json_fields": {**log_meta, "duration_ms": upsert_time}},
+        f"Upserting session: {session.session_id} with user id {session.user_id}, "
+        f"message_count={len(session.messages)}"
    )
+    session = await upsert_chat_session(session)
    assert session, "Session not found"

    # Generate title for new sessions on first user message (non-blocking)
    # Check: is_user_message, no title yet, and this is the first user message
-    user_messages = [m for m in session.messages if m.role == "user"]
-    first_user_msg = message or (user_messages[0].content if user_messages else None)
-    if is_user_message and first_user_msg and not session.title:
+    if is_user_message and message and not session.title:
+        user_messages = [m for m in session.messages if m.role == "user"]
        if len(user_messages) == 1:
            # First user message - generate title in background
            import asyncio
@@ -476,7 +430,7 @@ async def stream_chat_completion(
            # Capture only the values we need (not the session object) to avoid
            # stale data issues when the main flow modifies the session
            captured_session_id = session_id
-            captured_message = first_user_msg
+            captured_message = message
            captured_user_id = user_id

            async def _update_title():
@@ -500,13 +454,7 @@ async def stream_chat_completion(
            asyncio.create_task(_update_title())

    # Build system prompt with business understanding
-    prompt_start = time.monotonic()
    system_prompt, understanding = await _build_system_prompt(user_id)
-    prompt_time = (time.monotonic() - prompt_start) * 1000
-    logger.info(
-        f"[TIMING] _build_system_prompt took {prompt_time:.1f}ms",
-        extra={"json_fields": {**log_meta, "duration_ms": prompt_time}},
-    )

    # Initialize variables for streaming
    assistant_response = ChatMessage(
@@ -531,27 +479,13 @@ async def stream_chat_completion(
    # Generate unique IDs for AI SDK protocol
    import uuid as uuid_module

-    is_continuation = _continuation_message_id is not None
-    message_id = _continuation_message_id or str(uuid_module.uuid4())
+    message_id = str(uuid_module.uuid4())
    text_block_id = str(uuid_module.uuid4())

-    # Only yield message start for the initial call, not for continuations.
-    setup_time = (time.monotonic() - completion_start) * 1000
-    logger.info(
-        f"[TIMING] Setup complete, yielding StreamStart at {setup_time:.1f}ms",
-        extra={"json_fields": {**log_meta, "setup_time_ms": setup_time}},
-    )
-    if not is_continuation:
-        yield StreamStart(messageId=message_id, taskId=_task_id)
-
-    # Emit start-step before each LLM call (AI SDK uses this to add step boundaries)
-    yield StreamStartStep()
+    # Yield message start
+    yield StreamStart(messageId=message_id)

    try:
-        logger.info(
-            "[TIMING] Calling _stream_chat_chunks",
-            extra={"json_fields": log_meta},
-        )
        async for chunk in _stream_chat_chunks(
            session=session,
            tools=tools,
@@ -651,10 +585,6 @@ async def stream_chat_completion(
                    )
                yield chunk
            elif isinstance(chunk, StreamFinish):
-                if has_done_tool_call:
-                    # Tool calls happened — close the step but don't send message-level finish.
-                    # The continuation will open a new step, and finish will come at the end.
-                    yield StreamFinishStep()
                if not has_done_tool_call:
                    # Emit text-end before finish if we received text but haven't closed it
                    if has_received_text and not text_streaming_ended:
@@ -686,8 +616,6 @@ async def stream_chat_completion(
                            has_saved_assistant_message = True

                    has_yielded_end = True
-                    # Emit finish-step before finish (resets AI SDK text/reasoning state)
-                    yield StreamFinishStep()
                    yield chunk
            elif isinstance(chunk, StreamError):
                has_yielded_error = True
@@ -737,10 +665,6 @@ async def stream_chat_completion(
            logger.info(
                f"Retryable error encountered. Attempt {retry_count + 1}/{config.max_retries}"
            )
-            # Close the current step before retrying so the recursive call's
-            # StreamStartStep doesn't produce unbalanced step events.
-            if not has_yielded_end:
-                yield StreamFinishStep()
            should_retry = True
        else:
            # Non-retryable error or max retries exceeded
@@ -776,7 +700,6 @@ async def stream_chat_completion(
                error_response = StreamError(errorText=error_message)
                yield error_response
            if not has_yielded_end:
-                yield StreamFinishStep()
                yield StreamFinish()
            return

@@ -791,8 +714,6 @@ async def stream_chat_completion(
            retry_count=retry_count + 1,
            session=session,
            context=context,
-            _continuation_message_id=message_id,  # Reuse message ID since start was already sent
-            _task_id=_task_id,
        ):
            yield chunk
        return  # Exit after retry to avoid double-saving in finally block
@@ -808,13 +729,9 @@ async def stream_chat_completion(
        # Build the messages list in the correct order
        messages_to_save: list[ChatMessage] = []

-        # Add assistant message with tool_calls if any.
-        # Use extend (not assign) to preserve tool_calls already added by
-        # _yield_tool_call for long-running tools.
+        # Add assistant message with tool_calls if any
        if accumulated_tool_calls:
-            if not assistant_response.tool_calls:
-                assistant_response.tool_calls = []
-            assistant_response.tool_calls.extend(accumulated_tool_calls)
+            assistant_response.tool_calls = accumulated_tool_calls
            logger.info(
                f"Added {len(accumulated_tool_calls)} tool calls to assistant message"
            )
@@ -866,8 +783,6 @@ async def stream_chat_completion(
            session=session,  # Pass session object to avoid Redis refetch
            context=context,
            tool_call_response=str(tool_response_messages),
-            _continuation_message_id=message_id,  # Reuse message ID to avoid duplicates
-            _task_id=_task_id,
        ):
            yield chunk

@@ -978,21 +893,9 @@ async def _stream_chat_chunks(
        SSE formatted JSON response objects

    """
-    import time as time_module
-
-    stream_chunks_start = time_module.perf_counter()
    model = config.model

-    # Build log metadata for structured logging
-    log_meta = {"component": "ChatService", "session_id": session.session_id}
-    if session.user_id:
-        log_meta["user_id"] = session.user_id
-
-    logger.info(
-        f"[TIMING] _stream_chat_chunks STARTED, session={session.session_id}, "
-        f"user={session.user_id}, n_messages={len(session.messages)}",
-        extra={"json_fields": {**log_meta, "n_messages": len(session.messages)}},
-    )
+    logger.info("Starting pure chat stream")

    messages = session.to_openai_messages()
    if system_prompt:
@@ -1003,18 +906,12 @@ async def _stream_chat_chunks(
        messages = [system_message] + messages

    # Apply context window management
-    context_start = time_module.perf_counter()
    context_result = await _manage_context_window(
        messages=messages,
        model=model,
        api_key=config.api_key,
        base_url=config.base_url,
    )
-    context_time = (time_module.perf_counter() - context_start) * 1000
-    logger.info(
-        f"[TIMING] _manage_context_window took {context_time:.1f}ms",
-        extra={"json_fields": {**log_meta, "duration_ms": context_time}},
-    )

    if context_result.error:
        if "System prompt dropped" in context_result.error:
@@ -1049,19 +946,9 @@ async def _stream_chat_chunks(

        while retry_count <= MAX_RETRIES:
            try:
-                elapsed = (time_module.perf_counter() - stream_chunks_start) * 1000
-                retry_info = (
-                    f" (retry {retry_count}/{MAX_RETRIES})" if retry_count > 0 else ""
-                )
                logger.info(
-                    f"[TIMING] Creating OpenAI stream at {elapsed:.1f}ms{retry_info}",
-                    extra={
-                        "json_fields": {
-                            **log_meta,
-                            "elapsed_ms": elapsed,
-                            "retry_count": retry_count,
-                        }
-                    },
+                    f"Creating OpenAI chat completion stream..."
+                    f"{f' (retry {retry_count}/{MAX_RETRIES})' if retry_count > 0 else ''}"
                )

                # Build extra_body for OpenRouter tracing and PostHog analytics
@@ -1078,11 +965,6 @@ async def _stream_chat_chunks(
                        :128
                    ]  # OpenRouter limit

-                # Enable adaptive thinking for Anthropic models via OpenRouter
-                if config.thinking_enabled and "anthropic" in model.lower():
-                    extra_body["reasoning"] = {"enabled": True}
-
-                api_call_start = time_module.perf_counter()
                stream = await client.chat.completions.create(
                    model=model,
                    messages=cast(list[ChatCompletionMessageParam], messages),
@@ -1092,11 +974,6 @@ async def _stream_chat_chunks(
                    stream_options=ChatCompletionStreamOptionsParam(include_usage=True),
                    extra_body=extra_body,
                )
-                api_init_time = (time_module.perf_counter() - api_call_start) * 1000
-                logger.info(
-                    f"[TIMING] OpenAI stream object returned in {api_init_time:.1f}ms",
-                    extra={"json_fields": {**log_meta, "duration_ms": api_init_time}},
-                )

                # Variables to accumulate tool calls
                tool_calls: list[dict[str, Any]] = []
@@ -1107,13 +984,10 @@ async def _stream_chat_chunks(

                # Track if we've started the text block
                text_started = False
-                first_content_chunk = True
-                chunk_count = 0

                # Process the stream
                chunk: ChatCompletionChunk
                async for chunk in stream:
-                    chunk_count += 1
                    if chunk.usage:
                        yield StreamUsage(
                            promptTokens=chunk.usage.prompt_tokens,
@@ -1136,23 +1010,6 @@ async def _stream_chat_chunks(
                            if not text_started and text_block_id:
                                yield StreamTextStart(id=text_block_id)
                                text_started = True
-                            # Log timing for first content chunk
-                            if first_content_chunk:
-                                first_content_chunk = False
-                                ttfc = (
-                                    time_module.perf_counter() - api_call_start
-                                ) * 1000
-                                logger.info(
-                                    f"[TIMING] FIRST CONTENT CHUNK at {ttfc:.1f}ms "
-                                    f"(since API call), n_chunks={chunk_count}",
-                                    extra={
-                                        "json_fields": {
-                                            **log_meta,
-                                            "time_to_first_chunk_ms": ttfc,
-                                            "n_chunks": chunk_count,
-                                        }
-                                    },
-                                )
                            # Stream the text delta
                            text_response = StreamTextDelta(
                                id=text_block_id or "",
@@ -1209,21 +1066,7 @@ async def _stream_chat_chunks(
                                        toolName=tool_calls[idx]["function"]["name"],
                                    )
                                    emitted_start_for_idx.add(idx)
-                stream_duration = time_module.perf_counter() - api_call_start
-                logger.info(
-                    f"[TIMING] OpenAI stream COMPLETE, finish_reason={finish_reason}, "
-                    f"duration={stream_duration:.2f}s, "
-                    f"n_chunks={chunk_count}, n_tool_calls={len(tool_calls)}",
-                    extra={
-                        "json_fields": {
-                            **log_meta,
-                            "stream_duration_ms": stream_duration * 1000,
-                            "finish_reason": finish_reason,
-                            "n_chunks": chunk_count,
-                            "n_tool_calls": len(tool_calls),
-                        }
-                    },
-                )
+                logger.info(f"Stream complete. Finish reason: {finish_reason}")

                # Yield all accumulated tool calls after the stream is complete
                # This ensures all tool call arguments have been fully received
@@ -1243,17 +1086,10 @@ async def _stream_chat_chunks(
                        # Re-raise to trigger retry logic in the parent function
                        raise

-                total_time = (time_module.perf_counter() - stream_chunks_start) * 1000
-                logger.info(
-                    f"[TIMING] _stream_chat_chunks COMPLETED in {total_time / 1000:.1f}s; "
-                    f"session={session.session_id}, user={session.user_id}",
-                    extra={"json_fields": {**log_meta, "total_time_ms": total_time}},
-                )
                yield StreamFinish()
                return
            except Exception as e:
                last_error = e
-
                if _is_retryable_error(e) and retry_count < MAX_RETRIES:
                    retry_count += 1
                    # Calculate delay with exponential backoff
@@ -1269,27 +1105,12 @@ async def _stream_chat_chunks(
                    continue  # Retry the stream
                else:
                    # Non-retryable error or max retries exceeded
-                    _log_api_error(
-                        error=e,
-                        context="stream (not retrying)",
-                        session_id=session.session_id if session else None,
-                        message_count=len(messages) if messages else None,
-                        model=model,
-                        retry_count=retry_count,
+                    logger.error(
+                        f"Error in stream (not retrying): {e!s}",
+                        exc_info=True,
                    )
                    error_code = None
                    error_text = str(e)
-
-                    error_details = _extract_api_error_details(e)
-                    if error_details.get("response_body"):
-                        body = error_details["response_body"]
-                        if isinstance(body, dict):
-                            err = body.get("error")
-                            if isinstance(err, dict) and err.get("message"):
-                                error_text = err["message"]
-                            elif body.get("message"):
-                                error_text = body["message"]
-
                    if _is_region_blocked_error(e):
                        error_code = "MODEL_NOT_AVAILABLE_REGION"
                        error_text = (
@@ -1306,13 +1127,9 @@ async def _stream_chat_chunks(

        # If we exit the retry loop without returning, it means we exhausted retries
        if last_error:
-            _log_api_error(
-                error=last_error,
-                context=f"stream (max retries {MAX_RETRIES} exceeded)",
-                session_id=session.session_id if session else None,
-                message_count=len(messages) if messages else None,
-                model=model,
-                retry_count=MAX_RETRIES,
+            logger.error(
+                f"Max retries ({MAX_RETRIES}) exceeded. Last error: {last_error!s}",
+                exc_info=True,
            )
            yield StreamError(errorText=f"Max retries exceeded: {last_error!s}")
            yield StreamFinish()
@@ -1436,9 +1253,13 @@ async def _yield_tool_call(
                operation_id=operation_id,
            )

-            # Attach the tool_call to the current turn's assistant message
-            # (or create one if this is a tool-only response with no text).
-            session.add_tool_call_to_current_turn(tool_calls[yield_idx])
+            # Save assistant message with tool_call FIRST (required by LLM)
+            assistant_message = ChatMessage(
+                role="assistant",
+                content="",
+                tool_calls=[tool_calls[yield_idx]],
+            )
+            session.messages.append(assistant_message)

            # Then save pending tool result
            pending_message = ChatMessage(
@@ -1744,7 +1565,6 @@ async def _execute_long_running_tool_with_streaming(
            task_id,
            StreamError(errorText=str(e)),
        )
-        await stream_registry.publish_chunk(task_id, StreamFinishStep())
        await stream_registry.publish_chunk(task_id, StreamFinish())

        await _update_pending_operation(
@@ -1772,7 +1592,7 @@ async def _update_pending_operation(
    This is called by background tasks when long-running operations complete.
    """
    # Update the message in database
-    updated = await chat_db().update_tool_message_content(
+    updated = await chat_db.update_tool_message_content(
        session_id=session_id,
        tool_call_id=tool_call_id,
        new_content=result,
@@ -1861,10 +1681,6 @@ async def _generate_llm_continuation(
        if session_id:
            extra_body["session_id"] = session_id[:128]

-        # Enable adaptive thinking for Anthropic models via OpenRouter
-        if config.thinking_enabled and "anthropic" in config.model.lower():
-            extra_body["reasoning"] = {"enabled": True}
-
        retry_count = 0
        last_error: Exception | None = None
        response = None
@@ -1885,7 +1701,6 @@ async def _generate_llm_continuation(
                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(
@@ -1899,25 +1714,17 @@ async def _generate_llm_continuation(
                    await asyncio.sleep(delay)
                    continue
                else:
-                    # Non-retryable error - log details and exit gracefully
-                    _log_api_error(
-                        error=e,
-                        context="LLM continuation (not retrying)",
-                        session_id=session_id,
-                        message_count=len(messages) if messages else None,
-                        model=config.model,
-                        retry_count=retry_count,
+                    # Non-retryable error - log and exit gracefully
+                    logger.error(
+                        f"Non-retryable error in LLM continuation: {e!s}",
+                        exc_info=True,
                    )
                    return

        if last_error:
-            _log_api_error(
-                error=last_error,
-                context=f"LLM continuation (max retries {MAX_RETRIES} exceeded)",
-                session_id=session_id,
-                message_count=len(messages) if messages else None,
-                model=config.model,
-                retry_count=MAX_RETRIES,
+            logger.error(
+                f"Max retries ({MAX_RETRIES}) exceeded for LLM continuation. "
+                f"Last error: {last_error!s}"
            )
            return

@@ -1957,91 +1764,6 @@ async def _generate_llm_continuation(
        logger.error(f"Failed to generate LLM continuation: {e}", exc_info=True)


-def _log_api_error(
-    error: Exception,
-    context: str,
-    session_id: str | None = None,
-    message_count: int | None = None,
-    model: str | None = None,
-    retry_count: int = 0,
-) -> None:
-    """Log detailed API error information for debugging."""
-    details = _extract_api_error_details(error)
-    details["context"] = context
-    details["session_id"] = session_id
-    details["message_count"] = message_count
-    details["model"] = model
-    details["retry_count"] = retry_count
-
-    if isinstance(error, RateLimitError):
-        logger.warning(f"Rate limit error in {context}: {details}", exc_info=error)
-    elif isinstance(error, APIConnectionError):
-        logger.warning(f"API connection error in {context}: {details}", exc_info=error)
-    elif isinstance(error, APIStatusError) and error.status_code >= 500:
-        logger.error(f"API server error (5xx) in {context}: {details}", exc_info=error)
-    else:
-        logger.error(f"API error in {context}: {details}", exc_info=error)
-
-
-def _extract_api_error_details(error: Exception) -> dict[str, Any]:
-    """Extract detailed information from OpenAI/OpenRouter API errors."""
-    error_msg = str(error)
-    details: dict[str, Any] = {
-        "error_type": type(error).__name__,
-        "error_message": error_msg[:500] + "..." if len(error_msg) > 500 else error_msg,
-    }
-
-    if hasattr(error, "code"):
-        details["code"] = getattr(error, "code", None)
-    if hasattr(error, "param"):
-        details["param"] = getattr(error, "param", None)
-
-    if isinstance(error, APIStatusError):
-        details["status_code"] = error.status_code
-        details["request_id"] = getattr(error, "request_id", None)
-
-        if hasattr(error, "body") and error.body:
-            details["response_body"] = _sanitize_error_body(error.body)
-
-        if hasattr(error, "response") and error.response:
-            headers = error.response.headers
-            details["openrouter_provider"] = headers.get("x-openrouter-provider")
-            details["openrouter_model"] = headers.get("x-openrouter-model")
-            details["retry_after"] = headers.get("retry-after")
-            details["rate_limit_remaining"] = headers.get("x-ratelimit-remaining")
-
-    return details
-
-
-def _sanitize_error_body(
-    body: Any, max_length: int = 2000
-) -> dict[str, Any] | str | None:
-    """Extract only safe fields from error response body to avoid logging sensitive data."""
-    if not isinstance(body, dict):
-        # Non-dict bodies (e.g., HTML error pages) - return truncated string
-        if body is not None:
-            body_str = str(body)
-            if len(body_str) > max_length:
-                return body_str[:max_length] + "...[truncated]"
-            return body_str
-        return None
-
-    safe_fields = ("message", "type", "code", "param", "error")
-    sanitized: dict[str, Any] = {}
-
-    for field in safe_fields:
-        if field in body:
-            value = body[field]
-            if field == "error" and isinstance(value, dict):
-                sanitized[field] = _sanitize_error_body(value, max_length)
-            elif isinstance(value, str) and len(value) > max_length:
-                sanitized[field] = value[:max_length] + "...[truncated]"
-            else:
-                sanitized[field] = value
-
-    return sanitized if sanitized else None
-
-
 async def _generate_llm_continuation_with_streaming(
    session_id: str,
    user_id: str | None,
@@ -2089,10 +1811,6 @@ async def _generate_llm_continuation_with_streaming(
        if session_id:
            extra_body["session_id"] = session_id[:128]

-        # Enable adaptive thinking for Anthropic models via OpenRouter
-        if config.thinking_enabled and "anthropic" in config.model.lower():
-            extra_body["reasoning"] = {"enabled": True}
-
        # Make streaming LLM call (no tools - just text response)
        from typing import cast

@@ -2104,7 +1822,6 @@ async def _generate_llm_continuation_with_streaming(

        # Publish start event
        await stream_registry.publish_chunk(task_id, StreamStart(messageId=message_id))
-        await stream_registry.publish_chunk(task_id, StreamStartStep())
        await stream_registry.publish_chunk(task_id, StreamTextStart(id=text_block_id))

        # Stream the response
@@ -2128,7 +1845,6 @@ async def _generate_llm_continuation_with_streaming(

        # Publish end events
        await stream_registry.publish_chunk(task_id, StreamTextEnd(id=text_block_id))
-        await stream_registry.publish_chunk(task_id, StreamFinishStep())

        if assistant_content:
            # Reload session from DB to avoid race condition with user messages
@@ -2170,5 +1886,4 @@ async def _generate_llm_continuation_with_streaming(
            task_id,
            StreamError(errorText=f"Failed to generate response: {e}"),
        )
-        await stream_registry.publish_chunk(task_id, StreamFinishStep())
        await stream_registry.publish_chunk(task_id, StreamFinish())
--- a/autogpt_platform/backend/backend/api/features/chat/service_test.py
+++ b/autogpt_platform/backend/backend/api/features/chat/service_test.py
@@ -0,0 +1,82 @@
+import logging
+from os import getenv
+
+import pytest
+
+from . import service as chat_service
+from .model import create_chat_session, get_chat_session, upsert_chat_session
+from .response_model import (
+    StreamError,
+    StreamFinish,
+    StreamTextDelta,
+    StreamToolOutputAvailable,
+)
+
+logger = logging.getLogger(__name__)
+
+
+@pytest.mark.asyncio(loop_scope="session")
+async def test_stream_chat_completion(setup_test_user, test_user_id):
+    """
+    Test the stream_chat_completion function.
+    """
+    api_key: str | None = getenv("OPEN_ROUTER_API_KEY")
+    if not api_key:
+        return pytest.skip("OPEN_ROUTER_API_KEY is not set, skipping test")
+
+    session = await create_chat_session(test_user_id)
+
+    has_errors = False
+    has_ended = False
+    assistant_message = ""
+    async for chunk in chat_service.stream_chat_completion(
+        session.session_id, "Hello, how are you?", user_id=session.user_id
+    ):
+        logger.info(chunk)
+        if isinstance(chunk, StreamError):
+            has_errors = True
+        if isinstance(chunk, StreamTextDelta):
+            assistant_message += chunk.delta
+        if isinstance(chunk, StreamFinish):
+            has_ended = True
+
+    assert has_ended, "Chat completion did not end"
+    assert not has_errors, "Error occurred while streaming chat completion"
+    assert assistant_message, "Assistant message is empty"
+
+
+@pytest.mark.asyncio(loop_scope="session")
+async def test_stream_chat_completion_with_tool_calls(setup_test_user, test_user_id):
+    """
+    Test the stream_chat_completion function.
+    """
+    api_key: str | None = getenv("OPEN_ROUTER_API_KEY")
+    if not api_key:
+        return pytest.skip("OPEN_ROUTER_API_KEY is not set, skipping test")
+
+    session = await create_chat_session(test_user_id)
+    session = await upsert_chat_session(session)
+
+    has_errors = False
+    has_ended = False
+    had_tool_calls = False
+    async for chunk in chat_service.stream_chat_completion(
+        session.session_id,
+        "Please find me an agent that can help me with my business. Use the query 'moneny printing agent'",
+        user_id=session.user_id,
+    ):
+        logger.info(chunk)
+        if isinstance(chunk, StreamError):
+            has_errors = True
+
+        if isinstance(chunk, StreamFinish):
+            has_ended = True
+        if isinstance(chunk, StreamToolOutputAvailable):
+            had_tool_calls = True
+
+    assert has_ended, "Chat completion did not end"
+    assert not has_errors, "Error occurred while streaming chat completion"
+    assert had_tool_calls, "Tool calls did not occur"
+    session = await get_chat_session(session.session_id)
+    assert session, "Session not found"
+    assert session.usage, "Usage is empty"
--- a/autogpt_platform/backend/backend/api/features/chat/stream_registry.py
+++ b/autogpt_platform/backend/backend/api/features/chat/stream_registry.py
@@ -104,24 +104,6 @@ async def create_task(
    Returns:
        The created ActiveTask instance (metadata only)
    """
-    import time
-
-    start_time = time.perf_counter()
-
-    # Build log metadata for structured logging
-    log_meta = {
-        "component": "StreamRegistry",
-        "task_id": task_id,
-        "session_id": session_id,
-    }
-    if user_id:
-        log_meta["user_id"] = user_id
-
-    logger.info(
-        f"[TIMING] create_task STARTED, task={task_id}, session={session_id}, user={user_id}",
-        extra={"json_fields": log_meta},
-    )
-
    task = ActiveTask(
        task_id=task_id,
        session_id=session_id,
@@ -132,18 +114,10 @@ async def create_task(
    )

    # Store metadata in Redis
-    redis_start = time.perf_counter()
    redis = await get_redis_async()
-    redis_time = (time.perf_counter() - redis_start) * 1000
-    logger.info(
-        f"[TIMING] get_redis_async took {redis_time:.1f}ms",
-        extra={"json_fields": {**log_meta, "duration_ms": redis_time}},
-    )
-
    meta_key = _get_task_meta_key(task_id)
    op_key = _get_operation_mapping_key(operation_id)

-    hset_start = time.perf_counter()
    await redis.hset(  # type: ignore[misc]
        meta_key,
        mapping={
@@ -157,22 +131,12 @@ async def create_task(
            "created_at": task.created_at.isoformat(),
        },
    )
-    hset_time = (time.perf_counter() - hset_start) * 1000
-    logger.info(
-        f"[TIMING] redis.hset took {hset_time:.1f}ms",
-        extra={"json_fields": {**log_meta, "duration_ms": hset_time}},
-    )
-
    await redis.expire(meta_key, config.stream_ttl)

    # Create operation_id -> task_id mapping for webhook lookups
    await redis.set(op_key, task_id, ex=config.stream_ttl)

-    total_time = (time.perf_counter() - start_time) * 1000
-    logger.info(
-        f"[TIMING] create_task COMPLETED in {total_time:.1f}ms; task={task_id}, session={session_id}",
-        extra={"json_fields": {**log_meta, "total_time_ms": total_time}},
-    )
+    logger.debug(f"Created task {task_id} for session {session_id}")

    return task

@@ -192,60 +156,26 @@ async def publish_chunk(
    Returns:
        The Redis Stream message ID
    """
-    import time
-
-    start_time = time.perf_counter()
-    chunk_type = type(chunk).__name__
    chunk_json = chunk.model_dump_json()
    message_id = "0-0"

-    # Build log metadata
-    log_meta = {
-        "component": "StreamRegistry",
-        "task_id": task_id,
-        "chunk_type": chunk_type,
-    }
-
    try:
        redis = await get_redis_async()
        stream_key = _get_task_stream_key(task_id)

        # Write to Redis Stream for persistence and real-time delivery
-        xadd_start = time.perf_counter()
        raw_id = await redis.xadd(
            stream_key,
            {"data": chunk_json},
            maxlen=config.stream_max_length,
        )
-        xadd_time = (time.perf_counter() - xadd_start) * 1000
        message_id = raw_id if isinstance(raw_id, str) else raw_id.decode()

        # Set TTL on stream to match task metadata TTL
        await redis.expire(stream_key, config.stream_ttl)
-
-        total_time = (time.perf_counter() - start_time) * 1000
-        # Only log timing for significant chunks or slow operations
-        if (
-            chunk_type
-            in ("StreamStart", "StreamFinish", "StreamTextStart", "StreamTextEnd")
-            or total_time > 50
-        ):
-            logger.info(
-                f"[TIMING] publish_chunk {chunk_type} in {total_time:.1f}ms (xadd={xadd_time:.1f}ms)",
-                extra={
-                    "json_fields": {
-                        **log_meta,
-                        "total_time_ms": total_time,
-                        "xadd_time_ms": xadd_time,
-                        "message_id": message_id,
-                    }
-                },
-            )
    except Exception as e:
-        elapsed = (time.perf_counter() - start_time) * 1000
        logger.error(
-            f"[TIMING] Failed to publish chunk {chunk_type} after {elapsed:.1f}ms: {e}",
-            extra={"json_fields": {**log_meta, "elapsed_ms": elapsed, "error": str(e)}},
+            f"Failed to publish chunk for task {task_id}: {e}",
            exc_info=True,
        )

@@ -270,61 +200,24 @@ async def subscribe_to_task(
        An asyncio Queue that will receive stream chunks, or None if task not found
        or user doesn't have access
    """
-    import time
-
-    start_time = time.perf_counter()
-
-    # Build log metadata
-    log_meta = {"component": "StreamRegistry", "task_id": task_id}
-    if user_id:
-        log_meta["user_id"] = user_id
-
-    logger.info(
-        f"[TIMING] subscribe_to_task STARTED, task={task_id}, user={user_id}, last_msg={last_message_id}",
-        extra={"json_fields": {**log_meta, "last_message_id": last_message_id}},
-    )
-
-    redis_start = time.perf_counter()
    redis = await get_redis_async()
    meta_key = _get_task_meta_key(task_id)
    meta: dict[Any, Any] = await redis.hgetall(meta_key)  # type: ignore[misc]
-    hgetall_time = (time.perf_counter() - redis_start) * 1000
-    logger.info(
-        f"[TIMING] Redis hgetall took {hgetall_time:.1f}ms",
-        extra={"json_fields": {**log_meta, "duration_ms": hgetall_time}},
-    )

    if not meta:
-        elapsed = (time.perf_counter() - start_time) * 1000
-        logger.info(
-            f"[TIMING] Task not found in Redis after {elapsed:.1f}ms",
-            extra={
-                "json_fields": {
-                    **log_meta,
-                    "elapsed_ms": elapsed,
-                    "reason": "task_not_found",
-                }
-            },
-        )
+        logger.debug(f"Task {task_id} not found in Redis")
        return None

    # Note: Redis client uses decode_responses=True, so keys are strings
    task_status = meta.get("status", "")
    task_user_id = meta.get("user_id", "") or None
-    log_meta["session_id"] = meta.get("session_id", "")

    # Validate ownership - if task has an owner, requester must match
    if task_user_id:
        if user_id != task_user_id:
            logger.warning(
-                f"[TIMING] Access denied: user {user_id} tried to access task owned by {task_user_id}",
-                extra={
-                    "json_fields": {
-                        **log_meta,
-                        "task_owner": task_user_id,
-                        "reason": "access_denied",
-                    }
-                },
+                f"User {user_id} denied access to task {task_id} "
+                f"owned by {task_user_id}"
            )
            return None

@@ -332,19 +225,7 @@ async def subscribe_to_task(
    stream_key = _get_task_stream_key(task_id)

    # Step 1: Replay messages from Redis Stream
-    xread_start = time.perf_counter()
    messages = await redis.xread({stream_key: last_message_id}, block=0, count=1000)
-    xread_time = (time.perf_counter() - xread_start) * 1000
-    logger.info(
-        f"[TIMING] Redis xread (replay) took {xread_time:.1f}ms, status={task_status}",
-        extra={
-            "json_fields": {
-                **log_meta,
-                "duration_ms": xread_time,
-                "task_status": task_status,
-            }
-        },
-    )

    replayed_count = 0
    replay_last_id = last_message_id
@@ -363,48 +244,19 @@ async def subscribe_to_task(
                    except Exception as e:
                        logger.warning(f"Failed to replay message: {e}")

-    logger.info(
-        f"[TIMING] Replayed {replayed_count} messages, last_id={replay_last_id}",
-        extra={
-            "json_fields": {
-                **log_meta,
-                "n_messages_replayed": replayed_count,
-                "replay_last_id": replay_last_id,
-            }
-        },
-    )
+    logger.debug(f"Task {task_id}: replayed {replayed_count} messages")

    # Step 2: If task is still running, start stream listener for live updates
    if task_status == "running":
-        logger.info(
-            "[TIMING] Task still running, starting _stream_listener",
-            extra={"json_fields": {**log_meta, "task_status": task_status}},
-        )
        listener_task = asyncio.create_task(
-            _stream_listener(task_id, subscriber_queue, replay_last_id, log_meta)
+            _stream_listener(task_id, subscriber_queue, replay_last_id)
        )
        # Track listener task for cleanup on unsubscribe
        _listener_tasks[id(subscriber_queue)] = (task_id, listener_task)
    else:
        # Task is completed/failed - add finish marker
-        logger.info(
-            f"[TIMING] Task already {task_status}, adding StreamFinish",
-            extra={"json_fields": {**log_meta, "task_status": task_status}},
-        )
        await subscriber_queue.put(StreamFinish())

-    total_time = (time.perf_counter() - start_time) * 1000
-    logger.info(
-        f"[TIMING] subscribe_to_task COMPLETED in {total_time:.1f}ms; task={task_id}, "
-        f"n_messages_replayed={replayed_count}",
-        extra={
-            "json_fields": {
-                **log_meta,
-                "total_time_ms": total_time,
-                "n_messages_replayed": replayed_count,
-            }
-        },
-    )
    return subscriber_queue


@@ -412,7 +264,6 @@ async def _stream_listener(
    task_id: str,
    subscriber_queue: asyncio.Queue[StreamBaseResponse],
    last_replayed_id: str,
-    log_meta: dict | None = None,
 ) -> None:
    """Listen to Redis Stream for new messages using blocking XREAD.

@@ -423,27 +274,10 @@ async def _stream_listener(
        task_id: Task ID to listen for
        subscriber_queue: Queue to deliver messages to
        last_replayed_id: Last message ID from replay (continue from here)
-        log_meta: Structured logging metadata
    """
-    import time
-
-    start_time = time.perf_counter()
-
-    # Use provided log_meta or build minimal one
-    if log_meta is None:
-        log_meta = {"component": "StreamRegistry", "task_id": task_id}
-
-    logger.info(
-        f"[TIMING] _stream_listener STARTED, task={task_id}, last_id={last_replayed_id}",
-        extra={"json_fields": {**log_meta, "last_replayed_id": last_replayed_id}},
-    )
-
    queue_id = id(subscriber_queue)
    # Track the last successfully delivered message ID for recovery hints
    last_delivered_id = last_replayed_id
-    messages_delivered = 0
-    first_message_time = None
-    xread_count = 0

    try:
        redis = await get_redis_async()
@@ -453,39 +287,9 @@ async def _stream_listener(
        while True:
            # Block for up to 30 seconds waiting for new messages
            # This allows periodic checking if task is still running
-            xread_start = time.perf_counter()
-            xread_count += 1
            messages = await redis.xread(
                {stream_key: current_id}, block=30000, count=100
            )
-            xread_time = (time.perf_counter() - xread_start) * 1000
-
-            if messages:
-                msg_count = sum(len(msgs) for _, msgs in messages)
-                logger.info(
-                    f"[TIMING] xread #{xread_count} returned {msg_count} messages in {xread_time:.1f}ms",
-                    extra={
-                        "json_fields": {
-                            **log_meta,
-                            "xread_count": xread_count,
-                            "n_messages": msg_count,
-                            "duration_ms": xread_time,
-                        }
-                    },
-                )
-            elif xread_time > 1000:
-                # Only log timeouts (30s blocking)
-                logger.info(
-                    f"[TIMING] xread #{xread_count} timeout after {xread_time:.1f}ms",
-                    extra={
-                        "json_fields": {
-                            **log_meta,
-                            "xread_count": xread_count,
-                            "duration_ms": xread_time,
-                            "reason": "timeout",
-                        }
-                    },
-                )

            if not messages:
                # Timeout - check if task is still running
@@ -522,30 +326,10 @@ async def _stream_listener(
                                )
                                # Update last delivered ID on successful delivery
                                last_delivered_id = current_id
-                                messages_delivered += 1
-                                if first_message_time is None:
-                                    first_message_time = time.perf_counter()
-                                    elapsed = (first_message_time - start_time) * 1000
-                                    logger.info(
-                                        f"[TIMING] FIRST live message at {elapsed:.1f}ms, type={type(chunk).__name__}",
-                                        extra={
-                                            "json_fields": {
-                                                **log_meta,
-                                                "elapsed_ms": elapsed,
-                                                "chunk_type": type(chunk).__name__,
-                                            }
-                                        },
-                                    )
                            except asyncio.TimeoutError:
                                logger.warning(
-                                    f"[TIMING] Subscriber queue full, delivery timed out after {QUEUE_PUT_TIMEOUT}s",
-                                    extra={
-                                        "json_fields": {
-                                            **log_meta,
-                                            "timeout_s": QUEUE_PUT_TIMEOUT,
-                                            "reason": "queue_full",
-                                        }
-                                    },
+                                    f"Subscriber queue full for task {task_id}, "
+                                    f"message delivery timed out after {QUEUE_PUT_TIMEOUT}s"
                                )
                                # Send overflow error with recovery info
                                try:
@@ -567,44 +351,15 @@ async def _stream_listener(

                            # Stop listening on finish
                            if isinstance(chunk, StreamFinish):
-                                total_time = (time.perf_counter() - start_time) * 1000
-                                logger.info(
-                                    f"[TIMING] StreamFinish received in {total_time/1000:.1f}s; delivered={messages_delivered}",
-                                    extra={
-                                        "json_fields": {
-                                            **log_meta,
-                                            "total_time_ms": total_time,
-                                            "messages_delivered": messages_delivered,
-                                        }
-                                    },
-                                )
                                return
                    except Exception as e:
-                        logger.warning(
-                            f"Error processing stream message: {e}",
-                            extra={"json_fields": {**log_meta, "error": str(e)}},
-                        )
+                        logger.warning(f"Error processing stream message: {e}")

    except asyncio.CancelledError:
-        elapsed = (time.perf_counter() - start_time) * 1000
-        logger.info(
-            f"[TIMING] _stream_listener CANCELLED after {elapsed:.1f}ms, delivered={messages_delivered}",
-            extra={
-                "json_fields": {
-                    **log_meta,
-                    "elapsed_ms": elapsed,
-                    "messages_delivered": messages_delivered,
-                    "reason": "cancelled",
-                }
-            },
-        )
+        logger.debug(f"Stream listener cancelled for task {task_id}")
        raise  # Re-raise to propagate cancellation
    except Exception as e:
-        elapsed = (time.perf_counter() - start_time) * 1000
-        logger.error(
-            f"[TIMING] _stream_listener ERROR after {elapsed:.1f}ms: {e}",
-            extra={"json_fields": {**log_meta, "elapsed_ms": elapsed, "error": str(e)}},
-        )
+        logger.error(f"Stream listener error for task {task_id}: {e}")
        # On error, send finish to unblock subscriber
        try:
            await asyncio.wait_for(
@@ -613,24 +368,10 @@ async def _stream_listener(
            )
        except (asyncio.TimeoutError, asyncio.QueueFull):
            logger.warning(
-                "Could not deliver finish event after error",
-                extra={"json_fields": log_meta},
+                f"Could not deliver finish event for task {task_id} after error"
            )
    finally:
        # Clean up listener task mapping on exit
-        total_time = (time.perf_counter() - start_time) * 1000
-        logger.info(
-            f"[TIMING] _stream_listener FINISHED in {total_time/1000:.1f}s; task={task_id}, "
-            f"delivered={messages_delivered}, xread_count={xread_count}",
-            extra={
-                "json_fields": {
-                    **log_meta,
-                    "total_time_ms": total_time,
-                    "messages_delivered": messages_delivered,
-                    "xread_count": xread_count,
-                }
-            },
-        )
        _listener_tasks.pop(queue_id, None)


@@ -814,28 +555,6 @@ async def get_active_task_for_session(
                if task_user_id and user_id != task_user_id:
                    continue

-                # Auto-expire stale tasks that exceeded stream_timeout
-                created_at_str = meta.get("created_at", "")
-                if created_at_str:
-                    try:
-                        created_at = datetime.fromisoformat(created_at_str)
-                        age_seconds = (
-                            datetime.now(timezone.utc) - created_at
-                        ).total_seconds()
-                        if age_seconds > config.stream_timeout:
-                            logger.warning(
-                                f"[TASK_LOOKUP] Auto-expiring stale task {task_id[:8]}... "
-                                f"(age={age_seconds:.0f}s > timeout={config.stream_timeout}s)"
-                            )
-                            await mark_task_completed(task_id, "failed")
-                            continue
-                    except (ValueError, TypeError):
-                        pass
-
-                logger.info(
-                    f"[TASK_LOOKUP] Found running task {task_id[:8]}... for session {session_id[:8]}..."
-                )
-
                # Get the last message ID from Redis Stream
                stream_key = _get_task_stream_key(task_id)
                last_id = "0-0"
@@ -879,10 +598,8 @@ def _reconstruct_chunk(chunk_data: dict) -> StreamBaseResponse | None:
        ResponseType,
        StreamError,
        StreamFinish,
-        StreamFinishStep,
        StreamHeartbeat,
        StreamStart,
-        StreamStartStep,
        StreamTextDelta,
        StreamTextEnd,
        StreamTextStart,
@@ -896,8 +613,6 @@ def _reconstruct_chunk(chunk_data: dict) -> StreamBaseResponse | None:
    type_to_class: dict[str, type[StreamBaseResponse]] = {
        ResponseType.START.value: StreamStart,
        ResponseType.FINISH.value: StreamFinish,
-        ResponseType.START_STEP.value: StreamStartStep,
-        ResponseType.FINISH_STEP.value: StreamFinishStep,
        ResponseType.TEXT_START.value: StreamTextStart,
        ResponseType.TEXT_DELTA.value: StreamTextDelta,
        ResponseType.TEXT_END.value: StreamTextEnd,
--- a/autogpt_platform/backend/backend/api/features/chat/tools/IDEAS.md
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/IDEAS.md
--- a/autogpt_platform/backend/backend/api/features/chat/tools/init.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/init.py
@@ -3,18 +3,15 @@ from typing import TYPE_CHECKING, Any

 from openai.types.chat import ChatCompletionToolParam

-from backend.copilot.model import ChatSession
-from backend.copilot.tracking import track_tool_called
+from backend.api.features.chat.model import ChatSession
+from backend.api.features.chat.tracking import track_tool_called

 from .add_understanding import AddUnderstandingTool
 from .agent_output import AgentOutputTool
 from .base import BaseTool
-from .bash_exec import BashExecTool
-from .check_operation_status import CheckOperationStatusTool
 from .create_agent import CreateAgentTool
 from .customize_agent import CustomizeAgentTool
 from .edit_agent import EditAgentTool
-from .feature_requests import CreateFeatureRequestTool, SearchFeatureRequestsTool
 from .find_agent import FindAgentTool
 from .find_block import FindBlockTool
 from .find_library_agent import FindLibraryAgentTool
@@ -22,7 +19,6 @@ from .get_doc_page import GetDocPageTool
 from .run_agent import RunAgentTool
 from .run_block import RunBlockTool
 from .search_docs import SearchDocsTool
-from .web_fetch import WebFetchTool
 from .workspace_files import (
    DeleteWorkspaceFileTool,
    ListWorkspaceFilesTool,
@@ -31,7 +27,7 @@ from .workspace_files import (
 )

 if TYPE_CHECKING:
-    from backend.copilot.response_model import StreamToolOutputAvailable
+    from backend.api.features.chat.response_model import StreamToolOutputAvailable

 logger = logging.getLogger(__name__)

@@ -47,17 +43,8 @@ TOOL_REGISTRY: dict[str, BaseTool] = {
    "run_agent": RunAgentTool(),
    "run_block": RunBlockTool(),
    "view_agent_output": AgentOutputTool(),
-    "check_operation_status": CheckOperationStatusTool(),
    "search_docs": SearchDocsTool(),
    "get_doc_page": GetDocPageTool(),
-    # Web fetch for safe URL retrieval
-    "web_fetch": WebFetchTool(),
-    # Sandboxed code execution (bubblewrap)
-    "bash_exec": BashExecTool(),
-    # Persistent workspace tools (cloud storage, survives across sessions)
-    # Feature request tools
-    "search_feature_requests": SearchFeatureRequestsTool(),
-    "create_feature_request": CreateFeatureRequestTool(),
    # Workspace tools for CoPilot file operations
    "list_workspace_files": ListWorkspaceFilesTool(),
    "read_workspace_file": ReadWorkspaceFileTool(),
--- a/autogpt_platform/backend/backend/api/features/chat/tools/_test_data.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/_test_data.py
@@ -6,11 +6,11 @@ import pytest
 from prisma.types import ProfileCreateInput
 from pydantic import SecretStr

+from backend.api.features.chat.model import ChatSession
 from backend.api.features.store import db as store_db
 from backend.blocks.firecrawl.scrape import FirecrawlScrapeBlock
 from backend.blocks.io import AgentInputBlock, AgentOutputBlock
 from backend.blocks.llm import AITextGeneratorBlock
-from backend.copilot.model import ChatSession
 from backend.data.db import prisma
 from backend.data.graph import Graph, Link, Node, create_graph
 from backend.data.model import APIKeyCredentials
--- a/autogpt_platform/backend/backend/api/features/chat/tools/add_understanding.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/add_understanding.py
@@ -3,9 +3,11 @@
 import logging
 from typing import Any

-from backend.copilot.model import ChatSession
-from backend.data.db_accessors import understanding_db
-from backend.data.understanding import BusinessUnderstandingInput
+from backend.api.features.chat.model import ChatSession
+from backend.data.understanding import (
+    BusinessUnderstandingInput,
+    upsert_business_understanding,
+)

 from .base import BaseTool
 from .models import ErrorResponse, ToolResponseBase, UnderstandingUpdatedResponse
@@ -97,9 +99,7 @@ and automations for the user's specific needs."""
        ]

        # Upsert with merge
-        understanding = await understanding_db().upsert_business_understanding(
-            user_id, input_data
-        )
+        understanding = await upsert_business_understanding(user_id, input_data)

        # Build current understanding summary (filter out empty values)
        current_understanding = {
--- 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
--- 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
@@ -5,8 +5,9 @@ import re
 import uuid
 from typing import Any, NotRequired, TypedDict

-from backend.data.db_accessors import graph_db, library_db, store_db
-from backend.data.graph import Graph, Link, Node
+from backend.api.features.library import db as library_db
+from backend.api.features.store import db as store_db
+from backend.data.graph import Graph, Link, Node, get_graph, get_store_listed_graphs
 from backend.util.exceptions import DatabaseError, NotFoundError

 from .service import (
@@ -144,9 +145,8 @@ async def get_library_agent_by_id(
    Returns:
        LibraryAgentSummary if found, None otherwise
    """
-    db = library_db()
    try:
-        agent = await db.get_library_agent_by_graph_id(user_id, agent_id)
+        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(
@@ -163,7 +163,7 @@ async def get_library_agent_by_id(
        logger.debug(f"Could not fetch library agent by graph_id {agent_id}: {e}")

    try:
-        agent = await db.get_library_agent(agent_id, user_id)
+        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(
@@ -215,7 +215,7 @@ async def get_library_agents_for_generation(
        List of LibraryAgentSummary with schemas and recent executions for sub-agent composition
    """
    try:
-        response = await library_db().list_library_agents(
+        response = await library_db.list_library_agents(
            user_id=user_id,
            search_term=search_query,
            page=1,
@@ -272,7 +272,7 @@ async def search_marketplace_agents_for_generation(
        List of LibraryAgentSummary with full input/output schemas
    """
    try:
-        response = await store_db().get_store_agents(
+        response = await store_db.get_store_agents(
            search_query=search_query,
            page=1,
            page_size=max_results,
@@ -286,7 +286,7 @@ async def search_marketplace_agents_for_generation(
            return []

        graph_ids = [agent.agent_graph_id for agent in agents_with_graphs]
-        graphs = await graph_db().get_store_listed_graphs(graph_ids)
+        graphs = await get_store_listed_graphs(*graph_ids)

        results: list[LibraryAgentSummary] = []
        for agent in agents_with_graphs:
@@ -673,10 +673,9 @@ async def save_agent_to_library(
        Tuple of (created Graph, LibraryAgent)
    """
    graph = json_to_graph(agent_json)
-    db = library_db()
    if is_update:
-        return await db.update_graph_in_library(graph, user_id)
-    return await db.create_graph_in_library(graph, user_id)
+        return await library_db.update_graph_in_library(graph, user_id)
+    return await library_db.create_graph_in_library(graph, user_id)


 def graph_to_json(graph: Graph) -> dict[str, Any]:
@@ -736,14 +735,12 @@ async def get_agent_as_json(
    Returns:
        Agent as JSON dict or None if not found
    """
-    db = graph_db()
-
-    graph = await db.get_graph(agent_id, version=None, user_id=user_id)
+    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 db.get_graph(
+            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:
--- 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
--- 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
@@ -12,19 +12,8 @@ import httpx

 from backend.util.settings import Settings

-from .dummy import (
-    customize_template_dummy,
-    decompose_goal_dummy,
-    generate_agent_dummy,
-    generate_agent_patch_dummy,
-    get_blocks_dummy,
-    health_check_dummy,
-)
-
 logger = logging.getLogger(__name__)

-_dummy_mode_warned = False
-

 def _create_error_response(
    error_message: str,
@@ -101,26 +90,10 @@ def _get_settings() -> Settings:
    return _settings


-def _is_dummy_mode() -> bool:
-    """Check if dummy mode is enabled for testing."""
-    global _dummy_mode_warned
-    settings = _get_settings()
-    is_dummy = bool(settings.config.agentgenerator_use_dummy)
-    if is_dummy and not _dummy_mode_warned:
-        logger.warning(
-            "Agent Generator running in DUMMY MODE - returning mock responses. "
-            "Do not use in production!"
-        )
-        _dummy_mode_warned = True
-    return is_dummy
-
-
 def is_external_service_configured() -> bool:
-    """Check if external Agent Generator service is configured (or dummy mode)."""
+    """Check if external Agent Generator service is configured."""
    settings = _get_settings()
-    return bool(settings.config.agentgenerator_host) or bool(
-        settings.config.agentgenerator_use_dummy
-    )
+    return bool(settings.config.agentgenerator_host)


 def _get_base_url() -> str:
@@ -164,9 +137,6 @@ async def decompose_goal_external(
        - {"type": "error", "error": "...", "error_type": "..."} on error
        Or None on unexpected error
    """
-    if _is_dummy_mode():
-        return await decompose_goal_dummy(description, context, library_agents)
-
    client = _get_client()

    if context:
@@ -256,11 +226,6 @@ async def generate_agent_external(
    Returns:
        Agent JSON dict, {"status": "accepted"} for async, or error dict {"type": "error", ...} on error
    """
-    if _is_dummy_mode():
-        return await generate_agent_dummy(
-            instructions, library_agents, operation_id, task_id
-        )
-
    client = _get_client()

    # Build request payload
@@ -332,11 +297,6 @@ async def generate_agent_patch_external(
    Returns:
        Updated agent JSON, clarifying questions dict, {"status": "accepted"} for async, or error dict on error
    """
-    if _is_dummy_mode():
-        return await generate_agent_patch_dummy(
-            update_request, current_agent, library_agents, operation_id, task_id
-        )
-
    client = _get_client()

    # Build request payload
@@ -423,11 +383,6 @@ async def customize_template_external(
    Returns:
        Customized agent JSON, clarifying questions dict, or error dict on error
    """
-    if _is_dummy_mode():
-        return await customize_template_dummy(
-            template_agent, modification_request, context
-        )
-
    client = _get_client()

    request = modification_request
@@ -490,9 +445,6 @@ async def get_blocks_external() -> list[dict[str, Any]] | None:
    Returns:
        List of block info dicts or None on error
    """
-    if _is_dummy_mode():
-        return await get_blocks_dummy()
-
    client = _get_client()

    try:
@@ -526,9 +478,6 @@ async def health_check() -> bool:
    if not is_external_service_configured():
        return False

-    if _is_dummy_mode():
-        return await health_check_dummy()
-
    client = _get_client()

    try:
--- a/autogpt_platform/backend/backend/api/features/chat/tools/agent_output.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/agent_output.py
@@ -5,15 +5,15 @@ import re
 from datetime import datetime, timedelta, timezone
 from typing import Any

-from pydantic import BaseModel, Field, field_validator
+from pydantic import BaseModel, field_validator

+from backend.api.features.chat.model import ChatSession
+from backend.api.features.library import db as library_db
 from backend.api.features.library.model import LibraryAgent
-from backend.copilot.model import ChatSession
-from backend.data.db_accessors import execution_db, library_db
+from backend.data import execution as execution_db
 from backend.data.execution import ExecutionStatus, GraphExecution, GraphExecutionMeta

 from .base import BaseTool
-from .execution_utils import TERMINAL_STATUSES, wait_for_execution
 from .models import (
    AgentOutputResponse,
    ErrorResponse,
@@ -34,7 +34,6 @@ class AgentOutputInput(BaseModel):
    store_slug: str = ""
    execution_id: str = ""
    run_time: str = "latest"
-    wait_if_running: int = Field(default=0, ge=0, le=300)

    @field_validator(
        "agent_name",
@@ -118,11 +117,6 @@ class AgentOutputTool(BaseTool):
        Select which run to retrieve using:
        - execution_id: Specific execution ID
        - run_time: 'latest' (default), 'yesterday', 'last week', or ISO date 'YYYY-MM-DD'
-
-        Wait for completion (optional):
-        - wait_if_running: Max seconds to wait if execution is still running (0-300).
-          If the execution is running/queued, waits up to this many seconds for completion.
-          Returns current status on timeout. If already finished, returns immediately.
        """

    @property
@@ -152,13 +146,6 @@ class AgentOutputTool(BaseTool):
                        "Time filter: 'latest', 'yesterday', 'last week', or 'YYYY-MM-DD'"
                    ),
                },
-                "wait_if_running": {
-                    "type": "integer",
-                    "description": (
-                        "Max seconds to wait if execution is still running (0-300). "
-                        "If running, waits for completion. Returns current state on timeout."
-                    ),
-                },
            },
            "required": [],
        }
@@ -178,12 +165,10 @@ class AgentOutputTool(BaseTool):
        Resolve agent from provided identifiers.
        Returns (library_agent, error_message).
        """
-        lib_db = library_db()
-
        # Priority 1: Exact library agent ID
        if library_agent_id:
            try:
-                agent = await lib_db.get_library_agent(library_agent_id, user_id)
+                agent = await library_db.get_library_agent(library_agent_id, user_id)
                return agent, None
            except Exception as e:
                logger.warning(f"Failed to get library agent by ID: {e}")
@@ -197,7 +182,7 @@ class AgentOutputTool(BaseTool):
                return None, f"Agent '{store_slug}' not found in marketplace"

            # Find in user's library by graph_id
-            agent = await lib_db.get_library_agent_by_graph_id(user_id, graph.id)
+            agent = await library_db.get_library_agent_by_graph_id(user_id, graph.id)
            if not agent:
                return (
                    None,
@@ -209,7 +194,7 @@ class AgentOutputTool(BaseTool):
        # Priority 3: Fuzzy name search in library
        if agent_name:
            try:
-                response = await lib_db.list_library_agents(
+                response = await library_db.list_library_agents(
                    user_id=user_id,
                    search_term=agent_name,
                    page_size=5,
@@ -238,20 +223,14 @@ class AgentOutputTool(BaseTool):
        execution_id: str | None,
        time_start: datetime | None,
        time_end: datetime | None,
-        include_running: bool = False,
    ) -> tuple[GraphExecution | None, list[GraphExecutionMeta], str | None]:
        """
        Fetch execution(s) based on filters.
        Returns (single_execution, available_executions_meta, error_message).
-
-        Args:
-            include_running: If True, also look for running/queued executions (for waiting)
        """
-        exec_db = execution_db()
-
        # If specific execution_id provided, fetch it directly
        if execution_id:
-            execution = await exec_db.get_graph_execution(
+            execution = await execution_db.get_graph_execution(
                user_id=user_id,
                execution_id=execution_id,
                include_node_executions=False,
@@ -260,22 +239,11 @@ class AgentOutputTool(BaseTool):
                return None, [], f"Execution '{execution_id}' not found"
            return execution, [], None

-        # Determine which statuses to query
-        statuses = [ExecutionStatus.COMPLETED]
-        if include_running:
-            statuses.extend(
-                [
-                    ExecutionStatus.RUNNING,
-                    ExecutionStatus.QUEUED,
-                    ExecutionStatus.INCOMPLETE,
-                ]
-            )
-
-        # Get executions with time filters
-        executions = await exec_db.get_graph_executions(
+        # Get completed executions with time filters
+        executions = await execution_db.get_graph_executions(
            graph_id=graph_id,
            user_id=user_id,
-            statuses=statuses,
+            statuses=[ExecutionStatus.COMPLETED],
            created_time_gte=time_start,
            created_time_lte=time_end,
            limit=10,
@@ -286,7 +254,7 @@ class AgentOutputTool(BaseTool):

        # If only one execution, fetch full details
        if len(executions) == 1:
-            full_execution = await exec_db.get_graph_execution(
+            full_execution = await execution_db.get_graph_execution(
                user_id=user_id,
                execution_id=executions[0].id,
                include_node_executions=False,
@@ -294,7 +262,7 @@ class AgentOutputTool(BaseTool):
            return full_execution, [], None

        # Multiple executions - return latest with full details, plus list of available
-        full_execution = await exec_db.get_graph_execution(
+        full_execution = await execution_db.get_graph_execution(
            user_id=user_id,
            execution_id=executions[0].id,
            include_node_executions=False,
@@ -342,28 +310,10 @@ class AgentOutputTool(BaseTool):
                for e in available_executions[:5]
            ]

-        # Build appropriate message based on execution status
-        if execution.status == ExecutionStatus.COMPLETED:
-            message = f"Found execution outputs for agent '{agent.name}'"
-        elif execution.status == ExecutionStatus.FAILED:
-            message = f"Execution for agent '{agent.name}' failed"
-        elif execution.status == ExecutionStatus.TERMINATED:
-            message = f"Execution for agent '{agent.name}' was terminated"
-        elif execution.status in (
-            ExecutionStatus.RUNNING,
-            ExecutionStatus.QUEUED,
-            ExecutionStatus.INCOMPLETE,
-        ):
-            message = (
-                f"Execution for agent '{agent.name}' is still {execution.status.value}. "
-                "Results may be incomplete. Use wait_if_running to wait for completion."
-            )
-        else:
-            message = f"Found execution for agent '{agent.name}' (status: {execution.status.value})"
-
+        message = f"Found execution outputs for agent '{agent.name}'"
        if len(available_executions) > 1:
            message += (
-                f" Showing latest of {len(available_executions)} matching executions."
+                f". Showing latest of {len(available_executions)} matching executions."
            )

        return AgentOutputResponse(
@@ -430,7 +380,7 @@ class AgentOutputTool(BaseTool):
            and not input_data.store_slug
        ):
            # Fetch execution directly to get graph_id
-            execution = await execution_db().get_graph_execution(
+            execution = await execution_db.get_graph_execution(
                user_id=user_id,
                execution_id=input_data.execution_id,
                include_node_executions=False,
@@ -442,7 +392,7 @@ class AgentOutputTool(BaseTool):
                )

            # Find library agent by graph_id
-            agent = await library_db().get_library_agent_by_graph_id(
+            agent = await library_db.get_library_agent_by_graph_id(
                user_id, execution.graph_id
            )
            if not agent:
@@ -478,17 +428,13 @@ class AgentOutputTool(BaseTool):
        # Parse time expression
        time_start, time_end = parse_time_expression(input_data.run_time)

-        # Check if we should wait for running executions
-        wait_timeout = input_data.wait_if_running
-
-        # Fetch execution(s) - include running if we're going to wait
+        # Fetch execution(s)
        execution, available_executions, exec_error = await self._get_execution(
            user_id=user_id,
            graph_id=agent.graph_id,
            execution_id=input_data.execution_id or None,
            time_start=time_start,
            time_end=time_end,
-            include_running=wait_timeout > 0,
        )

        if exec_error:
@@ -497,17 +443,4 @@ class AgentOutputTool(BaseTool):
                session_id=session_id,
            )

-        # If we have an execution that's still running and we should wait
-        if execution and wait_timeout > 0 and execution.status not in TERMINAL_STATUSES:
-            logger.info(
-                f"Execution {execution.id} is {execution.status}, "
-                f"waiting up to {wait_timeout}s for completion"
-            )
-            execution = await wait_for_execution(
-                user_id=user_id,
-                graph_id=agent.graph_id,
-                execution_id=execution.id,
-                timeout_seconds=wait_timeout,
-            )
-
        return self._build_response(agent, execution, available_executions, session_id)
--- a/autogpt_platform/backend/backend/api/features/chat/tools/agent_search.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/agent_search.py
@@ -4,7 +4,8 @@ import logging
 import re
 from typing import Literal

-from backend.data.db_accessors import library_db, store_db
+from backend.api.features.library import db as library_db
+from backend.api.features.store import db as store_db
 from backend.util.exceptions import DatabaseError, NotFoundError

 from .models import (
@@ -44,10 +45,8 @@ async def _get_library_agent_by_id(user_id: str, agent_id: str) -> AgentInfo | N
    Returns:
        AgentInfo if found, None otherwise
    """
-    lib_db = library_db()
-
    try:
-        agent = await lib_db.get_library_agent_by_graph_id(user_id, agent_id)
+        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(
@@ -72,7 +71,7 @@ async def _get_library_agent_by_id(user_id: str, agent_id: str) -> AgentInfo | N
        )

    try:
-        agent = await lib_db.get_library_agent(agent_id, user_id)
+        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(
@@ -134,7 +133,7 @@ async def search_agents(
    try:
        if source == "marketplace":
            logger.info(f"Searching marketplace for: {query}")
-            results = await store_db().get_store_agents(search_query=query, page_size=5)
+            results = await store_db.get_store_agents(search_query=query, page_size=5)
            for agent in results.agents:
                agents.append(
                    AgentInfo(
@@ -160,7 +159,7 @@ async def search_agents(

            if not agents:
                logger.info(f"Searching user library for: {query}")
-                results = await library_db().list_library_agents(
+                results = await library_db.list_library_agents(
                    user_id=user_id,  # type: ignore[arg-type]
                    search_term=query,
                    page_size=10,
--- a/autogpt_platform/backend/backend/api/features/chat/tools/base.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/base.py
@@ -5,8 +5,8 @@ from typing import Any

 from openai.types.chat import ChatCompletionToolParam

-from backend.copilot.model import ChatSession
-from backend.copilot.response_model import StreamToolOutputAvailable
+from backend.api.features.chat.model import ChatSession
+from backend.api.features.chat.response_model import StreamToolOutputAvailable

 from .models import ErrorResponse, NeedLoginResponse, ToolResponseBase

--- a/autogpt_platform/backend/backend/api/features/chat/tools/binary_output_processor.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/binary_output_processor.py
@@ -0,0 +1,251 @@
+"""
+Detect and save embedded binary data in block outputs.
+
+Scans stdout_logs and other string outputs for embedded base64 patterns,
+saves detected binary content to workspace, and replaces the base64 with
+workspace:// references. This reduces LLM output token usage by ~97% for
+file generation tasks.
+
+Primary use case: ExecuteCodeBlock prints base64 to stdout, which appears
+in stdout_logs. Without this processor, the LLM would re-type the entire
+base64 string when saving files.
+"""
+
+import base64
+import binascii
+import hashlib
+import logging
+import re
+import uuid
+from typing import Any, Optional
+
+from backend.util.file import sanitize_filename
+from backend.util.virus_scanner import scan_content_safe
+from backend.util.workspace import WorkspaceManager
+
+logger = logging.getLogger(__name__)
+
+# Minimum decoded size to process (filters out small base64 strings)
+MIN_DECODED_SIZE = 1024  # 1KB
+
+# Pattern to find base64 chunks in text (at least 100 chars to be worth checking)
+# Matches continuous base64 characters (with optional whitespace for line wrapping),
+# optionally ending with = padding
+EMBEDDED_BASE64_PATTERN = re.compile(r"[A-Za-z0-9+/\s]{100,}={0,2}")
+
+# Magic numbers for binary file detection
+MAGIC_SIGNATURES = [
+    (b"\x89PNG\r\n\x1a\n", "png"),
+    (b"\xff\xd8\xff", "jpg"),
+    (b"%PDF-", "pdf"),
+    (b"GIF87a", "gif"),
+    (b"GIF89a", "gif"),
+    (b"RIFF", "webp"),  # Also check content[8:12] == b'WEBP'
+]
+
+
+async def process_binary_outputs(
+    outputs: dict[str, list[Any]],
+    workspace_manager: WorkspaceManager,
+    block_name: str,
+) -> dict[str, list[Any]]:
+    """
+    Scan all string values in outputs for embedded base64 binary content.
+    Save detected binaries to workspace and replace with references.
+
+    Args:
+        outputs: Block execution outputs (dict of output_name -> list of values)
+        workspace_manager: WorkspaceManager instance with session scoping
+        block_name: Name of the block (used in generated filenames)
+
+    Returns:
+        Processed outputs with embedded base64 replaced by workspace references
+    """
+    cache: dict[str, str] = {}  # content_hash -> workspace_ref
+
+    processed: dict[str, list[Any]] = {}
+    for name, items in outputs.items():
+        processed_items = []
+        for item in items:
+            processed_items.append(
+                await _process_value(item, workspace_manager, block_name, cache)
+            )
+        processed[name] = processed_items
+    return processed
+
+
+async def _process_value(
+    value: Any,
+    wm: WorkspaceManager,
+    block: str,
+    cache: dict[str, str],
+) -> Any:
+    """Recursively process a value, detecting embedded base64 in strings."""
+    if isinstance(value, dict):
+        result = {}
+        for k, v in value.items():
+            result[k] = await _process_value(v, wm, block, cache)
+        return result
+    if isinstance(value, list):
+        return [await _process_value(v, wm, block, cache) for v in value]
+    if isinstance(value, str) and len(value) > MIN_DECODED_SIZE:
+        return await _extract_and_replace_base64(value, wm, block, cache)
+    return value
+
+
+async def _extract_and_replace_base64(
+    text: str,
+    wm: WorkspaceManager,
+    block: str,
+    cache: dict[str, str],
+) -> str:
+    """
+    Find embedded base64 in text, save binaries, replace with references.
+
+    Scans for base64 patterns, validates each as binary via magic numbers,
+    saves valid binaries to workspace, and replaces the base64 portion
+    (plus any surrounding markers) with the workspace reference.
+    """
+    result = text
+    offset = 0
+
+    for match in EMBEDDED_BASE64_PATTERN.finditer(text):
+        b64_str = match.group(0)
+
+        # Try to decode and validate
+        detection = _decode_and_validate(b64_str)
+        if detection is None:
+            continue
+
+        content, ext = detection
+
+        # Save to workspace
+        ref = await _save_binary(content, ext, wm, block, cache)
+        if ref is None:
+            continue
+
+        # Calculate replacement bounds (include surrounding markers if present)
+        start, end = match.start(), match.end()
+        start, end = _expand_to_markers(text, start, end)
+
+        # Apply replacement with offset adjustment
+        adj_start = start + offset
+        adj_end = end + offset
+        result = result[:adj_start] + ref + result[adj_end:]
+        offset += len(ref) - (end - start)
+
+    return result
+
+
+def _decode_and_validate(b64_str: str) -> Optional[tuple[bytes, str]]:
+    """
+    Decode base64 and validate it's a known binary format.
+
+    Tries multiple 4-byte aligned offsets to handle cases where marker text
+    (e.g., "START" from "PDF_BASE64_START") bleeds into the regex match.
+    Base64 works in 4-char chunks, so we only check aligned offsets.
+
+    Returns (content, extension) if valid binary, None otherwise.
+    """
+    # Strip whitespace for RFC 2045 line-wrapped base64
+    normalized = re.sub(r"\s+", "", b64_str)
+
+    # Try offsets 0, 4, 8, ... up to 32 chars (handles markers up to ~24 chars)
+    # This handles cases like "STARTJVBERi0..." where "START" bleeds into match
+    for char_offset in range(0, min(33, len(normalized)), 4):
+        candidate = normalized[char_offset:]
+
+        try:
+            content = base64.b64decode(candidate, validate=True)
+        except (ValueError, binascii.Error):
+            continue
+
+        # Must meet minimum size
+        if len(content) < MIN_DECODED_SIZE:
+            continue
+
+        # Check magic numbers
+        for magic, ext in MAGIC_SIGNATURES:
+            if content.startswith(magic):
+                # Special case for WebP: RIFF container, verify "WEBP" at offset 8
+                if magic == b"RIFF":
+                    if len(content) < 12 or content[8:12] != b"WEBP":
+                        continue
+                return content, ext
+
+    return None
+
+
+def _expand_to_markers(text: str, start: int, end: int) -> tuple[int, int]:
+    """
+    Expand replacement bounds to include surrounding markers if present.
+
+    Handles patterns like:
+    - ---BASE64_START---\\n{base64}\\n---BASE64_END---
+    - [BASE64]{base64}[/BASE64]
+    - Or just the raw base64
+    """
+    # Common marker patterns to strip (order matters - check longer patterns first)
+    start_markers = [
+        "PDF_BASE64_START",
+        "---BASE64_START---\n",
+        "---BASE64_START---",
+        "[BASE64]\n",
+        "[BASE64]",
+    ]
+    end_markers = [
+        "PDF_BASE64_END",
+        "\n---BASE64_END---",
+        "---BASE64_END---",
+        "\n[/BASE64]",
+        "[/BASE64]",
+    ]
+
+    # Check for start markers
+    for marker in start_markers:
+        marker_start = start - len(marker)
+        if marker_start >= 0 and text[marker_start:start] == marker:
+            start = marker_start
+            break
+
+    # Check for end markers
+    for marker in end_markers:
+        marker_end = end + len(marker)
+        if marker_end <= len(text) and text[end:marker_end] == marker:
+            end = marker_end
+            break
+
+    return start, end
+
+
+async def _save_binary(
+    content: bytes,
+    ext: str,
+    wm: WorkspaceManager,
+    block: str,
+    cache: dict[str, str],
+) -> Optional[str]:
+    """
+    Save binary content to workspace with deduplication.
+
+    Returns workspace://file-id reference, or None on failure.
+    """
+    content_hash = hashlib.sha256(content).hexdigest()
+
+    if content_hash in cache:
+        return cache[content_hash]
+
+    try:
+        safe_block = sanitize_filename(block)[:20].lower()
+        filename = f"{safe_block}_{uuid.uuid4().hex[:12]}.{ext}"
+
+        # Scan for viruses before saving
+        await scan_content_safe(content, filename=filename)
+
+        file = await wm.write_file(content, filename)
+        ref = f"workspace://{file.id}"
+        cache[content_hash] = ref
+        return ref
+    except Exception as e:
+        logger.warning("Failed to save binary output: %s", e)
+        return None
--- a/autogpt_platform/backend/backend/api/features/chat/tools/create_agent.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/create_agent.py
@@ -3,7 +3,7 @@
 import logging
 from typing import Any

-from backend.copilot.model import ChatSession
+from backend.api.features.chat.model import ChatSession

 from .agent_generator import (
    AgentGeneratorNotConfiguredError,
--- a/autogpt_platform/backend/backend/api/features/chat/tools/customize_agent.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/customize_agent.py
@@ -3,9 +3,9 @@
 import logging
 from typing import Any

+from backend.api.features.chat.model import ChatSession
+from backend.api.features.store import db as store_db
 from backend.api.features.store.exceptions import AgentNotFoundError
-from backend.copilot.model import ChatSession
-from backend.data.db_accessors import store_db as get_store_db

 from .agent_generator import (
    AgentGeneratorNotConfiguredError,
@@ -137,8 +137,6 @@ class CustomizeAgentTool(BaseTool):

        creator_username, agent_slug = parts

-        store_db = get_store_db()
-
        # Fetch the marketplace agent details
        try:
            agent_details = await store_db.get_store_agent_details(
--- a/autogpt_platform/backend/backend/api/features/chat/tools/edit_agent.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/edit_agent.py
@@ -3,7 +3,7 @@
 import logging
 from typing import Any

-from backend.copilot.model import ChatSession
+from backend.api.features.chat.model import ChatSession

 from .agent_generator import (
    AgentGeneratorNotConfiguredError,
--- a/autogpt_platform/backend/backend/api/features/chat/tools/find_agent.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/find_agent.py
@@ -2,7 +2,7 @@

 from typing import Any

-from backend.copilot.model import ChatSession
+from backend.api.features.chat.model import ChatSession

 from .agent_search import search_agents
 from .base import BaseTool
--- a/autogpt_platform/backend/backend/api/features/chat/tools/find_block.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/find_block.py
@@ -3,18 +3,17 @@ from typing import Any

 from prisma.enums import ContentType

-from backend.blocks import get_block
-from backend.blocks._base import BlockType
-from backend.copilot.model import ChatSession
-from backend.data.db_accessors import search
-
-from .base import BaseTool, ToolResponseBase
-from .models import (
+from backend.api.features.chat.model import ChatSession
+from backend.api.features.chat.tools.base import BaseTool, ToolResponseBase
+from backend.api.features.chat.tools.models import (
    BlockInfoSummary,
+    BlockInputFieldInfo,
    BlockListResponse,
    ErrorResponse,
    NoResultsResponse,
 )
+from backend.api.features.store.hybrid_search import unified_hybrid_search
+from backend.data.block import BlockType, get_block

 logger = logging.getLogger(__name__)

@@ -55,8 +54,7 @@ class FindBlockTool(BaseTool):
            "Blocks are reusable components that perform specific tasks like "
            "sending emails, making API calls, processing text, etc. "
            "IMPORTANT: Use this tool FIRST to get the block's 'id' before calling run_block. "
-            "The response includes each block's id, name, and description. "
-            "Call run_block with the block's id **with no inputs** to see detailed inputs/outputs and execute it."
+            "The response includes each block's id, required_inputs, and input_schema."
        )

    @property
@@ -108,7 +106,7 @@ class FindBlockTool(BaseTool):

        try:
            # Search for blocks using hybrid search
-            results, total = await search().unified_hybrid_search(
+            results, total = await unified_hybrid_search(
                query=query,
                content_types=[ContentType.BLOCK],
                page=1,
@@ -125,7 +123,7 @@ class FindBlockTool(BaseTool):
                    session_id=session_id,
                )

-            # Enrich results with block information
+            # Enrich results with full block information
            blocks: list[BlockInfoSummary] = []
            for result in results:
                block_id = result["content_id"]
@@ -142,12 +140,65 @@ class FindBlockTool(BaseTool):
                ):
                    continue

+                # Get input/output schemas
+                input_schema = {}
+                output_schema = {}
+                try:
+                    input_schema = block.input_schema.jsonschema()
+                except Exception as e:
+                    logger.debug(
+                        "Failed to generate input schema for block %s: %s",
+                        block_id,
+                        e,
+                    )
+                try:
+                    output_schema = block.output_schema.jsonschema()
+                except Exception as e:
+                    logger.debug(
+                        "Failed to generate output schema for block %s: %s",
+                        block_id,
+                        e,
+                    )
+
+                # Get categories from block instance
+                categories = []
+                if hasattr(block, "categories") and block.categories:
+                    categories = [cat.value for cat in block.categories]
+
+                # Extract required inputs for easier use
+                required_inputs: list[BlockInputFieldInfo] = []
+                if input_schema:
+                    properties = input_schema.get("properties", {})
+                    required_fields = set(input_schema.get("required", []))
+                    # Get credential field names to exclude from required inputs
+                    credentials_fields = set(
+                        block.input_schema.get_credentials_fields().keys()
+                    )
+
+                    for field_name, field_schema in properties.items():
+                        # Skip credential fields - they're handled separately
+                        if field_name in credentials_fields:
+                            continue
+
+                        required_inputs.append(
+                            BlockInputFieldInfo(
+                                name=field_name,
+                                type=field_schema.get("type", "string"),
+                                description=field_schema.get("description", ""),
+                                required=field_name in required_fields,
+                                default=field_schema.get("default"),
+                            )
+                        )
+
                blocks.append(
                    BlockInfoSummary(
                        id=block_id,
                        name=block.name,
                        description=block.description or "",
-                        categories=[c.value for c in block.categories],
+                        categories=categories,
+                        input_schema=input_schema,
+                        output_schema=output_schema,
+                        required_inputs=required_inputs,
                    )
                )

@@ -176,7 +227,8 @@ class FindBlockTool(BaseTool):
            return BlockListResponse(
                message=(
                    f"Found {len(blocks)} block(s) matching '{query}'. "
-                    "To see a block's inputs/outputs and execute it, use run_block with the block's 'id' - providing no inputs."
+                    "To execute a block, use run_block with the block's 'id' field "
+                    "and provide 'input_data' matching the block's input_schema."
                ),
                blocks=blocks,
                count=len(blocks),
--- a/autogpt_platform/backend/backend/api/features/chat/tools/find_block_test.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/find_block_test.py
@@ -0,0 +1,139 @@
+"""Tests for block filtering in FindBlockTool."""
+
+from unittest.mock import AsyncMock, MagicMock, patch
+
+import pytest
+
+from backend.api.features.chat.tools.find_block import (
+    COPILOT_EXCLUDED_BLOCK_IDS,
+    COPILOT_EXCLUDED_BLOCK_TYPES,
+    FindBlockTool,
+)
+from backend.api.features.chat.tools.models import BlockListResponse
+from backend.data.block import BlockType
+
+from ._test_data import make_session
+
+_TEST_USER_ID = "test-user-find-block"
+
+
+def make_mock_block(
+    block_id: str, name: str, block_type: BlockType, disabled: bool = False
+):
+    """Create a mock block for testing."""
+    mock = MagicMock()
+    mock.id = block_id
+    mock.name = name
+    mock.description = f"{name} description"
+    mock.block_type = block_type
+    mock.disabled = disabled
+    mock.input_schema = MagicMock()
+    mock.input_schema.jsonschema.return_value = {"properties": {}, "required": []}
+    mock.input_schema.get_credentials_fields.return_value = {}
+    mock.output_schema = MagicMock()
+    mock.output_schema.jsonschema.return_value = {}
+    mock.categories = []
+    return mock
+
+
+class TestFindBlockFiltering:
+    """Tests for block filtering in FindBlockTool."""
+
+    def test_excluded_block_types_contains_expected_types(self):
+        """Verify COPILOT_EXCLUDED_BLOCK_TYPES contains all graph-only types."""
+        assert BlockType.INPUT in COPILOT_EXCLUDED_BLOCK_TYPES
+        assert BlockType.OUTPUT in COPILOT_EXCLUDED_BLOCK_TYPES
+        assert BlockType.WEBHOOK in COPILOT_EXCLUDED_BLOCK_TYPES
+        assert BlockType.WEBHOOK_MANUAL in COPILOT_EXCLUDED_BLOCK_TYPES
+        assert BlockType.NOTE in COPILOT_EXCLUDED_BLOCK_TYPES
+        assert BlockType.HUMAN_IN_THE_LOOP in COPILOT_EXCLUDED_BLOCK_TYPES
+        assert BlockType.AGENT in COPILOT_EXCLUDED_BLOCK_TYPES
+
+    def test_excluded_block_ids_contains_smart_decision_maker(self):
+        """Verify SmartDecisionMakerBlock is in COPILOT_EXCLUDED_BLOCK_IDS."""
+        assert "3b191d9f-356f-482d-8238-ba04b6d18381" in COPILOT_EXCLUDED_BLOCK_IDS
+
+    @pytest.mark.asyncio(loop_scope="session")
+    async def test_excluded_block_type_filtered_from_results(self):
+        """Verify blocks with excluded BlockTypes are filtered from search results."""
+        session = make_session(user_id=_TEST_USER_ID)
+
+        # Mock search returns an INPUT block (excluded) and a STANDARD block (included)
+        search_results = [
+            {"content_id": "input-block-id", "score": 0.9},
+            {"content_id": "standard-block-id", "score": 0.8},
+        ]
+
+        input_block = make_mock_block("input-block-id", "Input Block", BlockType.INPUT)
+        standard_block = make_mock_block(
+            "standard-block-id", "HTTP Request", BlockType.STANDARD
+        )
+
+        def mock_get_block(block_id):
+            return {
+                "input-block-id": input_block,
+                "standard-block-id": standard_block,
+            }.get(block_id)
+
+        with patch(
+            "backend.api.features.chat.tools.find_block.unified_hybrid_search",
+            new_callable=AsyncMock,
+            return_value=(search_results, 2),
+        ):
+            with patch(
+                "backend.api.features.chat.tools.find_block.get_block",
+                side_effect=mock_get_block,
+            ):
+                tool = FindBlockTool()
+                response = await tool._execute(
+                    user_id=_TEST_USER_ID, session=session, query="test"
+                )
+
+        # Should only return the standard block, not the INPUT block
+        assert isinstance(response, BlockListResponse)
+        assert len(response.blocks) == 1
+        assert response.blocks[0].id == "standard-block-id"
+
+    @pytest.mark.asyncio(loop_scope="session")
+    async def test_excluded_block_id_filtered_from_results(self):
+        """Verify SmartDecisionMakerBlock is filtered from search results."""
+        session = make_session(user_id=_TEST_USER_ID)
+
+        smart_decision_id = "3b191d9f-356f-482d-8238-ba04b6d18381"
+        search_results = [
+            {"content_id": smart_decision_id, "score": 0.9},
+            {"content_id": "normal-block-id", "score": 0.8},
+        ]
+
+        # SmartDecisionMakerBlock has STANDARD type but is excluded by ID
+        smart_block = make_mock_block(
+            smart_decision_id, "Smart Decision Maker", BlockType.STANDARD
+        )
+        normal_block = make_mock_block(
+            "normal-block-id", "Normal Block", BlockType.STANDARD
+        )
+
+        def mock_get_block(block_id):
+            return {
+                smart_decision_id: smart_block,
+                "normal-block-id": normal_block,
+            }.get(block_id)
+
+        with patch(
+            "backend.api.features.chat.tools.find_block.unified_hybrid_search",
+            new_callable=AsyncMock,
+            return_value=(search_results, 2),
+        ):
+            with patch(
+                "backend.api.features.chat.tools.find_block.get_block",
+                side_effect=mock_get_block,
+            ):
+                tool = FindBlockTool()
+                response = await tool._execute(
+                    user_id=_TEST_USER_ID, session=session, query="decision"
+                )
+
+        # Should only return normal block, not SmartDecisionMakerBlock
+        assert isinstance(response, BlockListResponse)
+        assert len(response.blocks) == 1
+        assert response.blocks[0].id == "normal-block-id"
--- a/autogpt_platform/backend/backend/api/features/chat/tools/find_library_agent.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/find_library_agent.py
@@ -2,7 +2,7 @@

 from typing import Any

-from backend.copilot.model import ChatSession
+from backend.api.features.chat.model import ChatSession

 from .agent_search import search_agents
 from .base import BaseTool
--- a/autogpt_platform/backend/backend/api/features/chat/tools/get_doc_page.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/get_doc_page.py
@@ -4,10 +4,13 @@ import logging
 from pathlib import Path
 from typing import Any

-from backend.copilot.model import ChatSession
-
-from .base import BaseTool
-from .models import DocPageResponse, ErrorResponse, ToolResponseBase
+from backend.api.features.chat.model import ChatSession
+from backend.api.features.chat.tools.base import BaseTool
+from backend.api.features.chat.tools.models import (
+    DocPageResponse,
+    ErrorResponse,
+    ToolResponseBase,
+)

 logger = logging.getLogger(__name__)

--- a/autogpt_platform/backend/backend/api/features/chat/tools/models.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/models.py
@@ -25,7 +25,6 @@ class ResponseType(str, Enum):
    AGENT_SAVED = "agent_saved"
    CLARIFICATION_NEEDED = "clarification_needed"
    BLOCK_LIST = "block_list"
-    BLOCK_DETAILS = "block_details"
    BLOCK_OUTPUT = "block_output"
    DOC_SEARCH_RESULTS = "doc_search_results"
    DOC_PAGE = "doc_page"
@@ -41,15 +40,6 @@ class ResponseType(str, Enum):
    OPERATION_IN_PROGRESS = "operation_in_progress"
    # Input validation
    INPUT_VALIDATION_ERROR = "input_validation_error"
-    # Web fetch
-    WEB_FETCH = "web_fetch"
-    # Code execution
-    BASH_EXEC = "bash_exec"
-    # Operation status check
-    OPERATION_STATUS = "operation_status"
-    # Feature request types
-    FEATURE_REQUEST_SEARCH = "feature_request_search"
-    FEATURE_REQUEST_CREATED = "feature_request_created"


 # Base response model
@@ -345,17 +335,11 @@ class BlockInfoSummary(BaseModel):
    name: str
    description: str
    categories: list[str]
-    input_schema: dict[str, Any] = Field(
-        default_factory=dict,
-        description="Full JSON schema for block inputs",
-    )
-    output_schema: dict[str, Any] = Field(
-        default_factory=dict,
-        description="Full JSON schema for block outputs",
-    )
+    input_schema: dict[str, Any]
+    output_schema: dict[str, Any]
    required_inputs: list[BlockInputFieldInfo] = Field(
        default_factory=list,
-        description="List of input fields for this block",
+        description="List of required input fields for this block",
    )


@@ -368,29 +352,10 @@ class BlockListResponse(ToolResponseBase):
    query: str
    usage_hint: str = Field(
        default="To execute a block, call run_block with block_id set to the block's "
-        "'id' field and input_data containing the fields listed in required_inputs."
+        "'id' field and input_data containing the required fields from input_schema."
    )


-class BlockDetails(BaseModel):
-    """Detailed block information."""
-
-    id: str
-    name: str
-    description: str
-    inputs: dict[str, Any] = {}
-    outputs: dict[str, Any] = {}
-    credentials: list[CredentialsMetaInput] = []
-
-
-class BlockDetailsResponse(ToolResponseBase):
-    """Response for block details (first run_block attempt)."""
-
-    type: ResponseType = ResponseType.BLOCK_DETAILS
-    block: BlockDetails
-    user_authenticated: bool = False
-
-
 class BlockOutputResponse(ToolResponseBase):
    """Response for run_block tool."""

@@ -456,55 +421,3 @@ class AsyncProcessingResponse(ToolResponseBase):
    status: str = "accepted"  # Must be "accepted" for detection
    operation_id: str | None = None
    task_id: str | None = None
-
-
-class WebFetchResponse(ToolResponseBase):
-    """Response for web_fetch tool."""
-
-    type: ResponseType = ResponseType.WEB_FETCH
-    url: str
-    status_code: int
-    content_type: str
-    content: str
-    truncated: bool = False
-
-
-class BashExecResponse(ToolResponseBase):
-    """Response for bash_exec tool."""
-
-    type: ResponseType = ResponseType.BASH_EXEC
-    stdout: str
-    stderr: str
-    exit_code: int
-    timed_out: bool = False
-
-
-# Feature request models
-class FeatureRequestInfo(BaseModel):
-    """Information about a feature request issue."""
-
-    id: str
-    identifier: str
-    title: str
-    description: str | None = None
-
-
-class FeatureRequestSearchResponse(ToolResponseBase):
-    """Response for search_feature_requests tool."""
-
-    type: ResponseType = ResponseType.FEATURE_REQUEST_SEARCH
-    results: list[FeatureRequestInfo]
-    count: int
-    query: str
-
-
-class FeatureRequestCreatedResponse(ToolResponseBase):
-    """Response for create_feature_request tool."""
-
-    type: ResponseType = ResponseType.FEATURE_REQUEST_CREATED
-    issue_id: str
-    issue_identifier: str
-    issue_title: str
-    issue_url: str
-    is_new_issue: bool  # False if added to existing
-    customer_name: str
--- a/autogpt_platform/backend/backend/api/features/chat/tools/run_agent.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/run_agent.py
@@ -5,13 +5,16 @@ from typing import Any

 from pydantic import BaseModel, Field, field_validator

-from backend.copilot.config import ChatConfig
-from backend.copilot.model import ChatSession
-from backend.copilot.tracking import track_agent_run_success, track_agent_scheduled
-from backend.data.db_accessors import graph_db, library_db, user_db
-from backend.data.execution import ExecutionStatus
+from backend.api.features.chat.config import ChatConfig
+from backend.api.features.chat.model import ChatSession
+from backend.api.features.chat.tracking import (
+    track_agent_run_success,
+    track_agent_scheduled,
+)
+from backend.api.features.library import db as library_db
 from backend.data.graph import GraphModel
 from backend.data.model import CredentialsMetaInput
+from backend.data.user import get_user_by_id
 from backend.executor import utils as execution_utils
 from backend.util.clients import get_scheduler_client
 from backend.util.exceptions import DatabaseError, NotFoundError
@@ -21,11 +24,9 @@ from backend.util.timezone_utils import (
 )

 from .base import BaseTool
-from .helpers import get_inputs_from_schema
 from .models import (
    AgentDetails,
    AgentDetailsResponse,
-    AgentOutputResponse,
    ErrorResponse,
    ExecutionOptions,
    ExecutionStartedResponse,
@@ -35,7 +36,6 @@ from .models import (
    ToolResponseBase,
    UserReadiness,
 )
-from .execution_utils import get_execution_outputs, wait_for_execution
 from .utils import (
    build_missing_credentials_from_graph,
    extract_credentials_from_schema,
@@ -69,7 +69,6 @@ class RunAgentInput(BaseModel):
    schedule_name: str = ""
    cron: str = ""
    timezone: str = "UTC"
-    wait_for_result: int = Field(default=0, ge=0, le=300)

    @field_validator(
        "username_agent_slug",
@@ -151,14 +150,6 @@ class RunAgentTool(BaseTool):
                    "type": "string",
                    "description": "IANA timezone for schedule (default: UTC)",
                },
-                "wait_for_result": {
-                    "type": "integer",
-                    "description": (
-                        "Max seconds to wait for execution to complete (0-300). "
-                        "If >0, blocks until the execution finishes or times out. "
-                        "Returns execution outputs when complete."
-                    ),
-                },
            },
            "required": [],
        }
@@ -208,7 +199,7 @@ class RunAgentTool(BaseTool):

            # Priority: library_agent_id if provided
            if has_library_id:
-                library_agent = await library_db().get_library_agent(
+                library_agent = await library_db.get_library_agent(
                    params.library_agent_id, user_id
                )
                if not library_agent:
@@ -217,7 +208,9 @@ class RunAgentTool(BaseTool):
                        session_id=session_id,
                    )
                # Get the graph from the library agent
-                graph = await graph_db().get_graph(
+                from backend.data.graph import get_graph
+
+                graph = await get_graph(
                    library_agent.graph_id,
                    library_agent.graph_version,
                    user_id=user_id,
@@ -268,7 +261,7 @@ class RunAgentTool(BaseTool):
                        ),
                        requirements={
                            "credentials": requirements_creds_list,
-                            "inputs": get_inputs_from_schema(graph.input_schema),
+                            "inputs": self._get_inputs_list(graph.input_schema),
                            "execution_modes": self._get_execution_modes(graph),
                        },
                    ),
@@ -353,7 +346,6 @@ class RunAgentTool(BaseTool):
                    graph=graph,
                    graph_credentials=graph_credentials,
                    inputs=params.inputs,
-                    wait_for_result=params.wait_for_result,
                )

        except NotFoundError as e:
@@ -377,6 +369,22 @@ class RunAgentTool(BaseTool):
                session_id=session_id,
            )

+    def _get_inputs_list(self, input_schema: dict[str, Any]) -> list[dict[str, Any]]:
+        """Extract inputs list from schema."""
+        inputs_list = []
+        if isinstance(input_schema, dict) and "properties" in input_schema:
+            for field_name, field_schema in input_schema["properties"].items():
+                inputs_list.append(
+                    {
+                        "name": field_name,
+                        "title": field_schema.get("title", field_name),
+                        "type": field_schema.get("type", "string"),
+                        "description": field_schema.get("description", ""),
+                        "required": field_name in input_schema.get("required", []),
+                    }
+                )
+        return inputs_list
+
    def _get_execution_modes(self, graph: GraphModel) -> list[str]:
        """Get available execution modes for the graph."""
        trigger_info = graph.trigger_setup_info
@@ -390,7 +398,7 @@ class RunAgentTool(BaseTool):
        suffix: str,
    ) -> str:
        """Build a message describing available inputs for an agent."""
-        inputs_list = get_inputs_from_schema(graph.input_schema)
+        inputs_list = self._get_inputs_list(graph.input_schema)
        required_names = [i["name"] for i in inputs_list if i["required"]]
        optional_names = [i["name"] for i in inputs_list if not i["required"]]

@@ -437,9 +445,8 @@ class RunAgentTool(BaseTool):
        graph: GraphModel,
        graph_credentials: dict[str, CredentialsMetaInput],
        inputs: dict[str, Any],
-        wait_for_result: int = 0,
    ) -> ToolResponseBase:
-        """Execute an agent immediately, optionally waiting for completion."""
+        """Execute an agent immediately."""
        session_id = session.session_id

        # Check rate limits
@@ -476,56 +483,6 @@ class RunAgentTool(BaseTool):
        )

        library_agent_link = f"/library/agents/{library_agent.id}"
-
-        # If wait_for_result is requested, wait for execution to complete
-        if wait_for_result > 0:
-            logger.info(
-                f"Waiting up to {wait_for_result}s for execution {execution.id}"
-            )
-            completed = await wait_for_execution(
-                user_id=user_id,
-                graph_id=library_agent.graph_id,
-                execution_id=execution.id,
-                timeout_seconds=wait_for_result,
-            )
-
-            if completed and completed.status == ExecutionStatus.COMPLETED:
-                outputs = get_execution_outputs(completed)
-                return AgentOutputResponse(
-                    message=(
-                        f"Agent '{library_agent.name}' completed successfully. "
-                        f"View at {library_agent_link}."
-                    ),
-                    session_id=session_id,
-                    execution_id=execution.id,
-                    graph_id=library_agent.graph_id,
-                    graph_name=library_agent.name,
-                    outputs=outputs or {},
-                )
-            elif completed and completed.status == ExecutionStatus.FAILED:
-                return ErrorResponse(
-                    message=(
-                        f"Agent '{library_agent.name}' execution failed. "
-                        f"View details at {library_agent_link}."
-                    ),
-                    session_id=session_id,
-                )
-            else:
-                status = completed.status.value if completed else "unknown"
-                return ExecutionStartedResponse(
-                    message=(
-                        f"Agent '{library_agent.name}' is still {status} after "
-                        f"{wait_for_result}s. Check results later at {library_agent_link}. "
-                        f"Use view_agent_output with wait_if_running to check again."
-                    ),
-                    session_id=session_id,
-                    execution_id=execution.id,
-                    graph_id=library_agent.graph_id,
-                    graph_name=library_agent.name,
-                    library_agent_id=library_agent.id,
-                    library_agent_link=library_agent_link,
-                )
-
        return ExecutionStartedResponse(
            message=(
                f"Agent '{library_agent.name}' execution started successfully. "
@@ -580,7 +537,7 @@ class RunAgentTool(BaseTool):
        library_agent = await get_or_create_library_agent(graph, user_id)

        # Get user timezone
-        user = await user_db().get_user_by_id(user_id)
+        user = await get_user_by_id(user_id)
        user_timezone = get_user_timezone_or_utc(user.timezone if user else timezone)

        # Create schedule
--- a/autogpt_platform/backend/backend/api/features/chat/tools/run_agent_test.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/run_agent_test.py
--- a/autogpt_platform/backend/backend/api/features/chat/tools/run_block.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/run_block.py
@@ -7,33 +7,30 @@ from typing import Any

 from pydantic_core import PydanticUndefined

-from backend.blocks import get_block
-from backend.blocks._base import AnyBlockSchema
-from backend.copilot.model import ChatSession
-from backend.data.db_accessors import workspace_db
+from backend.api.features.chat.model import ChatSession
+from backend.api.features.chat.tools.find_block import (
+    COPILOT_EXCLUDED_BLOCK_IDS,
+    COPILOT_EXCLUDED_BLOCK_TYPES,
+)
+from backend.data.block import get_block
 from backend.data.execution import ExecutionContext
-from backend.data.model import CredentialsFieldInfo, CredentialsMetaInput
+from backend.data.model import CredentialsMetaInput
+from backend.data.workspace import get_or_create_workspace
 from backend.integrations.creds_manager import IntegrationCredentialsManager
 from backend.util.exceptions import BlockError
+from backend.util.workspace import WorkspaceManager

 from .base import BaseTool
-from .find_block import COPILOT_EXCLUDED_BLOCK_IDS, COPILOT_EXCLUDED_BLOCK_TYPES
-from .helpers import get_inputs_from_schema
+from .binary_output_processor import process_binary_outputs
 from .models import (
-    BlockDetails,
-    BlockDetailsResponse,
    BlockOutputResponse,
    ErrorResponse,
-    InputValidationErrorResponse,
    SetupInfo,
    SetupRequirementsResponse,
    ToolResponseBase,
    UserReadiness,
 )
-from .utils import (
-    build_missing_credentials_from_field_info,
-    match_credentials_to_requirements,
-)
+from .utils import build_missing_credentials_from_field_info

 logger = logging.getLogger(__name__)

@@ -51,8 +48,8 @@ class RunBlockTool(BaseTool):
            "Execute a specific block with the provided input data. "
            "IMPORTANT: You MUST call find_block first to get the block's 'id' - "
            "do NOT guess or make up block IDs. "
-            "On first attempt (without input_data), returns detailed schema showing "
-            "required inputs and outputs. Then call again with proper input_data to execute."
+            "Use the 'id' from find_block results and provide input_data "
+            "matching the block's required_inputs."
        )

    @property
@@ -67,19 +64,11 @@ class RunBlockTool(BaseTool):
                        "NEVER guess this - always get it from find_block first."
                    ),
                },
-                "block_name": {
-                    "type": "string",
-                    "description": (
-                        "The block's human-readable name from find_block results. "
-                        "Used for display purposes in the UI."
-                    ),
-                },
                "input_data": {
                    "type": "object",
                    "description": (
-                        "Input values for the block. "
-                        "First call with empty {} to see the block's schema, "
-                        "then call again with proper values to execute."
+                        "Input values for the block. Use the 'required_inputs' field "
+                        "from find_block to see what fields are needed."
                    ),
                },
            },
@@ -90,6 +79,91 @@ class RunBlockTool(BaseTool):
    def requires_auth(self) -> bool:
        return True

+    async def _check_block_credentials(
+        self,
+        user_id: str,
+        block: Any,
+        input_data: dict[str, Any] | None = None,
+    ) -> tuple[dict[str, CredentialsMetaInput], list[CredentialsMetaInput]]:
+        """
+        Check if user has required credentials for a block.
+
+        Args:
+            user_id: User ID
+            block: Block to check credentials for
+            input_data: Input data for the block (used to determine provider via discriminator)
+
+        Returns:
+            tuple[matched_credentials, missing_credentials]
+        """
+        matched_credentials: dict[str, CredentialsMetaInput] = {}
+        missing_credentials: list[CredentialsMetaInput] = []
+        input_data = input_data or {}
+
+        # Get credential field info from block's input schema
+        credentials_fields_info = block.input_schema.get_credentials_fields_info()
+
+        if not credentials_fields_info:
+            return matched_credentials, missing_credentials
+
+        # Get user's available credentials
+        creds_manager = IntegrationCredentialsManager()
+        available_creds = await creds_manager.store.get_all_creds(user_id)
+
+        for field_name, field_info in credentials_fields_info.items():
+            effective_field_info = field_info
+            if field_info.discriminator and field_info.discriminator_mapping:
+                # Get discriminator from input, falling back to schema default
+                discriminator_value = input_data.get(field_info.discriminator)
+                if discriminator_value is None:
+                    field = block.input_schema.model_fields.get(
+                        field_info.discriminator
+                    )
+                    if field and field.default is not PydanticUndefined:
+                        discriminator_value = field.default
+
+                if (
+                    discriminator_value
+                    and discriminator_value in field_info.discriminator_mapping
+                ):
+                    effective_field_info = field_info.discriminate(discriminator_value)
+                    logger.debug(
+                        f"Discriminated provider for {field_name}: "
+                        f"{discriminator_value} -> {effective_field_info.provider}"
+                    )
+
+            matching_cred = next(
+                (
+                    cred
+                    for cred in available_creds
+                    if cred.provider in effective_field_info.provider
+                    and cred.type in effective_field_info.supported_types
+                ),
+                None,
+            )
+
+            if matching_cred:
+                matched_credentials[field_name] = CredentialsMetaInput(
+                    id=matching_cred.id,
+                    provider=matching_cred.provider,  # type: ignore
+                    type=matching_cred.type,
+                    title=matching_cred.title,
+                )
+            else:
+                # Create a placeholder for the missing credential
+                provider = next(iter(effective_field_info.provider), "unknown")
+                cred_type = next(iter(effective_field_info.supported_types), "api_key")
+                missing_credentials.append(
+                    CredentialsMetaInput(
+                        id=field_name,
+                        provider=provider,  # type: ignore
+                        type=cred_type,  # type: ignore
+                        title=field_name.replace("_", " ").title(),
+                    )
+                )
+
+        return matched_credentials, missing_credentials
+
    async def _execute(
        self,
        user_id: str | None,
@@ -160,38 +234,10 @@ class RunBlockTool(BaseTool):
        logger.info(f"Executing block {block.name} ({block_id}) for user {user_id}")

        creds_manager = IntegrationCredentialsManager()
-        matched_credentials, missing_credentials = (
-            await self._resolve_block_credentials(user_id, block, input_data)
+        matched_credentials, missing_credentials = await self._check_block_credentials(
+            user_id, block, input_data
        )

-        # Get block schemas for details/validation
-        try:
-            input_schema: dict[str, Any] = block.input_schema.jsonschema()
-        except Exception as e:
-            logger.warning(
-                "Failed to generate input schema for block %s: %s",
-                block_id,
-                e,
-            )
-            return ErrorResponse(
-                message=f"Block '{block.name}' has an invalid input schema",
-                error=str(e),
-                session_id=session_id,
-            )
-        try:
-            output_schema: dict[str, Any] = block.output_schema.jsonschema()
-        except Exception as e:
-            logger.warning(
-                "Failed to generate output schema for block %s: %s",
-                block_id,
-                e,
-            )
-            return ErrorResponse(
-                message=f"Block '{block.name}' has an invalid output schema",
-                error=str(e),
-                session_id=session_id,
-            )
-
        if missing_credentials:
            # Return setup requirements response with missing credentials
            credentials_fields_info = block.input_schema.get_credentials_fields_info()
@@ -224,56 +270,9 @@ class RunBlockTool(BaseTool):
                graph_version=None,
            )

-        # Check if this is a first attempt (required inputs missing)
-        # Return block details so user can see what inputs are needed
-        credentials_fields = set(block.input_schema.get_credentials_fields().keys())
-        required_keys = set(input_schema.get("required", []))
-        required_non_credential_keys = required_keys - credentials_fields
-        provided_input_keys = set(input_data.keys()) - credentials_fields
-
-        # Check for unknown input fields
-        valid_fields = (
-            set(input_schema.get("properties", {}).keys()) - credentials_fields
-        )
-        unrecognized_fields = provided_input_keys - valid_fields
-        if unrecognized_fields:
-            return InputValidationErrorResponse(
-                message=(
-                    f"Unknown input field(s) provided: {', '.join(sorted(unrecognized_fields))}. "
-                    f"Block was not executed. Please use the correct field names from the schema."
-                ),
-                session_id=session_id,
-                unrecognized_fields=sorted(unrecognized_fields),
-                inputs=input_schema,
-            )
-
-        # Show details when not all required non-credential inputs are provided
-        if not (required_non_credential_keys <= provided_input_keys):
-            # Get credentials info for the response
-            credentials_meta = []
-            for field_name, cred_meta in matched_credentials.items():
-                credentials_meta.append(cred_meta)
-
-            return BlockDetailsResponse(
-                message=(
-                    f"Block '{block.name}' details. "
-                    "Provide input_data matching the inputs schema to execute the block."
-                ),
-                session_id=session_id,
-                block=BlockDetails(
-                    id=block_id,
-                    name=block.name,
-                    description=block.description or "",
-                    inputs=input_schema,
-                    outputs=output_schema,
-                    credentials=credentials_meta,
-                ),
-                user_authenticated=True,
-            )
-
        try:
            # Get or create user's workspace for CoPilot file operations
-            workspace = await workspace_db().get_or_create_workspace(user_id)
+            workspace = await get_or_create_workspace(user_id)

            # Generate synthetic IDs for CoPilot context
            # Each chat session is treated as its own agent with one continuous run
@@ -341,6 +340,16 @@ class RunBlockTool(BaseTool):
            ):
                outputs[output_name].append(output_data)

+            # Post-process outputs to save binary content to workspace
+            workspace_manager = WorkspaceManager(
+                user_id=user_id,
+                workspace_id=workspace.id,
+                session_id=session.session_id,
+            )
+            outputs = await process_binary_outputs(
+                dict(outputs), workspace_manager, block.name
+            )
+
            return BlockOutputResponse(
                message=f"Block '{block.name}' executed successfully",
                block_id=block_id,
@@ -365,75 +374,29 @@ class RunBlockTool(BaseTool):
                session_id=session_id,
            )

-    async def _resolve_block_credentials(
-        self,
-        user_id: str,
-        block: AnyBlockSchema,
-        input_data: dict[str, Any] | None = None,
-    ) -> tuple[dict[str, CredentialsMetaInput], list[CredentialsMetaInput]]:
-        """
-        Resolve credentials for a block by matching user's available credentials.
-
-        Args:
-            user_id: User ID
-            block: Block to resolve credentials for
-            input_data: Input data for the block (used to determine provider via discriminator)
-
-        Returns:
-            tuple of (matched_credentials, missing_credentials) - matched credentials
-            are used for block execution, missing ones indicate setup requirements.
-        """
-        input_data = input_data or {}
-        requirements = self._resolve_discriminated_credentials(block, input_data)
-
-        if not requirements:
-            return {}, []
-
-        return await match_credentials_to_requirements(user_id, requirements)
-
-    def _get_inputs_list(self, block: AnyBlockSchema) -> list[dict[str, Any]]:
+    def _get_inputs_list(self, block: Any) -> list[dict[str, Any]]:
        """Extract non-credential inputs from block schema."""
+        inputs_list = []
        schema = block.input_schema.jsonschema()
+        properties = schema.get("properties", {})
+        required_fields = set(schema.get("required", []))
+
+        # Get credential field names to exclude
        credentials_fields = set(block.input_schema.get_credentials_fields().keys())
-        return get_inputs_from_schema(schema, exclude_fields=credentials_fields)

-    def _resolve_discriminated_credentials(
-        self,
-        block: AnyBlockSchema,
-        input_data: dict[str, Any],
-    ) -> dict[str, CredentialsFieldInfo]:
-        """Resolve credential requirements, applying discriminator logic where needed."""
-        credentials_fields_info = block.input_schema.get_credentials_fields_info()
-        if not credentials_fields_info:
-            return {}
+        for field_name, field_schema in properties.items():
+            # Skip credential fields
+            if field_name in credentials_fields:
+                continue

-        resolved: dict[str, CredentialsFieldInfo] = {}
+            inputs_list.append(
+                {
+                    "name": field_name,
+                    "title": field_schema.get("title", field_name),
+                    "type": field_schema.get("type", "string"),
+                    "description": field_schema.get("description", ""),
+                    "required": field_name in required_fields,
+                }
+            )

-        for field_name, field_info in credentials_fields_info.items():
-            effective_field_info = field_info
-
-            if field_info.discriminator and field_info.discriminator_mapping:
-                discriminator_value = input_data.get(field_info.discriminator)
-                if discriminator_value is None:
-                    field = block.input_schema.model_fields.get(
-                        field_info.discriminator
-                    )
-                    if field and field.default is not PydanticUndefined:
-                        discriminator_value = field.default
-
-                if (
-                    discriminator_value
-                    and discriminator_value in field_info.discriminator_mapping
-                ):
-                    effective_field_info = field_info.discriminate(discriminator_value)
-                    # For host-scoped credentials, add the discriminator value
-                    # (e.g., URL) so _credential_is_for_host can match it
-                    effective_field_info.discriminator_values.add(discriminator_value)
-                    logger.debug(
-                        f"Discriminated provider for {field_name}: "
-                        f"{discriminator_value} -> {effective_field_info.provider}"
-                    )
-
-            resolved[field_name] = effective_field_info
-
-        return resolved
+        return inputs_list
--- a/autogpt_platform/backend/backend/api/features/chat/tools/run_block_test.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/run_block_test.py
@@ -0,0 +1,106 @@
+"""Tests for block execution guards in RunBlockTool."""
+
+from unittest.mock import MagicMock, patch
+
+import pytest
+
+from backend.api.features.chat.tools.models import ErrorResponse
+from backend.api.features.chat.tools.run_block import RunBlockTool
+from backend.data.block import BlockType
+
+from ._test_data import make_session
+
+_TEST_USER_ID = "test-user-run-block"
+
+
+def make_mock_block(
+    block_id: str, name: str, block_type: BlockType, disabled: bool = False
+):
+    """Create a mock block for testing."""
+    mock = MagicMock()
+    mock.id = block_id
+    mock.name = name
+    mock.block_type = block_type
+    mock.disabled = disabled
+    mock.input_schema = MagicMock()
+    mock.input_schema.jsonschema.return_value = {"properties": {}, "required": []}
+    mock.input_schema.get_credentials_fields_info.return_value = []
+    return mock
+
+
+class TestRunBlockFiltering:
+    """Tests for block execution guards in RunBlockTool."""
+
+    @pytest.mark.asyncio(loop_scope="session")
+    async def test_excluded_block_type_returns_error(self):
+        """Attempting to execute a block with excluded BlockType returns error."""
+        session = make_session(user_id=_TEST_USER_ID)
+
+        input_block = make_mock_block("input-block-id", "Input Block", BlockType.INPUT)
+
+        with patch(
+            "backend.api.features.chat.tools.run_block.get_block",
+            return_value=input_block,
+        ):
+            tool = RunBlockTool()
+            response = await tool._execute(
+                user_id=_TEST_USER_ID,
+                session=session,
+                block_id="input-block-id",
+                input_data={},
+            )
+
+        assert isinstance(response, ErrorResponse)
+        assert "cannot be run directly in CoPilot" in response.message
+        assert "designed for use within graphs only" in response.message
+
+    @pytest.mark.asyncio(loop_scope="session")
+    async def test_excluded_block_id_returns_error(self):
+        """Attempting to execute SmartDecisionMakerBlock returns error."""
+        session = make_session(user_id=_TEST_USER_ID)
+
+        smart_decision_id = "3b191d9f-356f-482d-8238-ba04b6d18381"
+        smart_block = make_mock_block(
+            smart_decision_id, "Smart Decision Maker", BlockType.STANDARD
+        )
+
+        with patch(
+            "backend.api.features.chat.tools.run_block.get_block",
+            return_value=smart_block,
+        ):
+            tool = RunBlockTool()
+            response = await tool._execute(
+                user_id=_TEST_USER_ID,
+                session=session,
+                block_id=smart_decision_id,
+                input_data={},
+            )
+
+        assert isinstance(response, ErrorResponse)
+        assert "cannot be run directly in CoPilot" in response.message
+
+    @pytest.mark.asyncio(loop_scope="session")
+    async def test_non_excluded_block_passes_guard(self):
+        """Non-excluded blocks pass the filtering guard (may fail later for other reasons)."""
+        session = make_session(user_id=_TEST_USER_ID)
+
+        standard_block = make_mock_block(
+            "standard-id", "HTTP Request", BlockType.STANDARD
+        )
+
+        with patch(
+            "backend.api.features.chat.tools.run_block.get_block",
+            return_value=standard_block,
+        ):
+            tool = RunBlockTool()
+            response = await tool._execute(
+                user_id=_TEST_USER_ID,
+                session=session,
+                block_id="standard-id",
+                input_data={},
+            )
+
+        # Should NOT be an ErrorResponse about CoPilot exclusion
+        # (may be other errors like missing credentials, but not the exclusion guard)
+        if isinstance(response, ErrorResponse):
+            assert "cannot be run directly in CoPilot" not in response.message
--- a/autogpt_platform/backend/backend/api/features/chat/tools/search_docs.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/search_docs.py
@@ -5,17 +5,16 @@ from typing import Any

 from prisma.enums import ContentType

-from backend.copilot.model import ChatSession
-from backend.data.db_accessors import search
-
-from .base import BaseTool
-from .models import (
+from backend.api.features.chat.model import ChatSession
+from backend.api.features.chat.tools.base import BaseTool
+from backend.api.features.chat.tools.models import (
    DocSearchResult,
    DocSearchResultsResponse,
    ErrorResponse,
    NoResultsResponse,
    ToolResponseBase,
 )
+from backend.api.features.store.hybrid_search import unified_hybrid_search

 logger = logging.getLogger(__name__)

@@ -118,7 +117,7 @@ class SearchDocsTool(BaseTool):

        try:
            # Search using hybrid search for DOCUMENTATION content type only
-            results, total = await search().unified_hybrid_search(
+            results, total = await unified_hybrid_search(
                query=query,
                content_types=[ContentType.DOCUMENTATION],
                page=1,
--- a/autogpt_platform/backend/backend/api/features/chat/tools/test_binary_output_processor.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/test_binary_output_processor.py
@@ -0,0 +1,518 @@
+"""Tests for embedded binary detection in block outputs."""
+
+import base64
+from unittest.mock import AsyncMock, MagicMock, patch
+
+import pytest
+
+from .binary_output_processor import (
+    _decode_and_validate,
+    _expand_to_markers,
+    process_binary_outputs,
+)
+
+
+@pytest.fixture
+def mock_workspace_manager():
+    """Create a mock workspace manager that returns predictable file IDs."""
+    wm = MagicMock()
+
+    async def mock_write_file(content, filename):
+        file = MagicMock()
+        file.id = f"file-{filename[:10]}"
+        return file
+
+    wm.write_file = AsyncMock(side_effect=mock_write_file)
+    return wm
+
+
+def _make_pdf_base64(size: int = 2000) -> str:
+    """Create a valid PDF base64 string of specified size."""
+    pdf_content = b"%PDF-1.4 " + b"x" * size
+    return base64.b64encode(pdf_content).decode()
+
+
+def _make_png_base64(size: int = 2000) -> str:
+    """Create a valid PNG base64 string of specified size."""
+    png_content = b"\x89PNG\r\n\x1a\n" + b"\x00" * size
+    return base64.b64encode(png_content).decode()
+
+
+# =============================================================================
+# Decode and Validate Tests
+# =============================================================================
+
+
+class TestDecodeAndValidate:
+    """Tests for _decode_and_validate function."""
+
+    def test_detects_pdf_magic_number(self):
+        """Should detect valid PDF by magic number."""
+        pdf_b64 = _make_pdf_base64()
+        result = _decode_and_validate(pdf_b64)
+        assert result is not None
+        content, ext = result
+        assert ext == "pdf"
+        assert content.startswith(b"%PDF-")
+
+    def test_detects_png_magic_number(self):
+        """Should detect valid PNG by magic number."""
+        png_b64 = _make_png_base64()
+        result = _decode_and_validate(png_b64)
+        assert result is not None
+        content, ext = result
+        assert ext == "png"
+
+    def test_detects_jpeg_magic_number(self):
+        """Should detect valid JPEG by magic number."""
+        jpeg_content = b"\xff\xd8\xff\xe0" + b"\x00" * 2000
+        jpeg_b64 = base64.b64encode(jpeg_content).decode()
+        result = _decode_and_validate(jpeg_b64)
+        assert result is not None
+        _, ext = result
+        assert ext == "jpg"
+
+    def test_detects_gif_magic_number(self):
+        """Should detect valid GIF by magic number."""
+        gif_content = b"GIF89a" + b"\x00" * 2000
+        gif_b64 = base64.b64encode(gif_content).decode()
+        result = _decode_and_validate(gif_b64)
+        assert result is not None
+        _, ext = result
+        assert ext == "gif"
+
+    def test_detects_webp_magic_number(self):
+        """Should detect valid WebP by magic number."""
+        webp_content = b"RIFF\x00\x00\x00\x00WEBP" + b"\x00" * 2000
+        webp_b64 = base64.b64encode(webp_content).decode()
+        result = _decode_and_validate(webp_b64)
+        assert result is not None
+        _, ext = result
+        assert ext == "webp"
+
+    def test_rejects_small_content(self):
+        """Should reject content smaller than threshold."""
+        small_pdf = b"%PDF-1.4 small"
+        small_b64 = base64.b64encode(small_pdf).decode()
+        result = _decode_and_validate(small_b64)
+        assert result is None
+
+    def test_rejects_no_magic_number(self):
+        """Should reject content without recognized magic number."""
+        random_content = b"This is just random text" * 100
+        random_b64 = base64.b64encode(random_content).decode()
+        result = _decode_and_validate(random_b64)
+        assert result is None
+
+    def test_rejects_invalid_base64(self):
+        """Should reject invalid base64."""
+        result = _decode_and_validate("not-valid-base64!!!")
+        assert result is None
+
+    def test_rejects_riff_without_webp(self):
+        """Should reject RIFF files that aren't WebP (e.g., WAV)."""
+        wav_content = b"RIFF\x00\x00\x00\x00WAVE" + b"\x00" * 2000
+        wav_b64 = base64.b64encode(wav_content).decode()
+        result = _decode_and_validate(wav_b64)
+        assert result is None
+
+    def test_handles_line_wrapped_base64(self):
+        """Should handle RFC 2045 line-wrapped base64."""
+        pdf_content = b"%PDF-1.4 " + b"x" * 2000
+        pdf_b64 = base64.b64encode(pdf_content).decode()
+        # Simulate line wrapping at 76 chars
+        wrapped = "\n".join(pdf_b64[i : i + 76] for i in range(0, len(pdf_b64), 76))
+        result = _decode_and_validate(wrapped)
+        assert result is not None
+        content, ext = result
+        assert ext == "pdf"
+        assert content == pdf_content
+
+
+# =============================================================================
+# Marker Expansion Tests
+# =============================================================================
+
+
+class TestExpandToMarkers:
+    """Tests for _expand_to_markers function."""
+
+    def test_expands_base64_start_end_markers(self):
+        """Should expand to include ---BASE64_START--- and ---BASE64_END---."""
+        text = "prefix\n---BASE64_START---\nABCDEF\n---BASE64_END---\nsuffix"
+        # Base64 "ABCDEF" is at position 26-32
+        start, end = _expand_to_markers(text, 26, 32)
+        assert text[start:end] == "---BASE64_START---\nABCDEF\n---BASE64_END---"
+
+    def test_expands_bracket_markers(self):
+        """Should expand to include [BASE64] and [/BASE64] markers."""
+        text = "prefix[BASE64]ABCDEF[/BASE64]suffix"
+        # Base64 is at position 14-20
+        start, end = _expand_to_markers(text, 14, 20)
+        assert text[start:end] == "[BASE64]ABCDEF[/BASE64]"
+
+    def test_no_expansion_without_markers(self):
+        """Should not expand if no markers present."""
+        text = "prefix ABCDEF suffix"
+        start, end = _expand_to_markers(text, 7, 13)
+        assert start == 7
+        assert end == 13
+
+
+# =============================================================================
+# Process Binary Outputs Tests
+# =============================================================================
+
+
+@pytest.fixture
+def mock_scan():
+    """Patch virus scanner for tests."""
+    with patch(
+        "backend.api.features.chat.tools.binary_output_processor.scan_content_safe",
+        new_callable=AsyncMock,
+    ) as mock:
+        yield mock
+
+
+class TestProcessBinaryOutputs:
+    """Tests for process_binary_outputs function."""
+
+    @pytest.mark.asyncio
+    async def test_detects_embedded_pdf_in_stdout_logs(
+        self, mock_workspace_manager, mock_scan
+    ):
+        """Should detect and replace embedded PDF in stdout_logs."""
+        pdf_b64 = _make_pdf_base64()
+        stdout = f"PDF generated!\n---BASE64_START---\n{pdf_b64}\n---BASE64_END---\n"
+
+        outputs = {"stdout_logs": [stdout]}
+
+        result = await process_binary_outputs(
+            outputs, mock_workspace_manager, "ExecuteCodeBlock"
+        )
+
+        # Should contain workspace reference, not base64
+        assert "workspace://" in result["stdout_logs"][0]
+        assert pdf_b64 not in result["stdout_logs"][0]
+        assert "PDF generated!" in result["stdout_logs"][0]
+        mock_workspace_manager.write_file.assert_called_once()
+
+    @pytest.mark.asyncio
+    async def test_detects_embedded_png_without_markers(
+        self, mock_workspace_manager, mock_scan
+    ):
+        """Should detect embedded PNG even without markers."""
+        png_b64 = _make_png_base64()
+        stdout = f"Image created: {png_b64} done"
+
+        outputs = {"stdout_logs": [stdout]}
+
+        result = await process_binary_outputs(
+            outputs, mock_workspace_manager, "ExecuteCodeBlock"
+        )
+
+        assert "workspace://" in result["stdout_logs"][0]
+        assert "Image created:" in result["stdout_logs"][0]
+        assert "done" in result["stdout_logs"][0]
+
+    @pytest.mark.asyncio
+    async def test_preserves_small_strings(self, mock_workspace_manager, mock_scan):
+        """Should not process small strings."""
+        outputs = {"stdout_logs": ["small output"]}
+
+        result = await process_binary_outputs(
+            outputs, mock_workspace_manager, "TestBlock"
+        )
+
+        assert result["stdout_logs"][0] == "small output"
+        mock_workspace_manager.write_file.assert_not_called()
+
+    @pytest.mark.asyncio
+    async def test_preserves_non_binary_large_strings(
+        self, mock_workspace_manager, mock_scan
+    ):
+        """Should preserve large strings that don't contain valid binary."""
+        large_text = "A" * 5000  # Large string - decodes to nulls, no magic number
+
+        outputs = {"stdout_logs": [large_text]}
+
+        result = await process_binary_outputs(
+            outputs, mock_workspace_manager, "TestBlock"
+        )
+
+        assert result["stdout_logs"][0] == large_text
+        mock_workspace_manager.write_file.assert_not_called()
+
+    @pytest.mark.asyncio
+    async def test_deduplicates_identical_content(
+        self, mock_workspace_manager, mock_scan
+    ):
+        """Should save identical content only once."""
+        pdf_b64 = _make_pdf_base64()
+        stdout1 = f"First: {pdf_b64}"
+        stdout2 = f"Second: {pdf_b64}"
+
+        outputs = {"stdout_logs": [stdout1, stdout2]}
+
+        result = await process_binary_outputs(
+            outputs, mock_workspace_manager, "TestBlock"
+        )
+
+        # Both should have references
+        assert "workspace://" in result["stdout_logs"][0]
+        assert "workspace://" in result["stdout_logs"][1]
+        # But only one write
+        assert mock_workspace_manager.write_file.call_count == 1
+
+    @pytest.mark.asyncio
+    async def test_handles_multiple_binaries_in_one_string(
+        self, mock_workspace_manager, mock_scan
+    ):
+        """Should handle multiple embedded binaries in a single string."""
+        pdf_b64 = _make_pdf_base64()
+        png_b64 = _make_png_base64()
+        stdout = f"PDF: {pdf_b64}\nPNG: {png_b64}"
+
+        outputs = {"stdout_logs": [stdout]}
+
+        result = await process_binary_outputs(
+            outputs, mock_workspace_manager, "TestBlock"
+        )
+
+        # Should have two workspace references
+        assert result["stdout_logs"][0].count("workspace://") == 2
+        assert mock_workspace_manager.write_file.call_count == 2
+
+    @pytest.mark.asyncio
+    async def test_processes_nested_structures(self, mock_workspace_manager, mock_scan):
+        """Should recursively process nested dicts and lists."""
+        pdf_b64 = _make_pdf_base64()
+
+        outputs = {"result": [{"nested": {"deep": f"data: {pdf_b64}"}}]}
+
+        result = await process_binary_outputs(
+            outputs, mock_workspace_manager, "TestBlock"
+        )
+
+        assert "workspace://" in result["result"][0]["nested"]["deep"]
+
+    @pytest.mark.asyncio
+    async def test_graceful_degradation_on_save_failure(
+        self, mock_workspace_manager, mock_scan
+    ):
+        """Should preserve original on save failure."""
+        mock_workspace_manager.write_file = AsyncMock(
+            side_effect=Exception("Storage error")
+        )
+
+        pdf_b64 = _make_pdf_base64()
+        stdout = f"PDF: {pdf_b64}"
+
+        outputs = {"stdout_logs": [stdout]}
+
+        result = await process_binary_outputs(
+            outputs, mock_workspace_manager, "TestBlock"
+        )
+
+        # Should keep original since save failed
+        assert pdf_b64 in result["stdout_logs"][0]
+
+
+# =============================================================================
+# Offset Loop Tests (handling marker bleed-in)
+# =============================================================================
+
+
+class TestOffsetLoopHandling:
+    """Tests for the offset-aligned decoding that handles marker bleed-in."""
+
+    def test_handles_4char_aligned_prefix(self):
+        """Should detect base64 when a 4-char aligned prefix bleeds into match.
+
+        When 'TEST' (4 chars, aligned) bleeds in, offset 4 finds valid base64.
+        """
+        pdf_content = b"%PDF-1.4 " + b"x" * 2000
+        pdf_b64 = base64.b64encode(pdf_content).decode()
+        # 4-char prefix (aligned)
+        with_prefix = f"TEST{pdf_b64}"
+
+        result = _decode_and_validate(with_prefix)
+        assert result is not None
+        content, ext = result
+        assert ext == "pdf"
+        assert content == pdf_content
+
+    def test_handles_8char_aligned_prefix(self):
+        """Should detect base64 when an 8-char prefix bleeds into match."""
+        pdf_content = b"%PDF-1.4 " + b"x" * 2000
+        pdf_b64 = base64.b64encode(pdf_content).decode()
+        # 8-char prefix (aligned)
+        with_prefix = f"TESTTEST{pdf_b64}"
+
+        result = _decode_and_validate(with_prefix)
+        assert result is not None
+        content, ext = result
+        assert ext == "pdf"
+
+    def test_handles_misaligned_prefix(self):
+        """Should handle misaligned prefix by finding a valid aligned offset.
+
+        'START' is 5 chars (misaligned). The loop tries offsets 0, 4, 8...
+        Since characters 0-4 include 'START' which is invalid base64 on its own,
+        we need the full PDF base64 to eventually decode correctly at some offset.
+        """
+        pdf_content = b"%PDF-1.4 " + b"x" * 2000
+        pdf_b64 = base64.b64encode(pdf_content).decode()
+        # 5-char prefix - misaligned, but offset 4 should start mid-'START'
+        # and offset 8 will be past the prefix
+        with_prefix = f"START{pdf_b64}"
+
+        result = _decode_and_validate(with_prefix)
+        # Should find valid PDF at some offset (8 in this case)
+        assert result is not None
+        _, ext = result
+        assert ext == "pdf"
+
+    def test_handles_pdf_base64_start_marker_bleed(self):
+        """Should handle PDF_BASE64_START marker bleeding into regex match.
+
+        This is the real-world case: regex matches 'STARTJVBERi0...' because
+        'START' chars are in the base64 alphabet. Offset loop skips past it.
+        PDF_BASE64_START is 16 chars (4-aligned), so offset 16 finds valid base64.
+        """
+        pdf_content = b"%PDF-1.4 " + b"x" * 2000
+        pdf_b64 = base64.b64encode(pdf_content).decode()
+        # Simulate regex capturing 'PDF_BASE64_START' + base64 together
+        # This happens when there's no delimiter between marker and content
+        with_full_marker = f"PDF_BASE64_START{pdf_b64}"
+
+        result = _decode_and_validate(with_full_marker)
+        assert result is not None
+        _, ext = result
+        assert ext == "pdf"
+
+    def test_clean_base64_works_at_offset_zero(self):
+        """Should detect clean base64 at offset 0 without issues."""
+        pdf_content = b"%PDF-1.4 " + b"x" * 2000
+        pdf_b64 = base64.b64encode(pdf_content).decode()
+
+        result = _decode_and_validate(pdf_b64)
+        assert result is not None
+        content, ext = result
+        assert ext == "pdf"
+        assert content == pdf_content
+
+
+# =============================================================================
+# PDF Marker Tests
+# =============================================================================
+
+
+class TestPdfMarkerExpansion:
+    """Tests for PDF_BASE64_START/END marker handling."""
+
+    def test_expands_pdf_base64_start_marker(self):
+        """Should expand to include PDF_BASE64_START marker."""
+        text = "prefixPDF_BASE64_STARTABCDEF"
+        # Base64 'ABCDEF' is at position 22-28
+        start, end = _expand_to_markers(text, 22, 28)
+        assert text[start:end] == "PDF_BASE64_STARTABCDEF"
+
+    def test_expands_pdf_base64_end_marker(self):
+        """Should expand to include PDF_BASE64_END marker."""
+        text = "ABCDEFPDF_BASE64_ENDsuffix"
+        # Base64 'ABCDEF' is at position 0-6
+        start, end = _expand_to_markers(text, 0, 6)
+        assert text[start:end] == "ABCDEFPDF_BASE64_END"
+
+    def test_expands_both_pdf_markers(self):
+        """Should expand to include both PDF_BASE64_START and END."""
+        text = "xPDF_BASE64_STARTABCDEFPDF_BASE64_ENDy"
+        # Base64 'ABCDEF' is at position 17-23
+        start, end = _expand_to_markers(text, 17, 23)
+        assert text[start:end] == "PDF_BASE64_STARTABCDEFPDF_BASE64_END"
+
+    def test_partial_marker_not_expanded(self):
+        """Should not expand if only partial marker present."""
+        text = "BASE64_STARTABCDEF"  # Missing 'PDF_' prefix
+        start, end = _expand_to_markers(text, 12, 18)
+        # Should not expand since it's not the full marker
+        assert start == 12
+        assert end == 18
+
+    @pytest.mark.asyncio
+    async def test_full_pipeline_with_pdf_markers(self, mock_workspace_manager):
+        """Test full pipeline with PDF_BASE64_START/END markers."""
+        pdf_b64 = _make_pdf_base64()
+        stdout = f"Output: PDF_BASE64_START{pdf_b64}PDF_BASE64_END done"
+
+        outputs = {"stdout_logs": [stdout]}
+
+        with patch(
+            "backend.api.features.chat.tools.binary_output_processor.scan_content_safe",
+            new_callable=AsyncMock,
+        ):
+            result = await process_binary_outputs(
+                outputs, mock_workspace_manager, "TestBlock"
+            )
+
+        # Should have workspace reference
+        assert "workspace://" in result["stdout_logs"][0]
+        # Markers should be consumed along with base64
+        assert "PDF_BASE64_START" not in result["stdout_logs"][0]
+        assert "PDF_BASE64_END" not in result["stdout_logs"][0]
+        # Surrounding text preserved
+        assert "Output:" in result["stdout_logs"][0]
+        assert "done" in result["stdout_logs"][0]
+
+
+# =============================================================================
+# Virus Scanning Tests
+# =============================================================================
+
+
+class TestVirusScanning:
+    """Tests for virus scanning integration."""
+
+    @pytest.mark.asyncio
+    async def test_calls_virus_scanner_before_save(self, mock_workspace_manager):
+        """Should call scan_content_safe before writing file."""
+        pdf_b64 = _make_pdf_base64()
+        stdout = f"PDF: {pdf_b64}"
+        outputs = {"stdout_logs": [stdout]}
+
+        with patch(
+            "backend.api.features.chat.tools.binary_output_processor.scan_content_safe",
+            new_callable=AsyncMock,
+        ) as mock_scan:
+            result = await process_binary_outputs(
+                outputs, mock_workspace_manager, "TestBlock"
+            )
+
+            # Verify scanner was called
+            mock_scan.assert_called_once()
+            # Verify file was written after scan
+            mock_workspace_manager.write_file.assert_called_once()
+            # Verify result has workspace reference
+            assert "workspace://" in result["stdout_logs"][0]
+
+    @pytest.mark.asyncio
+    async def test_virus_scan_failure_preserves_original(self, mock_workspace_manager):
+        """Should preserve original if virus scan fails."""
+        pdf_b64 = _make_pdf_base64()
+        stdout = f"PDF: {pdf_b64}"
+        outputs = {"stdout_logs": [stdout]}
+
+        with patch(
+            "backend.api.features.chat.tools.binary_output_processor.scan_content_safe",
+            new_callable=AsyncMock,
+            side_effect=Exception("Virus detected"),
+        ):
+            result = await process_binary_outputs(
+                outputs, mock_workspace_manager, "TestBlock"
+            )
+
+            # Should keep original since scan failed
+            assert pdf_b64 in result["stdout_logs"][0]
+            # File should not have been written
+            mock_workspace_manager.write_file.assert_not_called()
--- a/autogpt_platform/backend/backend/api/features/chat/tools/utils.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/utils.py
@@ -3,18 +3,17 @@
 import logging
 from typing import Any

+from backend.api.features.library import db as library_db
 from backend.api.features.library import model as library_model
-from backend.data.db_accessors import library_db, store_db
+from backend.api.features.store import db as store_db
 from backend.data.graph import GraphModel
 from backend.data.model import (
-    Credentials,
    CredentialsFieldInfo,
    CredentialsMetaInput,
    HostScopedCredentials,
    OAuth2Credentials,
 )
 from backend.integrations.creds_manager import IntegrationCredentialsManager
-from backend.integrations.providers import ProviderName
 from backend.util.exceptions import NotFoundError

 logger = logging.getLogger(__name__)
@@ -38,14 +37,13 @@ async def fetch_graph_from_store_slug(
    Raises:
        DatabaseError: If there's a database error during lookup.
    """
-    sdb = store_db()
    try:
-        store_agent = await sdb.get_store_agent_details(username, agent_name)
+        store_agent = await store_db.get_store_agent_details(username, agent_name)
    except NotFoundError:
        return None, None

    # Get the graph from store listing version
-    graph = await sdb.get_available_graph(
+    graph = await store_db.get_available_graph(
        store_agent.store_listing_version_id, hide_nodes=False
    )
    return graph, store_agent
@@ -210,13 +208,13 @@ async def get_or_create_library_agent(
    Returns:
        LibraryAgent instance
    """
-    existing = await library_db().get_library_agent_by_graph_id(
+    existing = await library_db.get_library_agent_by_graph_id(
        graph_id=graph.id, user_id=user_id
    )
    if existing:
        return existing

-    library_agents = await library_db().create_library_agent(
+    library_agents = await library_db.create_library_agent(
        graph=graph,
        user_id=user_id,
        create_library_agents_for_sub_graphs=False,
@@ -225,99 +223,6 @@ async def get_or_create_library_agent(
    return library_agents[0]


-async def match_credentials_to_requirements(
-    user_id: str,
-    requirements: dict[str, CredentialsFieldInfo],
-) -> tuple[dict[str, CredentialsMetaInput], list[CredentialsMetaInput]]:
-    """
-    Match user's credentials against a dictionary of credential requirements.
-
-    This is the core matching logic shared by both graph and block credential matching.
-    """
-    matched: dict[str, CredentialsMetaInput] = {}
-    missing: list[CredentialsMetaInput] = []
-
-    if not requirements:
-        return matched, missing
-
-    available_creds = await get_user_credentials(user_id)
-
-    for field_name, field_info in requirements.items():
-        matching_cred = find_matching_credential(available_creds, field_info)
-
-        if matching_cred:
-            try:
-                matched[field_name] = create_credential_meta_from_match(matching_cred)
-            except Exception as e:
-                logger.error(
-                    f"Failed to create CredentialsMetaInput for field '{field_name}': "
-                    f"provider={matching_cred.provider}, type={matching_cred.type}, "
-                    f"credential_id={matching_cred.id}",
-                    exc_info=True,
-                )
-                provider = next(iter(field_info.provider), "unknown")
-                cred_type = next(iter(field_info.supported_types), "api_key")
-                missing.append(
-                    CredentialsMetaInput(
-                        id=field_name,
-                        provider=provider,  # type: ignore
-                        type=cred_type,  # type: ignore
-                        title=f"{field_name} (validation failed: {e})",
-                    )
-                )
-        else:
-            provider = next(iter(field_info.provider), "unknown")
-            cred_type = next(iter(field_info.supported_types), "api_key")
-            missing.append(
-                CredentialsMetaInput(
-                    id=field_name,
-                    provider=provider,  # type: ignore
-                    type=cred_type,  # type: ignore
-                    title=field_name.replace("_", " ").title(),
-                )
-            )
-
-    return matched, missing
-
-
-async def get_user_credentials(user_id: str) -> list[Credentials]:
-    """Get all available credentials for a user."""
-    creds_manager = IntegrationCredentialsManager()
-    return await creds_manager.store.get_all_creds(user_id)
-
-
-def find_matching_credential(
-    available_creds: list[Credentials],
-    field_info: CredentialsFieldInfo,
-) -> Credentials | None:
-    """Find a credential that matches the required provider, type, scopes, and host."""
-    for cred in available_creds:
-        if cred.provider not in field_info.provider:
-            continue
-        if cred.type not in field_info.supported_types:
-            continue
-        if cred.type == "oauth2" and not _credential_has_required_scopes(
-            cred, field_info
-        ):
-            continue
-        if cred.type == "host_scoped" and not _credential_is_for_host(cred, field_info):
-            continue
-        return cred
-    return None
-
-
-def create_credential_meta_from_match(
-    matching_cred: Credentials,
-) -> CredentialsMetaInput:
-    """Create a CredentialsMetaInput from a matched credential."""
-    return CredentialsMetaInput(
-        id=matching_cred.id,
-        provider=matching_cred.provider,  # type: ignore
-        type=matching_cred.type,
-        title=matching_cred.title,
-    )
-
-
 async def match_user_credentials_to_graph(
    user_id: str,
    graph: GraphModel,
@@ -360,7 +265,7 @@ async def match_user_credentials_to_graph(
        _,
        _,
    ) in aggregated_creds.items():
-        # Find first matching credential by provider, type, scopes, and host/URL
+        # Find first matching credential by provider, type, and scopes
        matching_cred = next(
            (
                cred
@@ -375,10 +280,6 @@ async def match_user_credentials_to_graph(
                    cred.type != "host_scoped"
                    or _credential_is_for_host(cred, credential_requirements)
                )
-                and (
-                    cred.provider != ProviderName.MCP
-                    or _credential_is_for_mcp_server(cred, credential_requirements)
-                )
            ),
            None,
        )
@@ -430,6 +331,8 @@ def _credential_has_required_scopes(
    # 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)


@@ -449,22 +352,6 @@ def _credential_is_for_host(
    return credential.matches_url(list(requirements.discriminator_values)[0])


-def _credential_is_for_mcp_server(
-    credential: Credentials,
-    requirements: CredentialsFieldInfo,
-) -> bool:
-    """Check if an MCP OAuth credential matches the required server URL."""
-    if not requirements.discriminator_values:
-        return True
-
-    server_url = (
-        credential.metadata.get("mcp_server_url")
-        if isinstance(credential, OAuth2Credentials)
-        else None
-    )
-    return server_url in requirements.discriminator_values if server_url else False
-
-
 async def check_user_has_required_credentials(
    user_id: str,
    required_credentials: list[CredentialsMetaInput],
--- a/autogpt_platform/backend/backend/api/features/chat/tools/workspace_files.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tools/workspace_files.py
@@ -6,8 +6,8 @@ from typing import Any, Optional

 from pydantic import BaseModel

-from backend.copilot.model import ChatSession
-from backend.data.db_accessors import workspace_db
+from backend.api.features.chat.model import ChatSession
+from backend.data.workspace import get_or_create_workspace
 from backend.util.settings import Config
 from backend.util.virus_scanner import scan_content_safe
 from backend.util.workspace import WorkspaceManager
@@ -88,9 +88,7 @@ class ListWorkspaceFilesTool(BaseTool):
    @property
    def description(self) -> str:
        return (
-            "List files in the user's persistent workspace (cloud storage). "
-            "These files survive across sessions. "
-            "For ephemeral session files, use the SDK Read/Glob tools instead. "
+            "List files in the user's workspace. "
            "Returns file names, paths, sizes, and metadata. "
            "Optionally filter by path prefix."
        )
@@ -148,7 +146,7 @@ class ListWorkspaceFilesTool(BaseTool):
        include_all_sessions: bool = kwargs.get("include_all_sessions", False)

        try:
-            workspace = await workspace_db().get_or_create_workspace(user_id)
+            workspace = await get_or_create_workspace(user_id)
            # Pass session_id for session-scoped file access
            manager = WorkspaceManager(user_id, workspace.id, session_id)

@@ -167,8 +165,8 @@ class ListWorkspaceFilesTool(BaseTool):
                    file_id=f.id,
                    name=f.name,
                    path=f.path,
-                    mime_type=f.mime_type,
-                    size_bytes=f.size_bytes,
+                    mime_type=f.mimeType,
+                    size_bytes=f.sizeBytes,
                )
                for f in files
            ]
@@ -206,9 +204,7 @@ class ReadWorkspaceFileTool(BaseTool):
    @property
    def description(self) -> str:
        return (
-            "Read a file from the user's persistent workspace (cloud storage). "
-            "These files survive across sessions. "
-            "For ephemeral session files, use the SDK Read tool instead. "
+            "Read a file from the user's workspace. "
            "Specify either file_id or path to identify the file. "
            "For small text files, returns content directly. "
            "For large or binary files, returns metadata and a download URL. "
@@ -284,7 +280,7 @@ class ReadWorkspaceFileTool(BaseTool):
            )

        try:
-            workspace = await workspace_db().get_or_create_workspace(user_id)
+            workspace = await get_or_create_workspace(user_id)
            # Pass session_id for session-scoped file access
            manager = WorkspaceManager(user_id, workspace.id, session_id)

@@ -309,8 +305,8 @@ class ReadWorkspaceFileTool(BaseTool):
                target_file_id = file_info.id

            # Decide whether to return inline content or metadata+URL
-            is_small_file = file_info.size_bytes <= self.MAX_INLINE_SIZE_BYTES
-            is_text_file = self._is_text_mime_type(file_info.mime_type)
+            is_small_file = file_info.sizeBytes <= self.MAX_INLINE_SIZE_BYTES
+            is_text_file = self._is_text_mime_type(file_info.mimeType)

            # Return inline content for small text files (unless force_download_url)
            if is_small_file and is_text_file and not force_download_url:
@@ -321,7 +317,7 @@ class ReadWorkspaceFileTool(BaseTool):
                    file_id=file_info.id,
                    name=file_info.name,
                    path=file_info.path,
-                    mime_type=file_info.mime_type,
+                    mime_type=file_info.mimeType,
                    content_base64=content_b64,
                    message=f"Successfully read file: {file_info.name}",
                    session_id=session_id,
@@ -350,11 +346,11 @@ class ReadWorkspaceFileTool(BaseTool):
                file_id=file_info.id,
                name=file_info.name,
                path=file_info.path,
-                mime_type=file_info.mime_type,
-                size_bytes=file_info.size_bytes,
+                mime_type=file_info.mimeType,
+                size_bytes=file_info.sizeBytes,
                download_url=download_url,
                preview=preview,
-                message=f"File: {file_info.name} ({file_info.size_bytes} bytes). Use download_url to retrieve content.",
+                message=f"File: {file_info.name} ({file_info.sizeBytes} bytes). Use download_url to retrieve content.",
                session_id=session_id,
            )

@@ -382,9 +378,7 @@ class WriteWorkspaceFileTool(BaseTool):
    @property
    def description(self) -> str:
        return (
-            "Write or create a file in the user's persistent workspace (cloud storage). "
-            "These files survive across sessions. "
-            "For ephemeral session files, use the SDK Write tool instead. "
+            "Write or create a file in the user's workspace. "
            "Provide the content as a base64-encoded string. "
            f"Maximum file size is {Config().max_file_size_mb}MB. "
            "Files are saved to the current session's folder by default. "
@@ -484,7 +478,7 @@ class WriteWorkspaceFileTool(BaseTool):
            # Virus scan
            await scan_content_safe(content, filename=filename)

-            workspace = await workspace_db().get_or_create_workspace(user_id)
+            workspace = await get_or_create_workspace(user_id)
            # Pass session_id for session-scoped file access
            manager = WorkspaceManager(user_id, workspace.id, session_id)

@@ -500,7 +494,7 @@ class WriteWorkspaceFileTool(BaseTool):
                file_id=file_record.id,
                name=file_record.name,
                path=file_record.path,
-                size_bytes=file_record.size_bytes,
+                size_bytes=file_record.sizeBytes,
                message=f"Successfully wrote file: {file_record.name}",
                session_id=session_id,
            )
@@ -529,7 +523,7 @@ class DeleteWorkspaceFileTool(BaseTool):
    @property
    def description(self) -> str:
        return (
-            "Delete a file from the user's persistent workspace (cloud storage). "
+            "Delete a file from the user's workspace. "
            "Specify either file_id or path to identify the file. "
            "Paths are scoped to the current session by default. "
            "Use /sessions/<session_id>/... for cross-session access."
@@ -583,7 +577,7 @@ class DeleteWorkspaceFileTool(BaseTool):
            )

        try:
-            workspace = await workspace_db().get_or_create_workspace(user_id)
+            workspace = await get_or_create_workspace(user_id)
            # Pass session_id for session-scoped file access
            manager = WorkspaceManager(user_id, workspace.id, session_id)

--- a/autogpt_platform/backend/backend/api/features/chat/tracking.py
+++ b/autogpt_platform/backend/backend/api/features/chat/tracking.py
--- a/autogpt_platform/backend/backend/api/features/integrations/router.py
+++ b/autogpt_platform/backend/backend/api/features/integrations/router.py
@@ -1,7 +1,7 @@
 import asyncio
 import logging
 from datetime import datetime, timedelta, timezone
-from typing import TYPE_CHECKING, Annotated, Any, List, Literal
+from typing import TYPE_CHECKING, Annotated, List, Literal

 from autogpt_libs.auth import get_user_id
 from fastapi import (
@@ -14,7 +14,7 @@ from fastapi import (
    Security,
    status,
 )
-from pydantic import BaseModel, Field, SecretStr, model_validator
+from pydantic import BaseModel, Field, SecretStr
 from starlette.status import HTTP_500_INTERNAL_SERVER_ERROR, HTTP_502_BAD_GATEWAY

 from backend.api.features.library.db import set_preset_webhook, update_preset
@@ -39,11 +39,7 @@ from backend.data.onboarding import OnboardingStep, complete_onboarding_step
 from backend.data.user import get_user_integrations
 from backend.executor.utils import add_graph_execution
 from backend.integrations.ayrshare import AyrshareClient, SocialPlatform
-from backend.integrations.credentials_store import provider_matches
-from backend.integrations.creds_manager import (
-    IntegrationCredentialsManager,
-    create_mcp_oauth_handler,
-)
+from backend.integrations.creds_manager import IntegrationCredentialsManager
 from backend.integrations.oauth import CREDENTIALS_BY_PROVIDER, HANDLERS_BY_NAME
 from backend.integrations.providers import ProviderName
 from backend.integrations.webhooks import get_webhook_manager
@@ -106,37 +102,9 @@ class CredentialsMetaResponse(BaseModel):
    scopes: list[str] | None
    username: str | None
    host: str | None = Field(
-        default=None,
-        description="Host pattern for host-scoped or MCP server URL for MCP credentials",
+        default=None, description="Host pattern for host-scoped credentials"
    )

-    @model_validator(mode="before")
-    @classmethod
-    def _normalize_provider(cls, data: Any) -> Any:
-        """Fix ``ProviderName.X`` format from Python 3.13 ``str(Enum)`` bug."""
-        if isinstance(data, dict):
-            prov = data.get("provider", "")
-            if isinstance(prov, str) and prov.startswith("ProviderName."):
-                member = prov.removeprefix("ProviderName.")
-                try:
-                    data = {**data, "provider": ProviderName[member].value}
-                except KeyError:
-                    pass
-        return data
-
-    @staticmethod
-    def get_host(cred: Credentials) -> str | None:
-        """Extract host from credential: HostScoped host or MCP server URL."""
-        if isinstance(cred, HostScopedCredentials):
-            return cred.host
-        if isinstance(cred, OAuth2Credentials) and cred.provider in (
-            ProviderName.MCP,
-            ProviderName.MCP.value,
-            "ProviderName.MCP",
-        ):
-            return (cred.metadata or {}).get("mcp_server_url")
-        return None
-

@router.post("/{provider}/callback", summary="Exchange OAuth code for tokens")
 async def callback(
@@ -211,7 +179,9 @@ async def callback(
        title=credentials.title,
        scopes=credentials.scopes,
        username=credentials.username,
-        host=(CredentialsMetaResponse.get_host(credentials)),
+        host=(
+            credentials.host if isinstance(credentials, HostScopedCredentials) else None
+        ),
    )


@@ -229,7 +199,7 @@ async def list_credentials(
            title=cred.title,
            scopes=cred.scopes if isinstance(cred, OAuth2Credentials) else None,
            username=cred.username if isinstance(cred, OAuth2Credentials) else None,
-            host=CredentialsMetaResponse.get_host(cred),
+            host=cred.host if isinstance(cred, HostScopedCredentials) else None,
        )
        for cred in credentials
    ]
@@ -252,7 +222,7 @@ async def list_credentials_by_provider(
            title=cred.title,
            scopes=cred.scopes if isinstance(cred, OAuth2Credentials) else None,
            username=cred.username if isinstance(cred, OAuth2Credentials) else None,
-            host=CredentialsMetaResponse.get_host(cred),
+            host=cred.host if isinstance(cred, HostScopedCredentials) else None,
        )
        for cred in credentials
    ]
@@ -352,11 +322,7 @@ async def delete_credentials(

    tokens_revoked = None
    if isinstance(creds, OAuth2Credentials):
-        if provider_matches(provider.value, ProviderName.MCP.value):
-            # MCP uses dynamic per-server OAuth — create handler from metadata
-            handler = create_mcp_oauth_handler(creds)
-        else:
-            handler = _get_provider_oauth_handler(request, provider)
+        handler = _get_provider_oauth_handler(request, provider)
        tokens_revoked = await handler.revoke_tokens(creds)

    return CredentialsDeletionResponse(revoked=tokens_revoked)
--- a/autogpt_platform/backend/backend/api/features/library/db.py
+++ b/autogpt_platform/backend/backend/api/features/library/db.py
@@ -12,11 +12,12 @@ import backend.api.features.store.image_gen as store_image_gen
 import backend.api.features.store.media as store_media
 import backend.data.graph as graph_db
 import backend.data.integrations as integrations_db
+from backend.data.block import BlockInput
 from backend.data.db import transaction
 from backend.data.execution import get_graph_execution
 from backend.data.graph import GraphSettings
 from backend.data.includes import AGENT_PRESET_INCLUDE, library_agent_include
-from backend.data.model import CredentialsMetaInput, GraphInput
+from backend.data.model import CredentialsMetaInput
 from backend.integrations.creds_manager import IntegrationCredentialsManager
 from backend.integrations.webhooks.graph_lifecycle_hooks import (
    on_graph_activate,
@@ -1129,7 +1130,7 @@ async def create_preset_from_graph_execution(
 async def update_preset(
    user_id: str,
    preset_id: str,
-    inputs: Optional[GraphInput] = None,
+    inputs: Optional[BlockInput] = None,
    credentials: Optional[dict[str, CredentialsMetaInput]] = None,
    name: Optional[str] = None,
    description: Optional[str] = None,
--- a/autogpt_platform/backend/backend/api/features/library/model.py
+++ b/autogpt_platform/backend/backend/api/features/library/model.py
@@ -6,12 +6,9 @@ import prisma.enums
 import prisma.models
 import pydantic

+from backend.data.block import BlockInput
 from backend.data.graph import GraphModel, GraphSettings, GraphTriggerInfo
-from backend.data.model import (
-    CredentialsMetaInput,
-    GraphInput,
-    is_credentials_field_name,
-)
+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

@@ -326,7 +323,7 @@ class LibraryAgentPresetCreatable(pydantic.BaseModel):
    graph_id: str
    graph_version: int

-    inputs: GraphInput
+    inputs: BlockInput
    credentials: dict[str, CredentialsMetaInput]

    name: str
@@ -355,7 +352,7 @@ class LibraryAgentPresetUpdatable(pydantic.BaseModel):
    Request model used when updating a preset for a library agent.
    """

-    inputs: Optional[GraphInput] = None
+    inputs: Optional[BlockInput] = None
    credentials: Optional[dict[str, CredentialsMetaInput]] = None

    name: Optional[str] = None
@@ -398,7 +395,7 @@ class LibraryAgentPreset(LibraryAgentPresetCreatable):
                "Webhook must be included in AgentPreset query when webhookId is set"
            )

-        input_data: GraphInput = {}
+        input_data: BlockInput = {}
        input_credentials: dict[str, CredentialsMetaInput] = {}

        for preset_input in preset.InputPresets:
--- a/autogpt_platform/backend/backend/api/features/mcp/init.py
+++ b/autogpt_platform/backend/backend/api/features/mcp/init.py
--- a/autogpt_platform/backend/backend/api/features/mcp/routes.py
+++ b/autogpt_platform/backend/backend/api/features/mcp/routes.py
@@ -1,404 +0,0 @@
-"""
-MCP (Model Context Protocol) API routes.
-
-Provides endpoints for MCP tool discovery and OAuth authentication so the
-frontend can list available tools on an MCP server before placing a block.
-"""
-
-import logging
-from typing import Annotated, Any
-from urllib.parse import urlparse
-
-import fastapi
-from autogpt_libs.auth import get_user_id
-from fastapi import Security
-from pydantic import BaseModel, Field
-
-from backend.api.features.integrations.router import CredentialsMetaResponse
-from backend.blocks.mcp.client import MCPClient, MCPClientError
-from backend.blocks.mcp.oauth import MCPOAuthHandler
-from backend.data.model import OAuth2Credentials
-from backend.integrations.creds_manager import IntegrationCredentialsManager
-from backend.integrations.providers import ProviderName
-from backend.util.request import HTTPClientError, Requests
-from backend.util.settings import Settings
-
-logger = logging.getLogger(__name__)
-
-settings = Settings()
-router = fastapi.APIRouter(tags=["mcp"])
-creds_manager = IntegrationCredentialsManager()
-
-
-# ====================== Tool Discovery ====================== #
-
-
-class DiscoverToolsRequest(BaseModel):
-    """Request to discover tools on an MCP server."""
-
-    server_url: str = Field(description="URL of the MCP server")
-    auth_token: str | None = Field(
-        default=None,
-        description="Optional Bearer token for authenticated MCP servers",
-    )
-
-
-class MCPToolResponse(BaseModel):
-    """A single MCP tool returned by discovery."""
-
-    name: str
-    description: str
-    input_schema: dict[str, Any]
-
-
-class DiscoverToolsResponse(BaseModel):
-    """Response containing the list of tools available on an MCP server."""
-
-    tools: list[MCPToolResponse]
-    server_name: str | None = None
-    protocol_version: str | None = None
-
-
-@router.post(
-    "/discover-tools",
-    summary="Discover available tools on an MCP server",
-    response_model=DiscoverToolsResponse,
-)
-async def discover_tools(
-    request: DiscoverToolsRequest,
-    user_id: Annotated[str, Security(get_user_id)],
-) -> DiscoverToolsResponse:
-    """
-    Connect to an MCP server and return its available tools.
-
-    If the user has a stored MCP credential for this server URL, it will be
-    used automatically — no need to pass an explicit auth token.
-    """
-    auth_token = request.auth_token
-
-    # Auto-use stored MCP credential when no explicit token is provided.
-    if not auth_token:
-        mcp_creds = await creds_manager.store.get_creds_by_provider(
-            user_id, ProviderName.MCP.value
-        )
-        # Find the freshest credential for this server URL
-        best_cred: OAuth2Credentials | None = None
-        for cred in mcp_creds:
-            if (
-                isinstance(cred, OAuth2Credentials)
-                and (cred.metadata or {}).get("mcp_server_url") == request.server_url
-            ):
-                if best_cred is None or (
-                    (cred.access_token_expires_at or 0)
-                    > (best_cred.access_token_expires_at or 0)
-                ):
-                    best_cred = cred
-        if best_cred:
-            # Refresh the token if expired before using it
-            best_cred = await creds_manager.refresh_if_needed(user_id, best_cred)
-            logger.info(
-                f"Using MCP credential {best_cred.id} for {request.server_url}, "
-                f"expires_at={best_cred.access_token_expires_at}"
-            )
-            auth_token = best_cred.access_token.get_secret_value()
-
-    client = MCPClient(request.server_url, auth_token=auth_token)
-
-    try:
-        init_result = await client.initialize()
-        tools = await client.list_tools()
-    except HTTPClientError as e:
-        if e.status_code in (401, 403):
-            raise fastapi.HTTPException(
-                status_code=401,
-                detail="This MCP server requires authentication. "
-                "Please provide a valid auth token.",
-            )
-        raise fastapi.HTTPException(status_code=502, detail=str(e))
-    except MCPClientError as e:
-        raise fastapi.HTTPException(status_code=502, detail=str(e))
-    except Exception as e:
-        raise fastapi.HTTPException(
-            status_code=502,
-            detail=f"Failed to connect to MCP server: {e}",
-        )
-
-    return DiscoverToolsResponse(
-        tools=[
-            MCPToolResponse(
-                name=t.name,
-                description=t.description,
-                input_schema=t.input_schema,
-            )
-            for t in tools
-        ],
-        server_name=(
-            init_result.get("serverInfo", {}).get("name")
-            or urlparse(request.server_url).hostname
-            or "MCP"
-        ),
-        protocol_version=init_result.get("protocolVersion"),
-    )
-
-
-# ======================== OAuth Flow ======================== #
-
-
-class MCPOAuthLoginRequest(BaseModel):
-    """Request to start an OAuth flow for an MCP server."""
-
-    server_url: str = Field(description="URL of the MCP server that requires OAuth")
-
-
-class MCPOAuthLoginResponse(BaseModel):
-    """Response with the OAuth login URL for the user to authenticate."""
-
-    login_url: str
-    state_token: str
-
-
-@router.post(
-    "/oauth/login",
-    summary="Initiate OAuth login for an MCP server",
-)
-async def mcp_oauth_login(
-    request: MCPOAuthLoginRequest,
-    user_id: Annotated[str, Security(get_user_id)],
-) -> MCPOAuthLoginResponse:
-    """
-    Discover OAuth metadata from the MCP server and return a login URL.
-
-    1. Discovers the protected-resource metadata (RFC 9728)
-    2. Fetches the authorization server metadata (RFC 8414)
-    3. Performs Dynamic Client Registration (RFC 7591) if available
-    4. Returns the authorization URL for the frontend to open in a popup
-    """
-    client = MCPClient(request.server_url)
-
-    # Step 1: Discover protected-resource metadata (RFC 9728)
-    protected_resource = await client.discover_auth()
-
-    metadata: dict[str, Any] | None = None
-
-    if protected_resource and protected_resource.get("authorization_servers"):
-        auth_server_url = protected_resource["authorization_servers"][0]
-        resource_url = protected_resource.get("resource", request.server_url)
-
-        # Step 2a: Discover auth-server metadata (RFC 8414)
-        metadata = await client.discover_auth_server_metadata(auth_server_url)
-    else:
-        # Fallback: Some MCP servers (e.g. Linear) are their own auth server
-        # and serve OAuth metadata directly without protected-resource metadata.
-        # Don't assume a resource_url — omitting it lets the auth server choose
-        # the correct audience for the token (RFC 8707 resource is optional).
-        resource_url = None
-        metadata = await client.discover_auth_server_metadata(request.server_url)
-
-    if (
-        not metadata
-        or "authorization_endpoint" not in metadata
-        or "token_endpoint" not in metadata
-    ):
-        raise fastapi.HTTPException(
-            status_code=400,
-            detail="This MCP server does not advertise OAuth support. "
-            "You may need to provide an auth token manually.",
-        )
-
-    authorize_url = metadata["authorization_endpoint"]
-    token_url = metadata["token_endpoint"]
-    registration_endpoint = metadata.get("registration_endpoint")
-    revoke_url = metadata.get("revocation_endpoint")
-
-    # Step 3: Dynamic Client Registration (RFC 7591) if available
-    frontend_base_url = settings.config.frontend_base_url
-    if not frontend_base_url:
-        raise fastapi.HTTPException(
-            status_code=500,
-            detail="Frontend base URL is not configured.",
-        )
-    redirect_uri = f"{frontend_base_url}/auth/integrations/mcp_callback"
-
-    client_id = ""
-    client_secret = ""
-    if registration_endpoint:
-        reg_result = await _register_mcp_client(
-            registration_endpoint, redirect_uri, request.server_url
-        )
-        if reg_result:
-            client_id = reg_result.get("client_id", "")
-            client_secret = reg_result.get("client_secret", "")
-
-    if not client_id:
-        client_id = "autogpt-platform"
-
-    # Step 4: Store state token with OAuth metadata for the callback
-    scopes = (protected_resource or {}).get("scopes_supported") or metadata.get(
-        "scopes_supported", []
-    )
-    state_token, code_challenge = await creds_manager.store.store_state_token(
-        user_id,
-        ProviderName.MCP.value,
-        scopes,
-        state_metadata={
-            "authorize_url": authorize_url,
-            "token_url": token_url,
-            "revoke_url": revoke_url,
-            "resource_url": resource_url,
-            "server_url": request.server_url,
-            "client_id": client_id,
-            "client_secret": client_secret,
-        },
-    )
-
-    # Step 5: Build and return the login URL
-    handler = MCPOAuthHandler(
-        client_id=client_id,
-        client_secret=client_secret,
-        redirect_uri=redirect_uri,
-        authorize_url=authorize_url,
-        token_url=token_url,
-        resource_url=resource_url,
-    )
-    login_url = handler.get_login_url(
-        scopes, state_token, code_challenge=code_challenge
-    )
-
-    return MCPOAuthLoginResponse(login_url=login_url, state_token=state_token)
-
-
-class MCPOAuthCallbackRequest(BaseModel):
-    """Request to exchange an OAuth code for tokens."""
-
-    code: str = Field(description="Authorization code from OAuth callback")
-    state_token: str = Field(description="State token for CSRF verification")
-
-
-class MCPOAuthCallbackResponse(BaseModel):
-    """Response after successfully storing OAuth credentials."""
-
-    credential_id: str
-
-
-@router.post(
-    "/oauth/callback",
-    summary="Exchange OAuth code for MCP tokens",
-)
-async def mcp_oauth_callback(
-    request: MCPOAuthCallbackRequest,
-    user_id: Annotated[str, Security(get_user_id)],
-) -> CredentialsMetaResponse:
-    """
-    Exchange the authorization code for tokens and store the credential.
-
-    The frontend calls this after receiving the OAuth code from the popup.
-    On success, subsequent ``/discover-tools`` calls for the same server URL
-    will automatically use the stored credential.
-    """
-    valid_state = await creds_manager.store.verify_state_token(
-        user_id, request.state_token, ProviderName.MCP.value
-    )
-    if not valid_state:
-        raise fastapi.HTTPException(
-            status_code=400,
-            detail="Invalid or expired state token.",
-        )
-
-    meta = valid_state.state_metadata
-    frontend_base_url = settings.config.frontend_base_url
-    if not frontend_base_url:
-        raise fastapi.HTTPException(
-            status_code=500,
-            detail="Frontend base URL is not configured.",
-        )
-    redirect_uri = f"{frontend_base_url}/auth/integrations/mcp_callback"
-
-    handler = MCPOAuthHandler(
-        client_id=meta["client_id"],
-        client_secret=meta.get("client_secret", ""),
-        redirect_uri=redirect_uri,
-        authorize_url=meta["authorize_url"],
-        token_url=meta["token_url"],
-        revoke_url=meta.get("revoke_url"),
-        resource_url=meta.get("resource_url"),
-    )
-
-    try:
-        credentials = await handler.exchange_code_for_tokens(
-            request.code, valid_state.scopes, valid_state.code_verifier
-        )
-    except Exception as e:
-        raise fastapi.HTTPException(
-            status_code=400,
-            detail=f"OAuth token exchange failed: {e}",
-        )
-
-    # Enrich credential metadata for future lookup and token refresh
-    if credentials.metadata is None:
-        credentials.metadata = {}
-    credentials.metadata["mcp_server_url"] = meta["server_url"]
-    credentials.metadata["mcp_client_id"] = meta["client_id"]
-    credentials.metadata["mcp_client_secret"] = meta.get("client_secret", "")
-    credentials.metadata["mcp_token_url"] = meta["token_url"]
-    credentials.metadata["mcp_resource_url"] = meta.get("resource_url", "")
-
-    hostname = urlparse(meta["server_url"]).hostname or meta["server_url"]
-    credentials.title = f"MCP: {hostname}"
-
-    # Remove old MCP credentials for the same server to prevent stale token buildup.
-    try:
-        old_creds = await creds_manager.store.get_creds_by_provider(
-            user_id, ProviderName.MCP.value
-        )
-        for old in old_creds:
-            if (
-                isinstance(old, OAuth2Credentials)
-                and (old.metadata or {}).get("mcp_server_url") == meta["server_url"]
-            ):
-                await creds_manager.store.delete_creds_by_id(user_id, old.id)
-                logger.info(
-                    f"Removed old MCP credential {old.id} for {meta['server_url']}"
-                )
-    except Exception:
-        logger.debug("Could not clean up old MCP credentials", exc_info=True)
-
-    await creds_manager.create(user_id, credentials)
-
-    return CredentialsMetaResponse(
-        id=credentials.id,
-        provider=credentials.provider,
-        type=credentials.type,
-        title=credentials.title,
-        scopes=credentials.scopes,
-        username=credentials.username,
-        host=credentials.metadata.get("mcp_server_url"),
-    )
-
-
-# ======================== Helpers ======================== #
-
-
-async def _register_mcp_client(
-    registration_endpoint: str,
-    redirect_uri: str,
-    server_url: str,
-) -> dict[str, Any] | None:
-    """Attempt Dynamic Client Registration (RFC 7591) with an MCP auth server."""
-    try:
-        response = await Requests(raise_for_status=True).post(
-            registration_endpoint,
-            json={
-                "client_name": "AutoGPT Platform",
-                "redirect_uris": [redirect_uri],
-                "grant_types": ["authorization_code"],
-                "response_types": ["code"],
-                "token_endpoint_auth_method": "client_secret_post",
-            },
-        )
-        data = response.json()
-        if isinstance(data, dict) and "client_id" in data:
-            return data
-        return None
-    except Exception as e:
-        logger.warning(f"Dynamic client registration failed for {server_url}: {e}")
-        return None
--- a/autogpt_platform/backend/backend/api/features/mcp/test_routes.py
+++ b/autogpt_platform/backend/backend/api/features/mcp/test_routes.py
@@ -1,436 +0,0 @@
-"""Tests for MCP API routes.
-
-Uses httpx.AsyncClient with ASGITransport instead of fastapi.testclient.TestClient
-to avoid creating blocking portals that can corrupt pytest-asyncio's session event loop.
-"""
-
-from unittest.mock import AsyncMock, patch
-
-import fastapi
-import httpx
-import pytest
-import pytest_asyncio
-from autogpt_libs.auth import get_user_id
-
-from backend.api.features.mcp.routes import router
-from backend.blocks.mcp.client import MCPClientError, MCPTool
-from backend.util.request import HTTPClientError
-
-app = fastapi.FastAPI()
-app.include_router(router)
-app.dependency_overrides[get_user_id] = lambda: "test-user-id"
-
-
-@pytest_asyncio.fixture(scope="module")
-async def client():
-    transport = httpx.ASGITransport(app=app)
-    async with httpx.AsyncClient(transport=transport, base_url="http://test") as c:
-        yield c
-
-
-class TestDiscoverTools:
-    @pytest.mark.asyncio(loop_scope="session")
-    async def test_discover_tools_success(self, client):
-        mock_tools = [
-            MCPTool(
-                name="get_weather",
-                description="Get weather for a city",
-                input_schema={
-                    "type": "object",
-                    "properties": {"city": {"type": "string"}},
-                    "required": ["city"],
-                },
-            ),
-            MCPTool(
-                name="add_numbers",
-                description="Add two numbers",
-                input_schema={
-                    "type": "object",
-                    "properties": {
-                        "a": {"type": "number"},
-                        "b": {"type": "number"},
-                    },
-                },
-            ),
-        ]
-
-        with (
-            patch("backend.api.features.mcp.routes.MCPClient") as MockClient,
-            patch("backend.api.features.mcp.routes.creds_manager") as mock_cm,
-        ):
-            mock_cm.store.get_creds_by_provider = AsyncMock(return_value=[])
-            instance = MockClient.return_value
-            instance.initialize = AsyncMock(
-                return_value={
-                    "protocolVersion": "2025-03-26",
-                    "serverInfo": {"name": "test-server"},
-                }
-            )
-            instance.list_tools = AsyncMock(return_value=mock_tools)
-
-            response = await client.post(
-                "/discover-tools",
-                json={"server_url": "https://mcp.example.com/mcp"},
-            )
-
-        assert response.status_code == 200
-        data = response.json()
-        assert len(data["tools"]) == 2
-        assert data["tools"][0]["name"] == "get_weather"
-        assert data["tools"][1]["name"] == "add_numbers"
-        assert data["server_name"] == "test-server"
-        assert data["protocol_version"] == "2025-03-26"
-
-    @pytest.mark.asyncio(loop_scope="session")
-    async def test_discover_tools_with_auth_token(self, client):
-        with patch("backend.api.features.mcp.routes.MCPClient") as MockClient:
-            instance = MockClient.return_value
-            instance.initialize = AsyncMock(
-                return_value={"serverInfo": {}, "protocolVersion": "2025-03-26"}
-            )
-            instance.list_tools = AsyncMock(return_value=[])
-
-            response = await client.post(
-                "/discover-tools",
-                json={
-                    "server_url": "https://mcp.example.com/mcp",
-                    "auth_token": "my-secret-token",
-                },
-            )
-
-        assert response.status_code == 200
-        MockClient.assert_called_once_with(
-            "https://mcp.example.com/mcp",
-            auth_token="my-secret-token",
-        )
-
-    @pytest.mark.asyncio(loop_scope="session")
-    async def test_discover_tools_auto_uses_stored_credential(self, client):
-        """When no explicit token is given, stored MCP credentials are used."""
-        from pydantic import SecretStr
-
-        from backend.data.model import OAuth2Credentials
-
-        stored_cred = OAuth2Credentials(
-            provider="mcp",
-            title="MCP: example.com",
-            access_token=SecretStr("stored-token-123"),
-            refresh_token=None,
-            access_token_expires_at=None,
-            refresh_token_expires_at=None,
-            scopes=[],
-            metadata={"mcp_server_url": "https://mcp.example.com/mcp"},
-        )
-
-        with (
-            patch("backend.api.features.mcp.routes.MCPClient") as MockClient,
-            patch("backend.api.features.mcp.routes.creds_manager") as mock_cm,
-        ):
-            mock_cm.store.get_creds_by_provider = AsyncMock(return_value=[stored_cred])
-            mock_cm.refresh_if_needed = AsyncMock(return_value=stored_cred)
-            instance = MockClient.return_value
-            instance.initialize = AsyncMock(
-                return_value={"serverInfo": {}, "protocolVersion": "2025-03-26"}
-            )
-            instance.list_tools = AsyncMock(return_value=[])
-
-            response = await client.post(
-                "/discover-tools",
-                json={"server_url": "https://mcp.example.com/mcp"},
-            )
-
-        assert response.status_code == 200
-        MockClient.assert_called_once_with(
-            "https://mcp.example.com/mcp",
-            auth_token="stored-token-123",
-        )
-
-    @pytest.mark.asyncio(loop_scope="session")
-    async def test_discover_tools_mcp_error(self, client):
-        with (
-            patch("backend.api.features.mcp.routes.MCPClient") as MockClient,
-            patch("backend.api.features.mcp.routes.creds_manager") as mock_cm,
-        ):
-            mock_cm.store.get_creds_by_provider = AsyncMock(return_value=[])
-            instance = MockClient.return_value
-            instance.initialize = AsyncMock(
-                side_effect=MCPClientError("Connection refused")
-            )
-
-            response = await client.post(
-                "/discover-tools",
-                json={"server_url": "https://bad-server.example.com/mcp"},
-            )
-
-        assert response.status_code == 502
-        assert "Connection refused" in response.json()["detail"]
-
-    @pytest.mark.asyncio(loop_scope="session")
-    async def test_discover_tools_generic_error(self, client):
-        with (
-            patch("backend.api.features.mcp.routes.MCPClient") as MockClient,
-            patch("backend.api.features.mcp.routes.creds_manager") as mock_cm,
-        ):
-            mock_cm.store.get_creds_by_provider = AsyncMock(return_value=[])
-            instance = MockClient.return_value
-            instance.initialize = AsyncMock(side_effect=Exception("Network timeout"))
-
-            response = await client.post(
-                "/discover-tools",
-                json={"server_url": "https://timeout.example.com/mcp"},
-            )
-
-        assert response.status_code == 502
-        assert "Failed to connect" in response.json()["detail"]
-
-    @pytest.mark.asyncio(loop_scope="session")
-    async def test_discover_tools_auth_required(self, client):
-        with (
-            patch("backend.api.features.mcp.routes.MCPClient") as MockClient,
-            patch("backend.api.features.mcp.routes.creds_manager") as mock_cm,
-        ):
-            mock_cm.store.get_creds_by_provider = AsyncMock(return_value=[])
-            instance = MockClient.return_value
-            instance.initialize = AsyncMock(
-                side_effect=HTTPClientError("HTTP 401 Error: Unauthorized", 401)
-            )
-
-            response = await client.post(
-                "/discover-tools",
-                json={"server_url": "https://auth-server.example.com/mcp"},
-            )
-
-        assert response.status_code == 401
-        assert "requires authentication" in response.json()["detail"]
-
-    @pytest.mark.asyncio(loop_scope="session")
-    async def test_discover_tools_forbidden(self, client):
-        with (
-            patch("backend.api.features.mcp.routes.MCPClient") as MockClient,
-            patch("backend.api.features.mcp.routes.creds_manager") as mock_cm,
-        ):
-            mock_cm.store.get_creds_by_provider = AsyncMock(return_value=[])
-            instance = MockClient.return_value
-            instance.initialize = AsyncMock(
-                side_effect=HTTPClientError("HTTP 403 Error: Forbidden", 403)
-            )
-
-            response = await client.post(
-                "/discover-tools",
-                json={"server_url": "https://auth-server.example.com/mcp"},
-            )
-
-        assert response.status_code == 401
-        assert "requires authentication" in response.json()["detail"]
-
-    @pytest.mark.asyncio(loop_scope="session")
-    async def test_discover_tools_missing_url(self, client):
-        response = await client.post("/discover-tools", json={})
-        assert response.status_code == 422
-
-
-class TestOAuthLogin:
-    @pytest.mark.asyncio(loop_scope="session")
-    async def test_oauth_login_success(self, client):
-        with (
-            patch("backend.api.features.mcp.routes.MCPClient") as MockClient,
-            patch("backend.api.features.mcp.routes.creds_manager") as mock_cm,
-            patch("backend.api.features.mcp.routes.settings") as mock_settings,
-            patch(
-                "backend.api.features.mcp.routes._register_mcp_client"
-            ) as mock_register,
-        ):
-            instance = MockClient.return_value
-            instance.discover_auth = AsyncMock(
-                return_value={
-                    "authorization_servers": ["https://auth.sentry.io"],
-                    "resource": "https://mcp.sentry.dev/mcp",
-                    "scopes_supported": ["openid"],
-                }
-            )
-            instance.discover_auth_server_metadata = AsyncMock(
-                return_value={
-                    "authorization_endpoint": "https://auth.sentry.io/authorize",
-                    "token_endpoint": "https://auth.sentry.io/token",
-                    "registration_endpoint": "https://auth.sentry.io/register",
-                }
-            )
-            mock_register.return_value = {
-                "client_id": "registered-client-id",
-                "client_secret": "registered-secret",
-            }
-            mock_cm.store.store_state_token = AsyncMock(
-                return_value=("state-token-123", "code-challenge-abc")
-            )
-            mock_settings.config.frontend_base_url = "http://localhost:3000"
-
-            response = await client.post(
-                "/oauth/login",
-                json={"server_url": "https://mcp.sentry.dev/mcp"},
-            )
-
-        assert response.status_code == 200
-        data = response.json()
-        assert "login_url" in data
-        assert data["state_token"] == "state-token-123"
-        assert "auth.sentry.io/authorize" in data["login_url"]
-        assert "registered-client-id" in data["login_url"]
-
-    @pytest.mark.asyncio(loop_scope="session")
-    async def test_oauth_login_no_oauth_support(self, client):
-        with patch("backend.api.features.mcp.routes.MCPClient") as MockClient:
-            instance = MockClient.return_value
-            instance.discover_auth = AsyncMock(return_value=None)
-            instance.discover_auth_server_metadata = AsyncMock(return_value=None)
-
-            response = await client.post(
-                "/oauth/login",
-                json={"server_url": "https://simple-server.example.com/mcp"},
-            )
-
-        assert response.status_code == 400
-        assert "does not advertise OAuth" in response.json()["detail"]
-
-    @pytest.mark.asyncio(loop_scope="session")
-    async def test_oauth_login_fallback_to_public_client(self, client):
-        """When DCR is unavailable, falls back to default public client ID."""
-        with (
-            patch("backend.api.features.mcp.routes.MCPClient") as MockClient,
-            patch("backend.api.features.mcp.routes.creds_manager") as mock_cm,
-            patch("backend.api.features.mcp.routes.settings") as mock_settings,
-        ):
-            instance = MockClient.return_value
-            instance.discover_auth = AsyncMock(
-                return_value={
-                    "authorization_servers": ["https://auth.example.com"],
-                    "resource": "https://mcp.example.com/mcp",
-                }
-            )
-            instance.discover_auth_server_metadata = AsyncMock(
-                return_value={
-                    "authorization_endpoint": "https://auth.example.com/authorize",
-                    "token_endpoint": "https://auth.example.com/token",
-                    # No registration_endpoint
-                }
-            )
-            mock_cm.store.store_state_token = AsyncMock(
-                return_value=("state-abc", "challenge-xyz")
-            )
-            mock_settings.config.frontend_base_url = "http://localhost:3000"
-
-            response = await client.post(
-                "/oauth/login",
-                json={"server_url": "https://mcp.example.com/mcp"},
-            )
-
-        assert response.status_code == 200
-        data = response.json()
-        assert "autogpt-platform" in data["login_url"]
-
-
-class TestOAuthCallback:
-    @pytest.mark.asyncio(loop_scope="session")
-    async def test_oauth_callback_success(self, client):
-        from pydantic import SecretStr
-
-        from backend.data.model import OAuth2Credentials
-
-        mock_creds = OAuth2Credentials(
-            provider="mcp",
-            title=None,
-            access_token=SecretStr("access-token-xyz"),
-            refresh_token=None,
-            access_token_expires_at=None,
-            refresh_token_expires_at=None,
-            scopes=[],
-            metadata={
-                "mcp_token_url": "https://auth.sentry.io/token",
-                "mcp_resource_url": "https://mcp.sentry.dev/mcp",
-            },
-        )
-
-        with (
-            patch("backend.api.features.mcp.routes.creds_manager") as mock_cm,
-            patch("backend.api.features.mcp.routes.settings") as mock_settings,
-            patch("backend.api.features.mcp.routes.MCPOAuthHandler") as MockHandler,
-        ):
-            mock_settings.config.frontend_base_url = "http://localhost:3000"
-
-            # Mock state verification
-            mock_state = AsyncMock()
-            mock_state.state_metadata = {
-                "authorize_url": "https://auth.sentry.io/authorize",
-                "token_url": "https://auth.sentry.io/token",
-                "client_id": "test-client-id",
-                "client_secret": "test-secret",
-                "server_url": "https://mcp.sentry.dev/mcp",
-            }
-            mock_state.scopes = ["openid"]
-            mock_state.code_verifier = "verifier-123"
-            mock_cm.store.verify_state_token = AsyncMock(return_value=mock_state)
-            mock_cm.create = AsyncMock()
-
-            handler_instance = MockHandler.return_value
-            handler_instance.exchange_code_for_tokens = AsyncMock(
-                return_value=mock_creds
-            )
-
-            # Mock old credential cleanup
-            mock_cm.store.get_creds_by_provider = AsyncMock(return_value=[])
-
-            response = await client.post(
-                "/oauth/callback",
-                json={"code": "auth-code-abc", "state_token": "state-token-123"},
-            )
-
-        assert response.status_code == 200
-        data = response.json()
-        assert "id" in data
-        assert data["provider"] == "mcp"
-        assert data["type"] == "oauth2"
-        mock_cm.create.assert_called_once()
-
-    @pytest.mark.asyncio(loop_scope="session")
-    async def test_oauth_callback_invalid_state(self, client):
-        with patch("backend.api.features.mcp.routes.creds_manager") as mock_cm:
-            mock_cm.store.verify_state_token = AsyncMock(return_value=None)
-
-            response = await client.post(
-                "/oauth/callback",
-                json={"code": "auth-code", "state_token": "bad-state"},
-            )
-
-        assert response.status_code == 400
-        assert "Invalid or expired" in response.json()["detail"]
-
-    @pytest.mark.asyncio(loop_scope="session")
-    async def test_oauth_callback_token_exchange_fails(self, client):
-        with (
-            patch("backend.api.features.mcp.routes.creds_manager") as mock_cm,
-            patch("backend.api.features.mcp.routes.settings") as mock_settings,
-            patch("backend.api.features.mcp.routes.MCPOAuthHandler") as MockHandler,
-        ):
-            mock_settings.config.frontend_base_url = "http://localhost:3000"
-            mock_state = AsyncMock()
-            mock_state.state_metadata = {
-                "authorize_url": "https://auth.example.com/authorize",
-                "token_url": "https://auth.example.com/token",
-                "client_id": "cid",
-                "server_url": "https://mcp.example.com/mcp",
-            }
-            mock_state.scopes = []
-            mock_state.code_verifier = "v"
-            mock_cm.store.verify_state_token = AsyncMock(return_value=mock_state)
-
-            handler_instance = MockHandler.return_value
-            handler_instance.exchange_code_for_tokens = AsyncMock(
-                side_effect=RuntimeError("Token exchange failed")
-            )
-
-            response = await client.post(
-                "/oauth/callback",
-                json={"code": "bad-code", "state_token": "state"},
-            )
-
-        assert response.status_code == 400
-        assert "token exchange failed" in response.json()["detail"].lower()
--- a/autogpt_platform/backend/backend/api/features/otto/service.py
+++ b/autogpt_platform/backend/backend/api/features/otto/service.py
@@ -5,8 +5,8 @@ from typing import Optional
 import aiohttp
 from fastapi import HTTPException

-from backend.blocks import get_block
 from backend.data import graph as graph_db
+from backend.data.block import get_block
 from backend.util.settings import Settings

 from .models import ApiResponse, ChatRequest, GraphData
--- a/autogpt_platform/backend/backend/api/features/store/content_handlers.py
+++ b/autogpt_platform/backend/backend/api/features/store/content_handlers.py
@@ -152,7 +152,7 @@ class BlockHandler(ContentHandler):

    async def get_missing_items(self, batch_size: int) -> list[ContentItem]:
        """Fetch blocks without embeddings."""
-        from backend.blocks import get_blocks
+        from backend.data.block import get_blocks

        # Get all available blocks
        all_blocks = get_blocks()
@@ -249,7 +249,7 @@ class BlockHandler(ContentHandler):

    async def get_stats(self) -> dict[str, int]:
        """Get statistics about block embedding coverage."""
-        from backend.blocks import get_blocks
+        from backend.data.block import get_blocks

        all_blocks = get_blocks()

--- a/autogpt_platform/backend/backend/api/features/store/content_handlers_test.py
+++ b/autogpt_platform/backend/backend/api/features/store/content_handlers_test.py
@@ -93,7 +93,7 @@ async def test_block_handler_get_missing_items(mocker):
    mock_existing = []

    with patch(
-        "backend.blocks.get_blocks",
+        "backend.data.block.get_blocks",
        return_value=mock_blocks,
    ):
        with patch(
@@ -135,7 +135,7 @@ async def test_block_handler_get_stats(mocker):
    mock_embedded = [{"count": 2}]

    with patch(
-        "backend.blocks.get_blocks",
+        "backend.data.block.get_blocks",
        return_value=mock_blocks,
    ):
        with patch(
@@ -327,7 +327,7 @@ async def test_block_handler_handles_missing_attributes():
    mock_blocks = {"block-minimal": mock_block_class}

    with patch(
-        "backend.blocks.get_blocks",
+        "backend.data.block.get_blocks",
        return_value=mock_blocks,
    ):
        with patch(
@@ -360,7 +360,7 @@ async def test_block_handler_skips_failed_blocks():
    mock_blocks = {"good-block": good_block, "bad-block": bad_block}

    with patch(
-        "backend.blocks.get_blocks",
+        "backend.data.block.get_blocks",
        return_value=mock_blocks,
    ):
        with patch(
--- a/autogpt_platform/backend/backend/api/features/store/embeddings.py
+++ b/autogpt_platform/backend/backend/api/features/store/embeddings.py
@@ -662,7 +662,7 @@ async def cleanup_orphaned_embeddings() -> dict[str, Any]:
                )
                current_ids = {row["id"] for row in valid_agents}
            elif content_type == ContentType.BLOCK:
-                from backend.blocks import get_blocks
+                from backend.data.block import get_blocks

                current_ids = set(get_blocks().keys())
            elif content_type == ContentType.DOCUMENTATION:
--- a/autogpt_platform/backend/backend/api/features/store/hybrid_search.py
+++ b/autogpt_platform/backend/backend/api/features/store/hybrid_search.py
@@ -8,7 +8,6 @@ Includes BM25 reranking for improved lexical relevance.

 import logging
 import re
-import time
 from dataclasses import dataclass
 from typing import Any, Literal

@@ -363,11 +362,7 @@ async def unified_hybrid_search(
        LIMIT {limit_param} OFFSET {offset_param}
    """

-    try:
-        results = await query_raw_with_schema(sql_query, *params)
-    except Exception as e:
-        await _log_vector_error_diagnostics(e)
-        raise
+    results = await query_raw_with_schema(sql_query, *params)

    total = results[0]["total_count"] if results else 0
    # Apply BM25 reranking
@@ -691,11 +686,7 @@ async def hybrid_search(
        LIMIT {limit_param} OFFSET {offset_param}
    """

-    try:
-        results = await query_raw_with_schema(sql_query, *params)
-    except Exception as e:
-        await _log_vector_error_diagnostics(e)
-        raise
+    results = await query_raw_with_schema(sql_query, *params)

    total = results[0]["total_count"] if results else 0

@@ -727,87 +718,6 @@ async def hybrid_search_simple(
    return await hybrid_search(query=query, page=page, page_size=page_size)


-# ============================================================================
-# Diagnostics
-# ============================================================================
-
-# Rate limit: only log vector error diagnostics once per this interval
-_VECTOR_DIAG_INTERVAL_SECONDS = 60
-_last_vector_diag_time: float = 0
-
-
-async def _log_vector_error_diagnostics(error: Exception) -> None:
-    """Log diagnostic info when 'type vector does not exist' error occurs.
-
-    Note: Diagnostic queries use query_raw_with_schema which may run on a different
-    pooled connection than the one that failed. Session-level search_path can differ,
-    so these diagnostics show cluster-wide state, not necessarily the failed session.
-
-    Includes rate limiting to avoid log spam - only logs once per minute.
-    Caller should re-raise the error after calling this function.
-    """
-    global _last_vector_diag_time
-
-    # Check if this is the vector type error
-    error_str = str(error).lower()
-    if not (
-        "type" in error_str and "vector" in error_str and "does not exist" in error_str
-    ):
-        return
-
-    # Rate limit: only log once per interval
-    now = time.time()
-    if now - _last_vector_diag_time < _VECTOR_DIAG_INTERVAL_SECONDS:
-        return
-    _last_vector_diag_time = now
-
-    try:
-        diagnostics: dict[str, object] = {}
-
-        try:
-            search_path_result = await query_raw_with_schema("SHOW search_path")
-            diagnostics["search_path"] = search_path_result
-        except Exception as e:
-            diagnostics["search_path"] = f"Error: {e}"
-
-        try:
-            schema_result = await query_raw_with_schema("SELECT current_schema()")
-            diagnostics["current_schema"] = schema_result
-        except Exception as e:
-            diagnostics["current_schema"] = f"Error: {e}"
-
-        try:
-            user_result = await query_raw_with_schema(
-                "SELECT current_user, session_user, current_database()"
-            )
-            diagnostics["user_info"] = user_result
-        except Exception as e:
-            diagnostics["user_info"] = f"Error: {e}"
-
-        try:
-            # Check pgvector extension installation (cluster-wide, stable info)
-            ext_result = await query_raw_with_schema(
-                "SELECT extname, extversion, nspname as schema "
-                "FROM pg_extension e "
-                "JOIN pg_namespace n ON e.extnamespace = n.oid "
-                "WHERE extname = 'vector'"
-            )
-            diagnostics["pgvector_extension"] = ext_result
-        except Exception as e:
-            diagnostics["pgvector_extension"] = f"Error: {e}"
-
-        logger.error(
-            f"Vector type error diagnostics:\n"
-            f"  Error: {error}\n"
-            f"  search_path: {diagnostics.get('search_path')}\n"
-            f"  current_schema: {diagnostics.get('current_schema')}\n"
-            f"  user_info: {diagnostics.get('user_info')}\n"
-            f"  pgvector_extension: {diagnostics.get('pgvector_extension')}"
-        )
-    except Exception as diag_error:
-        logger.error(f"Failed to collect vector error diagnostics: {diag_error}")
-
-
 # Backward compatibility alias - HybridSearchWeights maps to StoreAgentSearchWeights
 # for existing code that expects the popularity parameter
 HybridSearchWeights = StoreAgentSearchWeights
--- a/autogpt_platform/backend/backend/api/features/store/image_gen.py
+++ b/autogpt_platform/backend/backend/api/features/store/image_gen.py
@@ -7,6 +7,15 @@ from replicate.client import Client as ReplicateClient
 from replicate.exceptions import ReplicateError
 from replicate.helpers import FileOutput

+from backend.blocks.ideogram import (
+    AspectRatio,
+    ColorPalettePreset,
+    IdeogramModelBlock,
+    IdeogramModelName,
+    MagicPromptOption,
+    StyleType,
+    UpscaleOption,
+)
 from backend.data.graph import GraphBaseMeta
 from backend.data.model import CredentialsMetaInput, ProviderName
 from backend.integrations.credentials_store import ideogram_credentials
@@ -41,16 +50,6 @@ async def generate_agent_image_v2(graph: GraphBaseMeta | AgentGraph) -> io.Bytes
    if not ideogram_credentials.api_key:
        raise ValueError("Missing Ideogram API key")

-    from backend.blocks.ideogram import (
-        AspectRatio,
-        ColorPalettePreset,
-        IdeogramModelBlock,
-        IdeogramModelName,
-        MagicPromptOption,
-        StyleType,
-        UpscaleOption,
-    )
-
    name = graph.name
    description = f"{name} ({graph.description})" if graph.description else name

--- a/autogpt_platform/backend/backend/api/features/v1.py
+++ b/autogpt_platform/backend/backend/api/features/v1.py
@@ -40,11 +40,10 @@ from backend.api.model import (
    UpdateTimezoneRequest,
    UploadFileResponse,
 )
-from backend.blocks import get_block, get_blocks
 from backend.data import execution as execution_db
 from backend.data import graph as graph_db
 from backend.data.auth import api_key as api_key_db
-from backend.data.block import BlockInput, CompletedBlockOutput
+from backend.data.block import BlockInput, CompletedBlockOutput, get_block, get_blocks
 from backend.data.credit import (
    AutoTopUpConfig,
    RefundRequest,
--- a/autogpt_platform/backend/backend/api/features/workspace/routes.py
+++ b/autogpt_platform/backend/backend/api/features/workspace/routes.py
@@ -11,7 +11,7 @@ import fastapi
 from autogpt_libs.auth.dependencies import get_user_id, requires_user
 from fastapi.responses import Response

-from backend.data.workspace import WorkspaceFile, get_workspace, get_workspace_file
+from backend.data.workspace import get_workspace, get_workspace_file
 from backend.util.workspace_storage import get_workspace_storage


@@ -44,11 +44,11 @@ router = fastapi.APIRouter(
 )


-def _create_streaming_response(content: bytes, file: WorkspaceFile) -> Response:
+def _create_streaming_response(content: bytes, file) -> Response:
    """Create a streaming response for file content."""
    return Response(
        content=content,
-        media_type=file.mime_type,
+        media_type=file.mimeType,
        headers={
            "Content-Disposition": _sanitize_filename_for_header(file.name),
            "Content-Length": str(len(content)),
@@ -56,7 +56,7 @@ def _create_streaming_response(content: bytes, file: WorkspaceFile) -> Response:
    )


-async def _create_file_download_response(file: WorkspaceFile) -> Response:
+async def _create_file_download_response(file) -> Response:
    """
    Create a download response for a workspace file.

@@ -66,33 +66,33 @@ async def _create_file_download_response(file: WorkspaceFile) -> Response:
    storage = await get_workspace_storage()

    # For local storage, stream the file directly
-    if file.storage_path.startswith("local://"):
-        content = await storage.retrieve(file.storage_path)
+    if file.storagePath.startswith("local://"):
+        content = await storage.retrieve(file.storagePath)
        return _create_streaming_response(content, file)

    # For GCS, try to redirect to signed URL, fall back to streaming
    try:
-        url = await storage.get_download_url(file.storage_path, expires_in=300)
+        url = await storage.get_download_url(file.storagePath, expires_in=300)
        # If we got back an API path (fallback), stream directly instead
        if url.startswith("/api/"):
-            content = await storage.retrieve(file.storage_path)
+            content = await storage.retrieve(file.storagePath)
            return _create_streaming_response(content, file)
        return fastapi.responses.RedirectResponse(url=url, status_code=302)
    except Exception as e:
        # Log the signed URL failure with context
        logger.error(
            f"Failed to get signed URL for file {file.id} "
-            f"(storagePath={file.storage_path}): {e}",
+            f"(storagePath={file.storagePath}): {e}",
            exc_info=True,
        )
        # Fall back to streaming directly from GCS
        try:
-            content = await storage.retrieve(file.storage_path)
+            content = await storage.retrieve(file.storagePath)
            return _create_streaming_response(content, file)
        except Exception as fallback_error:
            logger.error(
                f"Fallback streaming also failed for file {file.id} "
-                f"(storagePath={file.storage_path}): {fallback_error}",
+                f"(storagePath={file.storagePath}): {fallback_error}",
                exc_info=True,
            )
            raise
--- a/autogpt_platform/backend/backend/api/rest_api.py
+++ b/autogpt_platform/backend/backend/api/rest_api.py
@@ -26,7 +26,6 @@ import backend.api.features.executions.review.routes
 import backend.api.features.library.db
 import backend.api.features.library.model
 import backend.api.features.library.routes
-import backend.api.features.mcp.routes as mcp_routes
 import backend.api.features.oauth
 import backend.api.features.otto.routes
 import backend.api.features.postmark.postmark
@@ -41,11 +40,11 @@ import backend.data.user
 import backend.integrations.webhooks.utils
 import backend.util.service
 import backend.util.settings
-from backend.blocks.llm import DEFAULT_LLM_MODEL
-from backend.copilot.completion_consumer import (
+from backend.api.features.chat.completion_consumer import (
    start_completion_consumer,
    stop_completion_consumer,
 )
+from backend.blocks.llm import DEFAULT_LLM_MODEL
 from backend.data.model import Credentials
 from backend.integrations.providers import ProviderName
 from backend.monitoring.instrumentation import instrument_fastapi
@@ -344,11 +343,6 @@ app.include_router(
    tags=["workspace"],
    prefix="/api/workspace",
 )
-app.include_router(
-    mcp_routes.router,
-    tags=["v2", "mcp"],
-    prefix="/api/mcp",
-)
 app.include_router(
    backend.api.features.oauth.router,
    tags=["oauth"],
--- a/autogpt_platform/backend/backend/app.py
+++ b/autogpt_platform/backend/backend/app.py
@@ -38,9 +38,7 @@ def main(**kwargs):

    from backend.api.rest_api import AgentServer
    from backend.api.ws_api import WebsocketServer
-    from backend.copilot.executor.manager import CoPilotExecutor
-    from backend.data.db_manager import DatabaseManager
-    from backend.executor import ExecutionManager, Scheduler
+    from backend.executor import DatabaseManager, ExecutionManager, Scheduler
    from backend.notifications import NotificationManager

    run_processes(
@@ -50,7 +48,6 @@ def main(**kwargs):
        WebsocketServer(),
        AgentServer(),
        ExecutionManager(),
-        CoPilotExecutor(),
        **kwargs,
    )

--- a/autogpt_platform/backend/backend/blocks/init.py
+++ b/autogpt_platform/backend/backend/blocks/init.py
@@ -3,19 +3,22 @@ import logging
 import os
 import re
 from pathlib import Path
-from typing import Sequence, Type, TypeVar
+from typing import TYPE_CHECKING, TypeVar

-from backend.blocks._base import AnyBlockSchema, BlockType
 from backend.util.cache import cached

 logger = logging.getLogger(__name__)

+
+if TYPE_CHECKING:
+    from backend.data.block import Block
+
 T = TypeVar("T")


@cached(ttl_seconds=3600)
-def load_all_blocks() -> dict[str, type["AnyBlockSchema"]]:
-    from backend.blocks._base import Block
+def load_all_blocks() -> dict[str, type["Block"]]:
+    from backend.data.block import Block
    from backend.util.settings import Config

    # Check if example blocks should be loaded from settings
@@ -47,8 +50,8 @@ def load_all_blocks() -> dict[str, type["AnyBlockSchema"]]:
        importlib.import_module(f".{module}", package=__name__)

    # Load all Block instances from the available modules
-    available_blocks: dict[str, type["AnyBlockSchema"]] = {}
-    for block_cls in _all_subclasses(Block):
+    available_blocks: dict[str, type["Block"]] = {}
+    for block_cls in all_subclasses(Block):
        class_name = block_cls.__name__

        if class_name.endswith("Base"):
@@ -61,7 +64,7 @@ def load_all_blocks() -> dict[str, type["AnyBlockSchema"]]:
                "please name the class with 'Base' at the end"
            )

-        block = block_cls()  # pyright: ignore[reportAbstractUsage]
+        block = block_cls.create()

        if not isinstance(block.id, str) or len(block.id) != 36:
            raise ValueError(
@@ -102,7 +105,7 @@ def load_all_blocks() -> dict[str, type["AnyBlockSchema"]]:
        available_blocks[block.id] = block_cls

    # Filter out blocks with incomplete auth configs, e.g. missing OAuth server secrets
-    from ._utils import is_block_auth_configured
+    from backend.data.block import is_block_auth_configured

    filtered_blocks = {}
    for block_id, block_cls in available_blocks.items():
@@ -112,48 +115,11 @@ def load_all_blocks() -> dict[str, type["AnyBlockSchema"]]:
    return filtered_blocks


-def _all_subclasses(cls: type[T]) -> list[type[T]]:
+__all__ = ["load_all_blocks"]
+
+
+def all_subclasses(cls: type[T]) -> list[type[T]]:
    subclasses = cls.__subclasses__()
    for subclass in subclasses:
-        subclasses += _all_subclasses(subclass)
+        subclasses += all_subclasses(subclass)
    return subclasses
-
-
-# ============== Block access helper functions ============== #
-
-
-def get_blocks() -> dict[str, Type["AnyBlockSchema"]]:
-    return load_all_blocks()
-
-
-# Note on the return type annotation: https://github.com/microsoft/pyright/issues/10281
-def get_block(block_id: str) -> "AnyBlockSchema | None":
-    cls = get_blocks().get(block_id)
-    return cls() if cls else None
-
-
-@cached(ttl_seconds=3600)
-def get_webhook_block_ids() -> Sequence[str]:
-    return [
-        id
-        for id, B in get_blocks().items()
-        if B().block_type in (BlockType.WEBHOOK, BlockType.WEBHOOK_MANUAL)
-    ]
-
-
-@cached(ttl_seconds=3600)
-def get_io_block_ids() -> Sequence[str]:
-    return [
-        id
-        for id, B in get_blocks().items()
-        if B().block_type in (BlockType.INPUT, BlockType.OUTPUT)
-    ]
-
-
-@cached(ttl_seconds=3600)
-def get_human_in_the_loop_block_ids() -> Sequence[str]:
-    return [
-        id
-        for id, B in get_blocks().items()
-        if B().block_type == BlockType.HUMAN_IN_THE_LOOP
-    ]
--- a/autogpt_platform/backend/backend/blocks/_base.py
+++ b/autogpt_platform/backend/backend/blocks/_base.py
@@ -1,740 +0,0 @@
-import inspect
-import logging
-from abc import ABC, abstractmethod
-from enum import Enum
-from typing import (
-    TYPE_CHECKING,
-    Any,
-    Callable,
-    ClassVar,
-    Generic,
-    Optional,
-    Type,
-    TypeAlias,
-    TypeVar,
-    cast,
-    get_origin,
-)
-
-import jsonref
-import jsonschema
-from pydantic import BaseModel
-
-from backend.data.block import BlockInput, BlockOutput, BlockOutputEntry
-from backend.data.model import (
-    Credentials,
-    CredentialsFieldInfo,
-    CredentialsMetaInput,
-    SchemaField,
-    is_credentials_field_name,
-)
-from backend.integrations.providers import ProviderName
-from backend.util import json
-from backend.util.exceptions import (
-    BlockError,
-    BlockExecutionError,
-    BlockInputError,
-    BlockOutputError,
-    BlockUnknownError,
-)
-from backend.util.settings import Config
-
-logger = logging.getLogger(__name__)
-
-if TYPE_CHECKING:
-    from backend.data.execution import ExecutionContext
-    from backend.data.model import ContributorDetails, NodeExecutionStats
-
-    from ..data.graph import Link
-
-app_config = Config()
-
-
-BlockTestOutput = BlockOutputEntry | tuple[str, Callable[[Any], bool]]
-
-
-class BlockType(Enum):
-    STANDARD = "Standard"
-    INPUT = "Input"
-    OUTPUT = "Output"
-    NOTE = "Note"
-    WEBHOOK = "Webhook"
-    WEBHOOK_MANUAL = "Webhook (manual)"
-    AGENT = "Agent"
-    AI = "AI"
-    AYRSHARE = "Ayrshare"
-    HUMAN_IN_THE_LOOP = "Human In The Loop"
-    MCP_TOOL = "MCP Tool"
-
-
-class BlockCategory(Enum):
-    AI = "Block that leverages AI to perform a task."
-    SOCIAL = "Block that interacts with social media platforms."
-    TEXT = "Block that processes text data."
-    SEARCH = "Block that searches or extracts information from the internet."
-    BASIC = "Block that performs basic operations."
-    INPUT = "Block that interacts with input of the graph."
-    OUTPUT = "Block that interacts with output of the graph."
-    LOGIC = "Programming logic to control the flow of your agent"
-    COMMUNICATION = "Block that interacts with communication platforms."
-    DEVELOPER_TOOLS = "Developer tools such as GitHub blocks."
-    DATA = "Block that interacts with structured data."
-    HARDWARE = "Block that interacts with hardware."
-    AGENT = "Block that interacts with other agents."
-    CRM = "Block that interacts with CRM services."
-    SAFETY = (
-        "Block that provides AI safety mechanisms such as detecting harmful content"
-    )
-    PRODUCTIVITY = "Block that helps with productivity"
-    ISSUE_TRACKING = "Block that helps with issue tracking"
-    MULTIMEDIA = "Block that interacts with multimedia content"
-    MARKETING = "Block that helps with marketing"
-
-    def dict(self) -> dict[str, str]:
-        return {"category": self.name, "description": self.value}
-
-
-class BlockCostType(str, Enum):
-    RUN = "run"  # cost X credits per run
-    BYTE = "byte"  # cost X credits per byte
-    SECOND = "second"  # cost X credits per second
-
-
-class BlockCost(BaseModel):
-    cost_amount: int
-    cost_filter: BlockInput
-    cost_type: BlockCostType
-
-    def __init__(
-        self,
-        cost_amount: int,
-        cost_type: BlockCostType = BlockCostType.RUN,
-        cost_filter: Optional[BlockInput] = None,
-        **data: Any,
-    ) -> None:
-        super().__init__(
-            cost_amount=cost_amount,
-            cost_filter=cost_filter or {},
-            cost_type=cost_type,
-            **data,
-        )
-
-
-class BlockInfo(BaseModel):
-    id: str
-    name: str
-    inputSchema: dict[str, Any]
-    outputSchema: dict[str, Any]
-    costs: list[BlockCost]
-    description: str
-    categories: list[dict[str, str]]
-    contributors: list[dict[str, Any]]
-    staticOutput: bool
-    uiType: str
-
-
-class BlockSchema(BaseModel):
-    cached_jsonschema: ClassVar[dict[str, Any]]
-
-    @classmethod
-    def jsonschema(cls) -> dict[str, Any]:
-        if cls.cached_jsonschema:
-            return cls.cached_jsonschema
-
-        model = jsonref.replace_refs(cls.model_json_schema(), merge_props=True)
-
-        def ref_to_dict(obj):
-            if isinstance(obj, dict):
-                # OpenAPI <3.1 does not support sibling fields that has a $ref key
-                # So sometimes, the schema has an "allOf"/"anyOf"/"oneOf" with 1 item.
-                keys = {"allOf", "anyOf", "oneOf"}
-                one_key = next((k for k in keys if k in obj and len(obj[k]) == 1), None)
-                if one_key:
-                    obj.update(obj[one_key][0])
-
-                return {
-                    key: ref_to_dict(value)
-                    for key, value in obj.items()
-                    if not key.startswith("$") and key != one_key
-                }
-            elif isinstance(obj, list):
-                return [ref_to_dict(item) for item in obj]
-
-            return obj
-
-        cls.cached_jsonschema = cast(dict[str, Any], ref_to_dict(model))
-
-        return cls.cached_jsonschema
-
-    @classmethod
-    def validate_data(cls, data: BlockInput) -> str | None:
-        return json.validate_with_jsonschema(
-            schema=cls.jsonschema(),
-            data={k: v for k, v in data.items() if v is not None},
-        )
-
-    @classmethod
-    def get_mismatch_error(cls, data: BlockInput) -> str | None:
-        return cls.validate_data(data)
-
-    @classmethod
-    def get_field_schema(cls, field_name: str) -> dict[str, Any]:
-        model_schema = cls.jsonschema().get("properties", {})
-        if not model_schema:
-            raise ValueError(f"Invalid model schema {cls}")
-
-        property_schema = model_schema.get(field_name)
-        if not property_schema:
-            raise ValueError(f"Invalid property name {field_name}")
-
-        return property_schema
-
-    @classmethod
-    def validate_field(cls, field_name: str, data: BlockInput) -> str | None:
-        """
-        Validate the data against a specific property (one of the input/output name).
-        Returns the validation error message if the data does not match the schema.
-        """
-        try:
-            property_schema = cls.get_field_schema(field_name)
-            jsonschema.validate(json.to_dict(data), property_schema)
-            return None
-        except jsonschema.ValidationError as e:
-            return str(e)
-
-    @classmethod
-    def get_fields(cls) -> set[str]:
-        return set(cls.model_fields.keys())
-
-    @classmethod
-    def get_required_fields(cls) -> set[str]:
-        return {
-            field
-            for field, field_info in cls.model_fields.items()
-            if field_info.is_required()
-        }
-
-    @classmethod
-    def __pydantic_init_subclass__(cls, **kwargs):
-        """Validates the schema definition. Rules:
-        - Fields with annotation `CredentialsMetaInput` MUST be
-          named `credentials` or `*_credentials`
-        - Fields named `credentials` or `*_credentials` MUST be
-          of type `CredentialsMetaInput`
-        """
-        super().__pydantic_init_subclass__(**kwargs)
-
-        # Reset cached JSON schema to prevent inheriting it from parent class
-        cls.cached_jsonschema = {}
-
-        credentials_fields = cls.get_credentials_fields()
-
-        for field_name in cls.get_fields():
-            if is_credentials_field_name(field_name):
-                if field_name not in credentials_fields:
-                    raise TypeError(
-                        f"Credentials field '{field_name}' on {cls.__qualname__} "
-                        f"is not of type {CredentialsMetaInput.__name__}"
-                    )
-
-                CredentialsMetaInput.validate_credentials_field_schema(
-                    cls.get_field_schema(field_name), field_name
-                )
-
-            elif field_name in credentials_fields:
-                raise KeyError(
-                    f"Credentials field '{field_name}' on {cls.__qualname__} "
-                    "has invalid name: must be 'credentials' or *_credentials"
-                )
-
-    @classmethod
-    def get_credentials_fields(cls) -> dict[str, type[CredentialsMetaInput]]:
-        return {
-            field_name: info.annotation
-            for field_name, info in cls.model_fields.items()
-            if (
-                inspect.isclass(info.annotation)
-                and issubclass(
-                    get_origin(info.annotation) or info.annotation,
-                    CredentialsMetaInput,
-                )
-            )
-        }
-
-    @classmethod
-    def get_auto_credentials_fields(cls) -> dict[str, dict[str, Any]]:
-        """
-        Get fields that have auto_credentials metadata (e.g., GoogleDriveFileInput).
-
-        Returns a dict mapping kwarg_name -> {field_name, auto_credentials_config}
-
-        Raises:
-            ValueError: If multiple fields have the same kwarg_name, as this would
-                cause silent overwriting and only the last field would be processed.
-        """
-        result: dict[str, dict[str, Any]] = {}
-        schema = cls.jsonschema()
-        properties = schema.get("properties", {})
-
-        for field_name, field_schema in properties.items():
-            auto_creds = field_schema.get("auto_credentials")
-            if auto_creds:
-                kwarg_name = auto_creds.get("kwarg_name", "credentials")
-                if kwarg_name in result:
-                    raise ValueError(
-                        f"Duplicate auto_credentials kwarg_name '{kwarg_name}' "
-                        f"in fields '{result[kwarg_name]['field_name']}' and "
-                        f"'{field_name}' on {cls.__qualname__}"
-                    )
-                result[kwarg_name] = {
-                    "field_name": field_name,
-                    "config": auto_creds,
-                }
-        return result
-
-    @classmethod
-    def get_credentials_fields_info(cls) -> dict[str, CredentialsFieldInfo]:
-        result = {}
-
-        # Regular credentials fields
-        for field_name in cls.get_credentials_fields().keys():
-            result[field_name] = CredentialsFieldInfo.model_validate(
-                cls.get_field_schema(field_name), by_alias=True
-            )
-
-        # Auto-generated credentials fields (from GoogleDriveFileInput etc.)
-        for kwarg_name, info in cls.get_auto_credentials_fields().items():
-            config = info["config"]
-            # Build a schema-like dict that CredentialsFieldInfo can parse
-            auto_schema = {
-                "credentials_provider": [config.get("provider", "google")],
-                "credentials_types": [config.get("type", "oauth2")],
-                "credentials_scopes": config.get("scopes"),
-            }
-            result[kwarg_name] = CredentialsFieldInfo.model_validate(
-                auto_schema, by_alias=True
-            )
-
-        return result
-
-    @classmethod
-    def get_input_defaults(cls, data: BlockInput) -> BlockInput:
-        return data  # Return as is, by default.
-
-    @classmethod
-    def get_missing_links(cls, data: BlockInput, links: list["Link"]) -> set[str]:
-        input_fields_from_nodes = {link.sink_name for link in links}
-        return input_fields_from_nodes - set(data)
-
-    @classmethod
-    def get_missing_input(cls, data: BlockInput) -> set[str]:
-        return cls.get_required_fields() - set(data)
-
-
-class BlockSchemaInput(BlockSchema):
-    """
-    Base schema class for block inputs.
-    All block input schemas should extend this class for consistency.
-    """
-
-    pass
-
-
-class BlockSchemaOutput(BlockSchema):
-    """
-    Base schema class for block outputs that includes a standard error field.
-    All block output schemas should extend this class to ensure consistent error handling.
-    """
-
-    error: str = SchemaField(
-        description="Error message if the operation failed", default=""
-    )
-
-
-BlockSchemaInputType = TypeVar("BlockSchemaInputType", bound=BlockSchemaInput)
-BlockSchemaOutputType = TypeVar("BlockSchemaOutputType", bound=BlockSchemaOutput)
-
-
-class EmptyInputSchema(BlockSchemaInput):
-    pass
-
-
-class EmptyOutputSchema(BlockSchemaOutput):
-    pass
-
-
-# For backward compatibility - will be deprecated
-EmptySchema = EmptyOutputSchema
-
-
-# --8<-- [start:BlockWebhookConfig]
-class BlockManualWebhookConfig(BaseModel):
-    """
-    Configuration model for webhook-triggered blocks on which
-    the user has to manually set up the webhook at the provider.
-    """
-
-    provider: ProviderName
-    """The service provider that the webhook connects to"""
-
-    webhook_type: str
-    """
-    Identifier for the webhook type. E.g. GitHub has repo and organization level hooks.
-
-    Only for use in the corresponding `WebhooksManager`.
-    """
-
-    event_filter_input: str = ""
-    """
-    Name of the block's event filter input.
-    Leave empty if the corresponding webhook doesn't have distinct event/payload types.
-    """
-
-    event_format: str = "{event}"
-    """
-    Template string for the event(s) that a block instance subscribes to.
-    Applied individually to each event selected in the event filter input.
-
-    Example: `"pull_request.{event}"` -> `"pull_request.opened"`
-    """
-
-
-class BlockWebhookConfig(BlockManualWebhookConfig):
-    """
-    Configuration model for webhook-triggered blocks for which
-    the webhook can be automatically set up through the provider's API.
-    """
-
-    resource_format: str
-    """
-    Template string for the resource that a block instance subscribes to.
-    Fields will be filled from the block's inputs (except `payload`).
-
-    Example: `f"{repo}/pull_requests"` (note: not how it's actually implemented)
-
-    Only for use in the corresponding `WebhooksManager`.
-    """
-    # --8<-- [end:BlockWebhookConfig]
-
-
-class Block(ABC, Generic[BlockSchemaInputType, BlockSchemaOutputType]):
-    def __init__(
-        self,
-        id: str = "",
-        description: str = "",
-        contributors: list["ContributorDetails"] = [],
-        categories: set[BlockCategory] | None = None,
-        input_schema: Type[BlockSchemaInputType] = EmptyInputSchema,
-        output_schema: Type[BlockSchemaOutputType] = EmptyOutputSchema,
-        test_input: BlockInput | list[BlockInput] | None = None,
-        test_output: BlockTestOutput | list[BlockTestOutput] | None = None,
-        test_mock: dict[str, Any] | None = None,
-        test_credentials: Optional[Credentials | dict[str, Credentials]] = None,
-        disabled: bool = False,
-        static_output: bool = False,
-        block_type: BlockType = BlockType.STANDARD,
-        webhook_config: Optional[BlockWebhookConfig | BlockManualWebhookConfig] = None,
-        is_sensitive_action: bool = False,
-    ):
-        """
-        Initialize the block with the given schema.
-
-        Args:
-            id: The unique identifier for the block, this value will be persisted in the
-                DB. So it should be a unique and constant across the application run.
-                Use the UUID format for the ID.
-            description: The description of the block, explaining what the block does.
-            contributors: The list of contributors who contributed to the block.
-            input_schema: The schema, defined as a Pydantic model, for the input data.
-            output_schema: The schema, defined as a Pydantic model, for the output data.
-            test_input: The list or single sample input data for the block, for testing.
-            test_output: The list or single expected output if the test_input is run.
-            test_mock: function names on the block implementation to mock on test run.
-            disabled: If the block is disabled, it will not be available for execution.
-            static_output: Whether the output links of the block are static by default.
-        """
-        from backend.data.model import NodeExecutionStats
-
-        self.id = id
-        self.input_schema = input_schema
-        self.output_schema = output_schema
-        self.test_input = test_input
-        self.test_output = test_output
-        self.test_mock = test_mock
-        self.test_credentials = test_credentials
-        self.description = description
-        self.categories = categories or set()
-        self.contributors = contributors or set()
-        self.disabled = disabled
-        self.static_output = static_output
-        self.block_type = block_type
-        self.webhook_config = webhook_config
-        self.is_sensitive_action = is_sensitive_action
-        self.execution_stats: "NodeExecutionStats" = NodeExecutionStats()
-
-        if self.webhook_config:
-            if isinstance(self.webhook_config, BlockWebhookConfig):
-                # Enforce presence of credentials field on auto-setup webhook blocks
-                if not (cred_fields := self.input_schema.get_credentials_fields()):
-                    raise TypeError(
-                        "credentials field is required on auto-setup webhook blocks"
-                    )
-                # Disallow multiple credentials inputs on webhook blocks
-                elif len(cred_fields) > 1:
-                    raise ValueError(
-                        "Multiple credentials inputs not supported on webhook blocks"
-                    )
-
-                self.block_type = BlockType.WEBHOOK
-            else:
-                self.block_type = BlockType.WEBHOOK_MANUAL
-
-            # Enforce shape of webhook event filter, if present
-            if self.webhook_config.event_filter_input:
-                event_filter_field = self.input_schema.model_fields[
-                    self.webhook_config.event_filter_input
-                ]
-                if not (
-                    isinstance(event_filter_field.annotation, type)
-                    and issubclass(event_filter_field.annotation, BaseModel)
-                    and all(
-                        field.annotation is bool
-                        for field in event_filter_field.annotation.model_fields.values()
-                    )
-                ):
-                    raise NotImplementedError(
-                        f"{self.name} has an invalid webhook event selector: "
-                        "field must be a BaseModel and all its fields must be boolean"
-                    )
-
-            # Enforce presence of 'payload' input
-            if "payload" not in self.input_schema.model_fields:
-                raise TypeError(
-                    f"{self.name} is webhook-triggered but has no 'payload' input"
-                )
-
-            # Disable webhook-triggered block if webhook functionality not available
-            if not app_config.platform_base_url:
-                self.disabled = True
-
-    @abstractmethod
-    async def run(self, input_data: BlockSchemaInputType, **kwargs) -> BlockOutput:
-        """
-        Run the block with the given input data.
-        Args:
-            input_data: The input data with the structure of input_schema.
-
-        Kwargs: Currently 14/02/2025 these include
-            graph_id: The ID of the graph.
-            node_id: The ID of the node.
-            graph_exec_id: The ID of the graph execution.
-            node_exec_id: The ID of the node execution.
-            user_id: The ID of the user.
-
-        Returns:
-            A Generator that yields (output_name, output_data).
-            output_name: One of the output name defined in Block's output_schema.
-            output_data: The data for the output_name, matching the defined schema.
-        """
-        # --- satisfy the type checker, never executed -------------
-        if False:  # noqa: SIM115
-            yield "name", "value"  # pyright: ignore[reportMissingYield]
-        raise NotImplementedError(f"{self.name} does not implement the run method.")
-
-    async def run_once(
-        self, input_data: BlockSchemaInputType, output: str, **kwargs
-    ) -> Any:
-        async for item in self.run(input_data, **kwargs):
-            name, data = item
-            if name == output:
-                return data
-        raise ValueError(f"{self.name} did not produce any output for {output}")
-
-    def merge_stats(self, stats: "NodeExecutionStats") -> "NodeExecutionStats":
-        self.execution_stats += stats
-        return self.execution_stats
-
-    @property
-    def name(self):
-        return self.__class__.__name__
-
-    def to_dict(self):
-        return {
-            "id": self.id,
-            "name": self.name,
-            "inputSchema": self.input_schema.jsonschema(),
-            "outputSchema": self.output_schema.jsonschema(),
-            "description": self.description,
-            "categories": [category.dict() for category in self.categories],
-            "contributors": [
-                contributor.model_dump() for contributor in self.contributors
-            ],
-            "staticOutput": self.static_output,
-            "uiType": self.block_type.value,
-        }
-
-    def get_info(self) -> BlockInfo:
-        from backend.data.credit import get_block_cost
-
-        return BlockInfo(
-            id=self.id,
-            name=self.name,
-            inputSchema=self.input_schema.jsonschema(),
-            outputSchema=self.output_schema.jsonschema(),
-            costs=get_block_cost(self),
-            description=self.description,
-            categories=[category.dict() for category in self.categories],
-            contributors=[
-                contributor.model_dump() for contributor in self.contributors
-            ],
-            staticOutput=self.static_output,
-            uiType=self.block_type.value,
-        )
-
-    async def execute(self, input_data: BlockInput, **kwargs) -> BlockOutput:
-        try:
-            async for output_name, output_data in self._execute(input_data, **kwargs):
-                yield output_name, output_data
-        except Exception as ex:
-            if isinstance(ex, BlockError):
-                raise ex
-            else:
-                raise (
-                    BlockExecutionError
-                    if isinstance(ex, ValueError)
-                    else BlockUnknownError
-                )(
-                    message=str(ex),
-                    block_name=self.name,
-                    block_id=self.id,
-                ) from ex
-
-    async def is_block_exec_need_review(
-        self,
-        input_data: BlockInput,
-        *,
-        user_id: str,
-        node_id: str,
-        node_exec_id: str,
-        graph_exec_id: str,
-        graph_id: str,
-        graph_version: int,
-        execution_context: "ExecutionContext",
-        **kwargs,
-    ) -> tuple[bool, BlockInput]:
-        """
-        Check if this block execution needs human review and handle the review process.
-
-        Returns:
-            Tuple of (should_pause, input_data_to_use)
-            - should_pause: True if execution should be paused for review
-            - input_data_to_use: The input data to use (may be modified by reviewer)
-        """
-        if not (
-            self.is_sensitive_action and execution_context.sensitive_action_safe_mode
-        ):
-            return False, input_data
-
-        from backend.blocks.helpers.review import HITLReviewHelper
-
-        # Handle the review request and get decision
-        decision = await HITLReviewHelper.handle_review_decision(
-            input_data=input_data,
-            user_id=user_id,
-            node_id=node_id,
-            node_exec_id=node_exec_id,
-            graph_exec_id=graph_exec_id,
-            graph_id=graph_id,
-            graph_version=graph_version,
-            block_name=self.name,
-            editable=True,
-        )
-
-        if decision is None:
-            # We're awaiting review - pause execution
-            return True, input_data
-
-        if not decision.should_proceed:
-            # Review was rejected, raise an error to stop execution
-            raise BlockExecutionError(
-                message=f"Block execution rejected by reviewer: {decision.message}",
-                block_name=self.name,
-                block_id=self.id,
-            )
-
-        # Review was approved - use the potentially modified data
-        # ReviewResult.data must be a dict for block inputs
-        reviewed_data = decision.review_result.data
-        if not isinstance(reviewed_data, dict):
-            raise BlockExecutionError(
-                message=f"Review data must be a dict for block input, got {type(reviewed_data).__name__}",
-                block_name=self.name,
-                block_id=self.id,
-            )
-        return False, reviewed_data
-
-    async def _execute(self, input_data: BlockInput, **kwargs) -> BlockOutput:
-        # Check for review requirement only if running within a graph execution context
-        # Direct block execution (e.g., from chat) skips the review process
-        has_graph_context = all(
-            key in kwargs
-            for key in (
-                "node_exec_id",
-                "graph_exec_id",
-                "graph_id",
-                "execution_context",
-            )
-        )
-        if has_graph_context:
-            should_pause, input_data = await self.is_block_exec_need_review(
-                input_data, **kwargs
-            )
-            if should_pause:
-                return
-
-        # Validate the input data (original or reviewer-modified) once
-        if error := self.input_schema.validate_data(input_data):
-            raise BlockInputError(
-                message=f"Unable to execute block with invalid input data: {error}",
-                block_name=self.name,
-                block_id=self.id,
-            )
-
-        # Use the validated input data
-        async for output_name, output_data in self.run(
-            self.input_schema(**{k: v for k, v in input_data.items() if v is not None}),
-            **kwargs,
-        ):
-            if output_name == "error":
-                raise BlockExecutionError(
-                    message=output_data, block_name=self.name, block_id=self.id
-                )
-            if self.block_type == BlockType.STANDARD and (
-                error := self.output_schema.validate_field(output_name, output_data)
-            ):
-                raise BlockOutputError(
-                    message=f"Block produced an invalid output data: {error}",
-                    block_name=self.name,
-                    block_id=self.id,
-                )
-            yield output_name, output_data
-
-    def is_triggered_by_event_type(
-        self, trigger_config: dict[str, Any], event_type: str
-    ) -> bool:
-        if not self.webhook_config:
-            raise TypeError("This method can't be used on non-trigger blocks")
-        if not self.webhook_config.event_filter_input:
-            return True
-        event_filter = trigger_config.get(self.webhook_config.event_filter_input)
-        if not event_filter:
-            raise ValueError("Event filter is not configured on trigger")
-        return event_type in [
-            self.webhook_config.event_format.format(event=k)
-            for k in event_filter
-            if event_filter[k] is True
-        ]
-
-
-# Type alias for any block with standard input/output schemas
-AnyBlockSchema: TypeAlias = Block[BlockSchemaInput, BlockSchemaOutput]
--- a/autogpt_platform/backend/backend/blocks/_utils.py
+++ b/autogpt_platform/backend/backend/blocks/_utils.py
@@ -1,122 +0,0 @@
-import logging
-import os
-
-from backend.integrations.providers import ProviderName
-
-from ._base import AnyBlockSchema
-
-logger = logging.getLogger(__name__)
-
-
-def is_block_auth_configured(
-    block_cls: type[AnyBlockSchema],
-) -> bool:
-    """
-    Check if a block has a valid authentication method configured at runtime.
-
-    For example if a block is an OAuth-only block and there env vars are not set,
-    do not show it in the UI.
-
-    """
-    from backend.sdk.registry import AutoRegistry
-
-    # Create an instance to access input_schema
-    try:
-        block = block_cls()
-    except Exception as e:
-        # If we can't create a block instance, assume it's not OAuth-only
-        logger.error(f"Error creating block instance for {block_cls.__name__}: {e}")
-        return True
-    logger.debug(
-        f"Checking if block {block_cls.__name__} has a valid provider configured"
-    )
-
-    # Get all credential inputs from input schema
-    credential_inputs = block.input_schema.get_credentials_fields_info()
-    required_inputs = block.input_schema.get_required_fields()
-    if not credential_inputs:
-        logger.debug(
-            f"Block {block_cls.__name__} has no credential inputs - Treating as valid"
-        )
-        return True
-
-    # Check credential inputs
-    if len(required_inputs.intersection(credential_inputs.keys())) == 0:
-        logger.debug(
-            f"Block {block_cls.__name__} has only optional credential inputs"
-            " - will work without credentials configured"
-        )
-
-    # Check if the credential inputs for this block are correctly configured
-    for field_name, field_info in credential_inputs.items():
-        provider_names = field_info.provider
-        if not provider_names:
-            logger.warning(
-                f"Block {block_cls.__name__} "
-                f"has credential input '{field_name}' with no provider options"
-                " - Disabling"
-            )
-            return False
-
-        # If a field has multiple possible providers, each one needs to be usable to
-        # prevent breaking the UX
-        for _provider_name in provider_names:
-            provider_name = _provider_name.value
-            if provider_name in ProviderName.__members__.values():
-                logger.debug(
-                    f"Block {block_cls.__name__} credential input '{field_name}' "
-                    f"provider '{provider_name}' is part of the legacy provider system"
-                    " - Treating as valid"
-                )
-                break
-
-            provider = AutoRegistry.get_provider(provider_name)
-            if not provider:
-                logger.warning(
-                    f"Block {block_cls.__name__} credential input '{field_name}' "
-                    f"refers to unknown provider '{provider_name}' - Disabling"
-                )
-                return False
-
-            # Check the provider's supported auth types
-            if field_info.supported_types != provider.supported_auth_types:
-                logger.warning(
-                    f"Block {block_cls.__name__} credential input '{field_name}' "
-                    f"has mismatched supported auth types (field <> Provider): "
-                    f"{field_info.supported_types} != {provider.supported_auth_types}"
-                )
-
-            if not (supported_auth_types := provider.supported_auth_types):
-                # No auth methods are been configured for this provider
-                logger.warning(
-                    f"Block {block_cls.__name__} credential input '{field_name}' "
-                    f"provider '{provider_name}' "
-                    "has no authentication methods configured - Disabling"
-                )
-                return False
-
-            # Check if provider supports OAuth
-            if "oauth2" in supported_auth_types:
-                # Check if OAuth environment variables are set
-                if (oauth_config := provider.oauth_config) and bool(
-                    os.getenv(oauth_config.client_id_env_var)
-                    and os.getenv(oauth_config.client_secret_env_var)
-                ):
-                    logger.debug(
-                        f"Block {block_cls.__name__} credential input '{field_name}' "
-                        f"provider '{provider_name}' is configured for OAuth"
-                    )
-                else:
-                    logger.error(
-                        f"Block {block_cls.__name__} credential input '{field_name}' "
-                        f"provider '{provider_name}' "
-                        "is missing OAuth client ID or secret - Disabling"
-                    )
-                    return False
-
-        logger.debug(
-            f"Block {block_cls.__name__} credential input '{field_name}' is valid; "
-            f"supported credential types: {', '.join(field_info.supported_types)}"
-        )
-
-    return True
--- a/autogpt_platform/backend/backend/blocks/agent.py
+++ b/autogpt_platform/backend/backend/blocks/agent.py
@@ -1,7 +1,7 @@
 import logging
-from typing import TYPE_CHECKING, Any, Optional
+from typing import Any, Optional

-from backend.blocks._base import (
+from backend.data.block import (
    Block,
    BlockCategory,
    BlockInput,
@@ -9,15 +9,13 @@ from backend.blocks._base import (
    BlockSchema,
    BlockSchemaInput,
    BlockType,
+    get_block,
 )
 from backend.data.execution import ExecutionContext, ExecutionStatus, NodesInputMasks
 from backend.data.model import NodeExecutionStats, SchemaField
 from backend.util.json import validate_with_jsonschema
 from backend.util.retry import func_retry

-if TYPE_CHECKING:
-    from backend.executor.utils import LogMetadata
-
 _logger = logging.getLogger(__name__)


@@ -126,10 +124,9 @@ class AgentExecutorBlock(Block):
        graph_version: int,
        graph_exec_id: str,
        user_id: str,
-        logger: "LogMetadata",
+        logger,
    ) -> BlockOutput:

-        from backend.blocks import get_block
        from backend.data.execution import ExecutionEventType
        from backend.executor import utils as execution_utils

@@ -201,7 +198,7 @@ class AgentExecutorBlock(Block):
        self,
        graph_exec_id: str,
        user_id: str,
-        logger: "LogMetadata",
+        logger,
    ) -> None:
        from backend.executor import utils as execution_utils

--- a/autogpt_platform/backend/backend/blocks/ai_condition.py
+++ b/autogpt_platform/backend/backend/blocks/ai_condition.py
@@ -1,11 +1,5 @@
 from typing import Any

-from backend.blocks._base import (
-    BlockCategory,
-    BlockOutput,
-    BlockSchemaInput,
-    BlockSchemaOutput,
-)
 from backend.blocks.llm import (
    DEFAULT_LLM_MODEL,
    TEST_CREDENTIALS,
@@ -17,6 +11,12 @@ from backend.blocks.llm import (
    LLMResponse,
    llm_call,
 )
+from backend.data.block import (
+    BlockCategory,
+    BlockOutput,
+    BlockSchemaInput,
+    BlockSchemaOutput,
+)
 from backend.data.model import APIKeyCredentials, NodeExecutionStats, SchemaField


--- a/autogpt_platform/backend/backend/blocks/ai_image_customizer.py
+++ b/autogpt_platform/backend/backend/blocks/ai_image_customizer.py
@@ -6,7 +6,7 @@ from pydantic import SecretStr
 from replicate.client import Client as ReplicateClient
 from replicate.helpers import FileOutput

-from backend.blocks._base import (
+from backend.data.block import (
    Block,
    BlockCategory,
    BlockOutput,
--- a/autogpt_platform/backend/backend/blocks/ai_image_generator_block.py
+++ b/autogpt_platform/backend/backend/blocks/ai_image_generator_block.py
@@ -5,12 +5,7 @@ from pydantic import SecretStr
 from replicate.client import Client as ReplicateClient
 from replicate.helpers import FileOutput

-from backend.blocks._base import (
-    Block,
-    BlockCategory,
-    BlockSchemaInput,
-    BlockSchemaOutput,
-)
+from backend.data.block import Block, BlockCategory, BlockSchemaInput, BlockSchemaOutput
 from backend.data.execution import ExecutionContext
 from backend.data.model import (
    APIKeyCredentials,
--- a/autogpt_platform/backend/backend/blocks/ai_music_generator.py
+++ b/autogpt_platform/backend/backend/blocks/ai_music_generator.py
@@ -6,7 +6,7 @@ from typing import Literal
 from pydantic import SecretStr
 from replicate.client import Client as ReplicateClient

-from backend.blocks._base import (
+from backend.data.block import (
    Block,
    BlockCategory,
    BlockOutput,
--- a/autogpt_platform/backend/backend/blocks/ai_shortform_video_block.py
+++ b/autogpt_platform/backend/backend/blocks/ai_shortform_video_block.py
@@ -6,7 +6,7 @@ from typing import Literal

 from pydantic import SecretStr

-from backend.blocks._base import (
+from backend.data.block import (
    Block,
    BlockCategory,
    BlockOutput,
--- a/autogpt_platform/backend/backend/blocks/apollo/organization.py
+++ b/autogpt_platform/backend/backend/blocks/apollo/organization.py
@@ -1,10 +1,3 @@
-from backend.blocks._base import (
-    Block,
-    BlockCategory,
-    BlockOutput,
-    BlockSchemaInput,
-    BlockSchemaOutput,
-)
 from backend.blocks.apollo._api import ApolloClient
 from backend.blocks.apollo._auth import (
    TEST_CREDENTIALS,
@@ -17,6 +10,13 @@ from backend.blocks.apollo.models import (
    PrimaryPhone,
    SearchOrganizationsRequest,
 )
+from backend.data.block import (
+    Block,
+    BlockCategory,
+    BlockOutput,
+    BlockSchemaInput,
+    BlockSchemaOutput,
+)
 from backend.data.model import CredentialsField, SchemaField


--- a/autogpt_platform/backend/backend/blocks/apollo/people.py
+++ b/autogpt_platform/backend/backend/blocks/apollo/people.py
@@ -1,12 +1,5 @@
 import asyncio

-from backend.blocks._base import (
-    Block,
-    BlockCategory,
-    BlockOutput,
-    BlockSchemaInput,
-    BlockSchemaOutput,
-)
 from backend.blocks.apollo._api import ApolloClient
 from backend.blocks.apollo._auth import (
    TEST_CREDENTIALS,
@@ -21,6 +14,13 @@ from backend.blocks.apollo.models import (
    SearchPeopleRequest,
    SenorityLevels,
 )
+from backend.data.block import (
+    Block,
+    BlockCategory,
+    BlockOutput,
+    BlockSchemaInput,
+    BlockSchemaOutput,
+)
 from backend.data.model import CredentialsField, SchemaField


--- a/autogpt_platform/backend/backend/blocks/apollo/person.py
+++ b/autogpt_platform/backend/backend/blocks/apollo/person.py
@@ -1,10 +1,3 @@
-from backend.blocks._base import (
-    Block,
-    BlockCategory,
-    BlockOutput,
-    BlockSchemaInput,
-    BlockSchemaOutput,
-)
 from backend.blocks.apollo._api import ApolloClient
 from backend.blocks.apollo._auth import (
    TEST_CREDENTIALS,
@@ -13,6 +6,13 @@ from backend.blocks.apollo._auth import (
    ApolloCredentialsInput,
 )
 from backend.blocks.apollo.models import Contact, EnrichPersonRequest
+from backend.data.block import (
+    Block,
+    BlockCategory,
+    BlockOutput,
+    BlockSchemaInput,
+    BlockSchemaOutput,
+)
 from backend.data.model import CredentialsField, SchemaField


--- a/autogpt_platform/backend/backend/blocks/ayrshare/_util.py
+++ b/autogpt_platform/backend/backend/blocks/ayrshare/_util.py
@@ -3,7 +3,7 @@ from typing import Optional

 from pydantic import BaseModel, Field

-from backend.blocks._base import BlockSchemaInput
+from backend.data.block import BlockSchemaInput
 from backend.data.model import SchemaField, UserIntegrations
 from backend.integrations.ayrshare import AyrshareClient
 from backend.util.clients import get_database_manager_async_client
--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
Otto	bddb7d213c	Merge branch 'dev' into otto/secrt-1887-content-based-detection	2026-02-09 08:04:01 +00:00
Otto	b739f36826	fix(binary-processor): add virus scanning + tests for offset loop Security fix: Add scan_content_safe() call before writing files to prevent bypass of ClamAV scanning infrastructure. Malicious base64-encoded content matching a supported magic number would previously be saved without scanning. Test coverage: - Add tests for 4-byte aligned offset loop (handles marker bleed-in) - Add tests for PDF_BASE64_START/END marker expansion - Add tests verifying virus scanner integration - Mock virus scanner in existing process_binary_outputs tests Addresses review feedback from Claude sub-agents and ntindle.	2026-02-05 23:29:57 +00:00
Toran Bruce Richards	4e75279ea3	Merge branch 'dev' into otto/secrt-1887-content-based-detection	2026-02-05 23:50:58 +01:00
Otto	cbf3f16792	fix(binary-processor): handle marker text bleeding into base64 match When markers like PDF_BASE64_START are concatenated directly with base64 content (no delimiter), the regex matches both together. This corrupts the decode because 'START' becomes part of the base64 string. Fix: Try decoding from multiple 4-byte aligned offsets (0, 4, 8, ... 32) until we find valid magic numbers. Also add PDF_BASE64_START/END to the marker patterns for cleanup. Tested with: PDF_BASE64_STARTJVBERi0... (no delimiter case)	2026-02-05 22:46:15 +00:00
Otto	950f950985	style: Address review nits - Remove redundant extension from filename (was: block_png_uuid.png, now: block_uuid.png) - Use lazy logging format instead of f-string	2026-02-05 20:26:36 +00:00
Otto	4fe74d8e7f	fix: Correct test position calculation for marker expansion The test had wrong indices - ABCDEF starts at position 26, not 27.	2026-02-05 20:25:42 +00:00
Otto	513ec16357	fix: Handle RFC 2045 line-wrapped base64 in embedded detection Sentry correctly identified that the regex and decode would fail on line-wrapped base64. Fixed by: - Updated regex to allow whitespace in base64 chunks - Strip whitespace before decoding with validate=True - Added test for line-wrapped base64	2026-02-05 20:00:09 +00:00
Toran Bruce Richards	8cbf07ce40	Merge branch 'dev' into otto/secrt-1887-content-based-detection	2026-02-05 20:53:18 +01:00
Otto	9914d0ef2d	refactor: Focus binary detection on embedded base64 in stdout_logs Previous implementation detected data URIs and pure base64 strings, but these cases already worked in CoPilot. The actual problem was code execution printing base64 to stdout with markers. Changes: - Remove data URI detection (no user-facing impact) - Remove pure base64 detection (covered by embedded detection) - Add embedded base64 pattern matching within text strings - Handle marker patterns (---BASE64_START---/---BASE64_END---) - Replace base64 + markers with workspace:// reference This now solves the actual problem: ExecuteCodeBlock stdout_logs containing embedded base64 that would otherwise cost 17k+ output tokens for the LLM to re-type.	2026-02-05 19:52:14 +00:00
Otto	c32c7fb959	fix: Handle RFC 2045 line-wrapped base64 in binary detection - Strip whitespace before base64 decoding (encoders often add line breaks) - Update regex comment to clarify whitespace handling - Add test case for line-wrapped base64 Addresses CodeRabbit review feedback.	2026-02-05 19:11:56 +00:00
Otto	dcb3550910	feat(copilot): Auto-save binary block outputs using content-based detection When CoPilot executes blocks that produce binary outputs (images, PDFs), the data is now automatically detected and saved to the user's workspace, replacing raw base64 data with workspace:// references. Uses content-based detection (not field-name based) because: - Code execution blocks return base64 in stdout_logs, not structured fields - The png/jpeg/pdf fields only populate from Jupyter display mechanisms - Other blocks use various field names (image, result, output, etc.) Detection strategy: 1. Data URI detection with mimetype whitelist 2. Raw base64 with magic number validation (PNG, JPEG, PDF, GIF, WebP) 3. Size threshold > 1KB to filter tokens/hashes Features: - Scans ALL string values recursively, regardless of field name - Content deduplication within single block execution via SHA-256 hash - Graceful degradation (original value preserved on save failure) - ~97% token reduction observed (17k -> 500 tokens for PDF generation) Closes SECRT-1887	2026-02-05 18:39:31 +00:00