fix: pass webhook request metadata explicitly

fix: normalize signed call depth for root webhook URLs
fix: propagate workflow call depth through HTTP recursion
2026-03-20 04:51:37 +08:00 · 2026-03-20 04:27:22 +08:00 · 2026-03-20 04:03:00 +08:00 · 2026-03-20 01:13:26 +09:00 · 2026-03-20 00:16:44 +09:00 · 2026-03-19 22:03:07 +08:00
1137 changed files with 41552 additions and 14600 deletions
@@ -63,7 +63,8 @@ pnpm analyze-component <path> --review

 ### File Naming

- Test files: `ComponentName.spec.tsx` (same directory as component)
+- Test files: `ComponentName.spec.tsx` inside a same-level `__tests__/` directory
+- Placement rule: Component, hook, and utility tests must live in a sibling `__tests__/` folder at the same level as the source under test. For example, `foo/index.tsx` maps to `foo/__tests__/index.spec.tsx`, and `foo/bar.ts` maps to `foo/__tests__/bar.spec.ts`.
 - Integration tests: `web/__tests__/` directory

 ## Test Structure Template
@@ -41,7 +41,7 @@ import userEvent from '@testing-library/user-event'
 // Router (if component uses useRouter, usePathname, useSearchParams)
 // WHY: Isolates tests from Next.js routing, enables testing navigation behavior
 // const mockPush = vi.fn()
-// vi.mock('next/navigation', () => ({
+// vi.mock('@/next/navigation', () => ({
 //   useRouter: () => ({ push: mockPush }),
 //   usePathname: () => '/test-path',
 // }))
@@ -4,10 +4,10 @@ runs:
  using: composite
  steps:
    - name: Setup Vite+
-      uses: voidzero-dev/setup-vp@b5d848f5a62488f3d3d920f8aa6ac318a60c5f07 # v1
+      uses: voidzero-dev/setup-vp@4a524139920f87f9f7080d3b8545acac019e1852 # v1.0.0
      with:
-        node-version-file: "./web/.nvmrc"
+        node-version-file: web/.nvmrc
        cache: true
+        cache-dependency-path: web/pnpm-lock.yaml
        run-install: |
-          - cwd: ./web
-            args: ['--frozen-lockfile']
+          cwd: ./web
@@ -12,7 +12,7 @@ jobs:
  anti-slop:
    runs-on: ubuntu-latest
    steps:
-      - uses: peakoss/anti-slop@v0
+      - uses: peakoss/anti-slop@85daca1880e9e1af197fc06ea03349daf08f4202 # v0.2.1
        with:
          github-token: ${{ secrets.GITHUB_TOKEN }}
          close-pr: false
@@ -2,6 +2,12 @@ name: Run Pytest

 on:
  workflow_call:
+    secrets:
+      CODECOV_TOKEN:
+        required: false
+
+permissions:
+  contents: read

 concurrency:
  group: api-tests-${{ github.head_ref || github.run_id }}
@@ -11,6 +17,8 @@ jobs:
  test:
    name: API Tests
    runs-on: ubuntu-latest
+    env:
+      CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
    defaults:
      run:
        shell: bash
@@ -24,10 +32,11 @@ jobs:
      - name: Checkout code
        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
        with:
+          fetch-depth: 0
          persist-credentials: false

      - name: Setup UV and Python
-        uses: astral-sh/setup-uv@e06108dd0aef18192324c70427afc47652e63a82 # v7.5.0
+        uses: astral-sh/setup-uv@37802adc94f370d6bfd71619e3f0bf239e1f3b78 # v7.6.0
        with:
          enable-cache: true
          python-version: ${{ matrix.python-version }}
@@ -79,21 +88,12 @@ jobs:
            api/tests/test_containers_integration_tests \
            api/tests/unit_tests

-      - name: Coverage Summary
-        run: |
-          set -x
-          # Extract coverage percentage and create a summary
-          TOTAL_COVERAGE=$(python -c 'import json; print(json.load(open("coverage.json"))["totals"]["percent_covered_display"])')
-
-          # Create a detailed coverage summary
-          echo "### Test Coverage Summary :test_tube:" >> $GITHUB_STEP_SUMMARY
-          echo "Total Coverage: ${TOTAL_COVERAGE}%" >> $GITHUB_STEP_SUMMARY
-          {
-            echo ""
-            echo "<details><summary>File-level coverage (click to expand)</summary>"
-            echo ""
-            echo '```'
-            uv run --project api coverage report -m
-            echo '```'
-            echo "</details>"
-          } >> $GITHUB_STEP_SUMMARY
+      - name: Report coverage
+        if: ${{ env.CODECOV_TOKEN != '' && matrix.python-version == '3.12' }}
+        uses: codecov/codecov-action@1af58845a975a7985b0beb0cbe6fbbb71a41dbad # v5.5.3
+        with:
+          files: ./coverage.xml
+          disable_search: true
+          flags: api
+        env:
+          CODECOV_TOKEN: ${{ env.CODECOV_TOKEN }}
@@ -39,7 +39,7 @@ jobs:
        with:
          python-version: "3.11"

-      - uses: astral-sh/setup-uv@e06108dd0aef18192324c70427afc47652e63a82 # v7.5.0
+      - uses: astral-sh/setup-uv@37802adc94f370d6bfd71619e3f0bf239e1f3b78 # v7.6.0

      - name: Generate Docker Compose
        if: steps.docker-compose-changes.outputs.any_changed == 'true'
@@ -19,7 +19,7 @@ jobs:
          persist-credentials: false

      - name: Setup UV and Python
-        uses: astral-sh/setup-uv@e06108dd0aef18192324c70427afc47652e63a82 # v7.5.0
+        uses: astral-sh/setup-uv@37802adc94f370d6bfd71619e3f0bf239e1f3b78 # v7.6.0
        with:
          enable-cache: true
          python-version: "3.12"
@@ -69,7 +69,7 @@ jobs:
          persist-credentials: false

      - name: Setup UV and Python
-        uses: astral-sh/setup-uv@e06108dd0aef18192324c70427afc47652e63a82 # v7.5.0
+        uses: astral-sh/setup-uv@37802adc94f370d6bfd71619e3f0bf239e1f3b78 # v7.6.0
        with:
          enable-cache: true
          python-version: "3.12"
@@ -56,16 +56,14 @@ jobs:
    needs: check-changes
    if: needs.check-changes.outputs.api-changed == 'true'
    uses: ./.github/workflows/api-tests.yml
+    secrets: inherit

  web-tests:
    name: Web Tests
    needs: check-changes
    if: needs.check-changes.outputs.web-changed == 'true'
    uses: ./.github/workflows/web-tests.yml
-    with:
-      base_sha: ${{ github.event.before || github.event.pull_request.base.sha }}
-      diff_range_mode: ${{ github.event.before && 'exact' || 'merge-base' }}
-      head_sha: ${{ github.event.after || github.event.pull_request.head.sha || github.sha }}
+    secrets: inherit

  style-check:
    name: Style Check
@@ -22,7 +22,7 @@ jobs:
          fetch-depth: 0

      - name: Setup Python & UV
-        uses: astral-sh/setup-uv@e06108dd0aef18192324c70427afc47652e63a82 # v7.5.0
+        uses: astral-sh/setup-uv@37802adc94f370d6bfd71619e3f0bf239e1f3b78 # v7.6.0
        with:
          enable-cache: true

@@ -33,7 +33,7 @@ jobs:

      - name: Setup UV and Python
        if: steps.changed-files.outputs.any_changed == 'true'
-        uses: astral-sh/setup-uv@e06108dd0aef18192324c70427afc47652e63a82 # v7.5.0
+        uses: astral-sh/setup-uv@37802adc94f370d6bfd71619e3f0bf239e1f3b78 # v7.6.0
        with:
          enable-cache: false
          python-version: "3.12"
@@ -120,7 +120,7 @@ jobs:

      - name: Run Claude Code for Translation Sync
        if: steps.detect_changes.outputs.CHANGED_FILES != ''
-        uses: anthropics/claude-code-action@cd77b50d2b0808657f8e6774085c8bf54484351c # v1.0.72
+        uses: anthropics/claude-code-action@df37d2f0760a4b5683a6e617c9325bc1a36443f6 # v1.0.75
        with:
          anthropic_api_key: ${{ secrets.ANTHROPIC_API_KEY }}
          github_token: ${{ secrets.GITHUB_TOKEN }}
@@ -31,7 +31,7 @@ jobs:
          remove_tool_cache: true

      - name: Setup UV and Python
-        uses: astral-sh/setup-uv@e06108dd0aef18192324c70427afc47652e63a82 # v7.5.0
+        uses: astral-sh/setup-uv@37802adc94f370d6bfd71619e3f0bf239e1f3b78 # v7.6.0
        with:
          enable-cache: true
          python-version: ${{ matrix.python-version }}
@@ -2,16 +2,9 @@ name: Web Tests

 on:
  workflow_call:
-    inputs:
-      base_sha:
+    secrets:
+      CODECOV_TOKEN:
        required: false
-        type: string
-      diff_range_mode:
-        required: false
-        type: string
-      head_sha:
-        required: false
-        type: string

 permissions:
  contents: read
@@ -29,8 +22,8 @@ jobs:
    strategy:
      fail-fast: false
      matrix:
-        shardIndex: [1, 2, 3, 4]
-        shardTotal: [4]
+        shardIndex: [1, 2, 3, 4, 5, 6]
+        shardTotal: [6]
    defaults:
      run:
        shell: bash
@@ -63,7 +56,7 @@ jobs:
    needs: [test]
    runs-on: ubuntu-latest
    env:
-      VITEST_COVERAGE_SCOPE: app-components
+      CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
    defaults:
      run:
        shell: bash
@@ -87,52 +80,16 @@ jobs:
          merge-multiple: true

      - name: Merge reports
-        run: vp test --merge-reports --reporter=json --reporter=agent --coverage
+        run: vp test --merge-reports --coverage --silent=passed-only

-      - name: Report app/components baseline coverage
-        run: node ./scripts/report-components-coverage-baseline.mjs
-
-      - name: Report app/components test touch
-        env:
-          BASE_SHA: ${{ inputs.base_sha }}
-          DIFF_RANGE_MODE: ${{ inputs.diff_range_mode }}
-          HEAD_SHA: ${{ inputs.head_sha }}
-        run: node ./scripts/report-components-test-touch.mjs
-
-      - name: Check app/components pure diff coverage
-        env:
-          BASE_SHA: ${{ inputs.base_sha }}
-          DIFF_RANGE_MODE: ${{ inputs.diff_range_mode }}
-          HEAD_SHA: ${{ inputs.head_sha }}
-        run: node ./scripts/check-components-diff-coverage.mjs
-
-      - name: Check Coverage Summary
-        if: always()
-        id: coverage-summary
-        run: |
-          set -eo pipefail
-
-          COVERAGE_FILE="coverage/coverage-final.json"
-          COVERAGE_SUMMARY_FILE="coverage/coverage-summary.json"
-
-          if [ -f "$COVERAGE_FILE" ] || [ -f "$COVERAGE_SUMMARY_FILE" ]; then
-            echo "has_coverage=true" >> "$GITHUB_OUTPUT"
-            exit 0
-          fi
-
-          echo "has_coverage=false" >> "$GITHUB_OUTPUT"
-          echo "### 🚨 app/components Diff Coverage" >> "$GITHUB_STEP_SUMMARY"
-          echo "" >> "$GITHUB_STEP_SUMMARY"
-          echo "Coverage artifacts not found. Ensure Vitest merge reports ran with coverage enabled." >> "$GITHUB_STEP_SUMMARY"
-
-      - name: Upload Coverage Artifact
-        if: steps.coverage-summary.outputs.has_coverage == 'true'
-        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
+      - name: Report coverage
+        if: ${{ env.CODECOV_TOKEN != '' }}
+        uses: codecov/codecov-action@1af58845a975a7985b0beb0cbe6fbbb71a41dbad # v5.5.3
        with:
-          name: web-coverage-report
-          path: web/coverage
-          retention-days: 30
-          if-no-files-found: error
+          directory: web/coverage
+          flags: web
+        env:
+          CODECOV_TOKEN: ${{ env.CODECOV_TOKEN }}

  web-build:
    name: Web Build
@@ -1,9 +1,11 @@
 import json
 import logging
-from typing import Any
+from typing import Any, cast

 import click
 from pydantic import TypeAdapter
+from sqlalchemy import delete, select
+from sqlalchemy.engine import CursorResult

 from configs import dify_config
 from core.helper import encrypter
@@ -48,14 +50,15 @@ def setup_system_tool_oauth_client(provider, client_params):
        click.echo(click.style(f"Error parsing client params: {str(e)}", fg="red"))
        return

-    deleted_count = (
-        db.session.query(ToolOAuthSystemClient)
-        .filter_by(
-            provider=provider_name,
-            plugin_id=plugin_id,
-        )
-        .delete()
-    )
+    deleted_count = cast(
+        CursorResult,
+        db.session.execute(
+            delete(ToolOAuthSystemClient).where(
+                ToolOAuthSystemClient.provider == provider_name,
+                ToolOAuthSystemClient.plugin_id == plugin_id,
+            )
+        ),
+    ).rowcount
    if deleted_count > 0:
        click.echo(click.style(f"Deleted {deleted_count} existing oauth client params.", fg="yellow"))

@@ -97,14 +100,15 @@ def setup_system_trigger_oauth_client(provider, client_params):
        click.echo(click.style(f"Error parsing client params: {str(e)}", fg="red"))
        return

-    deleted_count = (
-        db.session.query(TriggerOAuthSystemClient)
-        .filter_by(
-            provider=provider_name,
-            plugin_id=plugin_id,
-        )
-        .delete()
-    )
+    deleted_count = cast(
+        CursorResult,
+        db.session.execute(
+            delete(TriggerOAuthSystemClient).where(
+                TriggerOAuthSystemClient.provider == provider_name,
+                TriggerOAuthSystemClient.plugin_id == plugin_id,
+            )
+        ),
+    ).rowcount
    if deleted_count > 0:
        click.echo(click.style(f"Deleted {deleted_count} existing oauth client params.", fg="yellow"))

@@ -139,14 +143,15 @@ def setup_datasource_oauth_client(provider, client_params):
        return

    click.echo(click.style(f"Ready to delete existing oauth client params: {provider_name}", fg="yellow"))
-    deleted_count = (
-        db.session.query(DatasourceOauthParamConfig)
-        .filter_by(
-            provider=provider_name,
-            plugin_id=plugin_id,
-        )
-        .delete()
-    )
+    deleted_count = cast(
+        CursorResult,
+        db.session.execute(
+            delete(DatasourceOauthParamConfig).where(
+                DatasourceOauthParamConfig.provider == provider_name,
+                DatasourceOauthParamConfig.plugin_id == plugin_id,
+            )
+        ),
+    ).rowcount
    if deleted_count > 0:
        click.echo(click.style(f"Deleted {deleted_count} existing oauth client params.", fg="yellow"))

@@ -192,7 +197,9 @@ def transform_datasource_credentials(environment: str):

        # deal notion credentials
        deal_notion_count = 0
-        notion_credentials = db.session.query(DataSourceOauthBinding).filter_by(provider="notion").all()
+        notion_credentials = db.session.scalars(
+            select(DataSourceOauthBinding).where(DataSourceOauthBinding.provider == "notion")
+        ).all()
        if notion_credentials:
            notion_credentials_tenant_mapping: dict[str, list[DataSourceOauthBinding]] = {}
            for notion_credential in notion_credentials:
@@ -201,7 +208,7 @@ def transform_datasource_credentials(environment: str):
                    notion_credentials_tenant_mapping[tenant_id] = []
                notion_credentials_tenant_mapping[tenant_id].append(notion_credential)
            for tenant_id, notion_tenant_credentials in notion_credentials_tenant_mapping.items():
-                tenant = db.session.query(Tenant).filter_by(id=tenant_id).first()
+                tenant = db.session.scalar(select(Tenant).where(Tenant.id == tenant_id))
                if not tenant:
                    continue
                try:
@@ -250,7 +257,9 @@ def transform_datasource_credentials(environment: str):
                db.session.commit()
        # deal firecrawl credentials
        deal_firecrawl_count = 0
-        firecrawl_credentials = db.session.query(DataSourceApiKeyAuthBinding).filter_by(provider="firecrawl").all()
+        firecrawl_credentials = db.session.scalars(
+            select(DataSourceApiKeyAuthBinding).where(DataSourceApiKeyAuthBinding.provider == "firecrawl")
+        ).all()
        if firecrawl_credentials:
            firecrawl_credentials_tenant_mapping: dict[str, list[DataSourceApiKeyAuthBinding]] = {}
            for firecrawl_credential in firecrawl_credentials:
@@ -259,7 +268,7 @@ def transform_datasource_credentials(environment: str):
                    firecrawl_credentials_tenant_mapping[tenant_id] = []
                firecrawl_credentials_tenant_mapping[tenant_id].append(firecrawl_credential)
            for tenant_id, firecrawl_tenant_credentials in firecrawl_credentials_tenant_mapping.items():
-                tenant = db.session.query(Tenant).filter_by(id=tenant_id).first()
+                tenant = db.session.scalar(select(Tenant).where(Tenant.id == tenant_id))
                if not tenant:
                    continue
                try:
@@ -312,7 +321,9 @@ def transform_datasource_credentials(environment: str):
                db.session.commit()
        # deal jina credentials
        deal_jina_count = 0
-        jina_credentials = db.session.query(DataSourceApiKeyAuthBinding).filter_by(provider="jinareader").all()
+        jina_credentials = db.session.scalars(
+            select(DataSourceApiKeyAuthBinding).where(DataSourceApiKeyAuthBinding.provider == "jinareader")
+        ).all()
        if jina_credentials:
            jina_credentials_tenant_mapping: dict[str, list[DataSourceApiKeyAuthBinding]] = {}
            for jina_credential in jina_credentials:
@@ -321,7 +332,7 @@ def transform_datasource_credentials(environment: str):
                    jina_credentials_tenant_mapping[tenant_id] = []
                jina_credentials_tenant_mapping[tenant_id].append(jina_credential)
            for tenant_id, jina_tenant_credentials in jina_credentials_tenant_mapping.items():
-                tenant = db.session.query(Tenant).filter_by(id=tenant_id).first()
+                tenant = db.session.scalar(select(Tenant).where(Tenant.id == tenant_id))
                if not tenant:
                    continue
                try:
@@ -1,7 +1,10 @@
 import json
+from typing import cast

 import click
 import sqlalchemy as sa
+from sqlalchemy import update
+from sqlalchemy.engine import CursorResult

 from configs import dify_config
 from extensions.ext_database import db
@@ -740,14 +743,17 @@ def migrate_oss(
        else:
            try:
                source_storage_type = StorageType.LOCAL if is_source_local else StorageType.OPENDAL
-                updated = (
-                    db.session.query(UploadFile)
-                    .where(
-                        UploadFile.storage_type == source_storage_type,
-                        UploadFile.key.in_(copied_upload_file_keys),
-                    )
-                    .update({UploadFile.storage_type: dify_config.STORAGE_TYPE}, synchronize_session=False)
-                )
+                updated = cast(
+                    CursorResult,
+                    db.session.execute(
+                        update(UploadFile)
+                        .where(
+                            UploadFile.storage_type == source_storage_type,
+                            UploadFile.key.in_(copied_upload_file_keys),
+                        )
+                        .values(storage_type=dify_config.STORAGE_TYPE)
+                    ),
+                ).rowcount
                db.session.commit()
                click.echo(click.style(f"Updated storage_type for {updated} upload_files records.", fg="green"))
            except Exception as e:
@@ -2,6 +2,7 @@ import logging

 import click
 import sqlalchemy as sa
+from sqlalchemy import delete, select, update
 from sqlalchemy.orm import sessionmaker

 from configs import dify_config
@@ -41,7 +42,7 @@ def reset_encrypt_key_pair():
        click.echo(click.style("This command is only for SELF_HOSTED installations.", fg="red"))
        return
    with sessionmaker(db.engine, expire_on_commit=False).begin() as session:
-        tenants = session.query(Tenant).all()
+        tenants = session.scalars(select(Tenant)).all()
        for tenant in tenants:
            if not tenant:
                click.echo(click.style("No workspaces found. Run /install first.", fg="red"))
@@ -49,8 +50,8 @@ def reset_encrypt_key_pair():

            tenant.encrypt_public_key = generate_key_pair(tenant.id)

-            session.query(Provider).where(Provider.provider_type == "custom", Provider.tenant_id == tenant.id).delete()
-            session.query(ProviderModel).where(ProviderModel.tenant_id == tenant.id).delete()
+            session.execute(delete(Provider).where(Provider.provider_type == "custom", Provider.tenant_id == tenant.id))
+            session.execute(delete(ProviderModel).where(ProviderModel.tenant_id == tenant.id))

            click.echo(
                click.style(
@@ -93,7 +94,7 @@ def convert_to_agent_apps():
                app_id = str(i.id)
                if app_id not in proceeded_app_ids:
                    proceeded_app_ids.append(app_id)
-                    app = db.session.query(App).where(App.id == app_id).first()
+                    app = db.session.scalar(select(App).where(App.id == app_id))
                    if app is not None:
                        apps.append(app)

@@ -108,8 +109,8 @@ def convert_to_agent_apps():
                db.session.commit()

                # update conversation mode to agent
-                db.session.query(Conversation).where(Conversation.app_id == app.id).update(
-                    {Conversation.mode: AppMode.AGENT_CHAT}
+                db.session.execute(
+                    update(Conversation).where(Conversation.app_id == app.id).values(mode=AppMode.AGENT_CHAT)
                )

                db.session.commit()
@@ -177,7 +178,7 @@ where sites.id is null limit 1000"""
                    continue

                try:
-                    app = db.session.query(App).where(App.id == app_id).first()
+                    app = db.session.scalar(select(App).where(App.id == app_id))
                    if not app:
                        logger.info("App %s not found", app_id)
                        continue
@@ -14,6 +14,7 @@ from core.rag.models.document import ChildDocument, Document
 from extensions.ext_database import db
 from models.dataset import Dataset, DatasetCollectionBinding, DatasetMetadata, DatasetMetadataBinding, DocumentSegment
 from models.dataset import Document as DatasetDocument
+from models.enums import DatasetMetadataType, IndexingStatus, SegmentStatus
 from models.model import App, AppAnnotationSetting, MessageAnnotation


@@ -40,14 +41,13 @@ def migrate_annotation_vector_database():
            # get apps info
            per_page = 50
            with sessionmaker(db.engine, expire_on_commit=False).begin() as session:
-                apps = (
-                    session.query(App)
+                apps = session.scalars(
+                    select(App)
                    .where(App.status == "normal")
                    .order_by(App.created_at.desc())
                    .limit(per_page)
                    .offset((page - 1) * per_page)
-                    .all()
-                )
+                ).all()
            if not apps:
                break
        except SQLAlchemyError:
@@ -62,8 +62,8 @@ def migrate_annotation_vector_database():
            try:
                click.echo(f"Creating app annotation index: {app.id}")
                with sessionmaker(db.engine, expire_on_commit=False).begin() as session:
-                    app_annotation_setting = (
-                        session.query(AppAnnotationSetting).where(AppAnnotationSetting.app_id == app.id).first()
+                    app_annotation_setting = session.scalar(
+                        select(AppAnnotationSetting).where(AppAnnotationSetting.app_id == app.id).limit(1)
                    )

                    if not app_annotation_setting:
@@ -71,10 +71,10 @@ def migrate_annotation_vector_database():
                        click.echo(f"App annotation setting disabled: {app.id}")
                        continue
                    # get dataset_collection_binding info
-                    dataset_collection_binding = (
-                        session.query(DatasetCollectionBinding)
-                        .where(DatasetCollectionBinding.id == app_annotation_setting.collection_binding_id)
-                        .first()
+                    dataset_collection_binding = session.scalar(
+                        select(DatasetCollectionBinding).where(
+                            DatasetCollectionBinding.id == app_annotation_setting.collection_binding_id
+                        )
                    )
                    if not dataset_collection_binding:
                        click.echo(f"App annotation collection binding not found: {app.id}")
@@ -204,11 +204,11 @@ def migrate_knowledge_vector_database():
                    collection_name = Dataset.gen_collection_name_by_id(dataset_id)
                elif vector_type == VectorType.QDRANT:
                    if dataset.collection_binding_id:
-                        dataset_collection_binding = (
-                            db.session.query(DatasetCollectionBinding)
-                            .where(DatasetCollectionBinding.id == dataset.collection_binding_id)
-                            .one_or_none()
-                        )
+                        dataset_collection_binding = db.session.execute(
+                            select(DatasetCollectionBinding).where(
+                                DatasetCollectionBinding.id == dataset.collection_binding_id
+                            )
+                        ).scalar_one_or_none()
                        if dataset_collection_binding:
                            collection_name = dataset_collection_binding.collection_name
                        else:
@@ -242,7 +242,7 @@ def migrate_knowledge_vector_database():
                dataset_documents = db.session.scalars(
                    select(DatasetDocument).where(
                        DatasetDocument.dataset_id == dataset.id,
-                        DatasetDocument.indexing_status == "completed",
+                        DatasetDocument.indexing_status == IndexingStatus.COMPLETED,
                        DatasetDocument.enabled == True,
                        DatasetDocument.archived == False,
                    )
@@ -254,7 +254,7 @@ def migrate_knowledge_vector_database():
                    segments = db.session.scalars(
                        select(DocumentSegment).where(
                            DocumentSegment.document_id == dataset_document.id,
-                            DocumentSegment.status == "completed",
+                            DocumentSegment.status == SegmentStatus.COMPLETED,
                            DocumentSegment.enabled == True,
                        )
                    ).all()
@@ -333,7 +333,7 @@ def add_qdrant_index(field: str):
    create_count = 0

    try:
-        bindings = db.session.query(DatasetCollectionBinding).all()
+        bindings = db.session.scalars(select(DatasetCollectionBinding)).all()
        if not bindings:
            click.echo(click.style("No dataset collection bindings found.", fg="red"))
            return
@@ -420,22 +420,22 @@ def old_metadata_migration():
                        if field.value == key:
                            break
                    else:
-                        dataset_metadata = (
-                            db.session.query(DatasetMetadata)
+                        dataset_metadata = db.session.scalar(
+                            select(DatasetMetadata)
                            .where(DatasetMetadata.dataset_id == document.dataset_id, DatasetMetadata.name == key)
-                            .first()
+                            .limit(1)
                        )
                        if not dataset_metadata:
                            dataset_metadata = DatasetMetadata(
                                tenant_id=document.tenant_id,
                                dataset_id=document.dataset_id,
                                name=key,
-                                type="string",
+                                type=DatasetMetadataType.STRING,
                                created_by=document.created_by,
                            )
                            db.session.add(dataset_metadata)
                            db.session.flush()
-                            dataset_metadata_binding = DatasetMetadataBinding(
+                            dataset_metadata_binding: DatasetMetadataBinding | None = DatasetMetadataBinding(
                                tenant_id=document.tenant_id,
                                dataset_id=document.dataset_id,
                                metadata_id=dataset_metadata.id,
@@ -444,14 +444,14 @@ def old_metadata_migration():
                            )
                            db.session.add(dataset_metadata_binding)
                        else:
-                            dataset_metadata_binding = (
-                                db.session.query(DatasetMetadataBinding)  # type: ignore
+                            dataset_metadata_binding = db.session.scalar(
+                                select(DatasetMetadataBinding)
                                .where(
                                    DatasetMetadataBinding.dataset_id == document.dataset_id,
                                    DatasetMetadataBinding.document_id == document.id,
                                    DatasetMetadataBinding.metadata_id == dataset_metadata.id,
                                )
-                                .first()
+                                .limit(1)
                            )
                            if not dataset_metadata_binding:
                                dataset_metadata_binding = DatasetMetadataBinding(
@@ -1,4 +1,4 @@
-from pydantic import Field, NonNegativeInt, PositiveFloat, PositiveInt
+from pydantic import Field, NonNegativeInt, PositiveFloat, PositiveInt, field_validator
 from pydantic_settings import BaseSettings


@@ -116,3 +116,13 @@ class RedisConfig(BaseSettings):
        description="Maximum connections in the Redis connection pool (unset for library default)",
        default=None,
    )
+
+    @field_validator("REDIS_MAX_CONNECTIONS", mode="before")
+    @classmethod
+    def _empty_string_to_none_for_max_conns(cls, v):
+        """Allow empty string in env/.env to mean 'unset' (None)."""
+        if v is None:
+            return None
+        if isinstance(v, str) and v.strip() == "":
+            return None
+        return v
@@ -103,13 +103,13 @@ class AppMCPServerController(Resource):
            raise NotFound()

        description = payload.description
-        if description is None:
-            pass
-        elif not description:
+        if description is None or not description:
            server.description = app_model.description or ""
        else:
            server.description = description

+        server.name = app_model.name
+
        server.parameters = json.dumps(payload.parameters, ensure_ascii=False)
        if payload.status:
            try:
@@ -30,6 +30,7 @@ from fields.raws import FilesContainedField
 from libs.helper import TimestampField, uuid_value
 from libs.infinite_scroll_pagination import InfiniteScrollPagination
 from libs.login import current_account_with_tenant, login_required
+from models.enums import FeedbackFromSource, FeedbackRating
 from models.model import AppMode, Conversation, Message, MessageAnnotation, MessageFeedback
 from services.errors.conversation import ConversationNotExistsError
 from services.errors.message import MessageNotExistsError, SuggestedQuestionsAfterAnswerDisabledError
@@ -335,7 +336,7 @@ class MessageFeedbackApi(Resource):
        if not args.rating and feedback:
            db.session.delete(feedback)
        elif args.rating and feedback:
-            feedback.rating = args.rating
+            feedback.rating = FeedbackRating(args.rating)
            feedback.content = args.content
        elif not args.rating and not feedback:
            raise ValueError("rating cannot be None when feedback not exists")
@@ -347,9 +348,9 @@ class MessageFeedbackApi(Resource):
                app_id=app_model.id,
                conversation_id=message.conversation_id,
                message_id=message.id,
-                rating=rating_value,
+                rating=FeedbackRating(rating_value),
                content=args.content,
-                from_source="admin",
+                from_source=FeedbackFromSource.ADMIN,
                from_account_id=current_user.id,
            )
            db.session.add(feedback)
@@ -54,6 +54,7 @@ from fields.document_fields import document_status_fields
 from libs.login import current_account_with_tenant, login_required
 from models import ApiToken, Dataset, Document, DocumentSegment, UploadFile
 from models.dataset import DatasetPermission, DatasetPermissionEnum
+from models.enums import SegmentStatus
 from models.provider_ids import ModelProviderID
 from services.api_token_service import ApiTokenCache
 from services.dataset_service import DatasetPermissionService, DatasetService, DocumentService
@@ -741,13 +742,15 @@ class DatasetIndexingStatusApi(Resource):
                .where(
                    DocumentSegment.completed_at.isnot(None),
                    DocumentSegment.document_id == str(document.id),
-                    DocumentSegment.status != "re_segment",
+                    DocumentSegment.status != SegmentStatus.RE_SEGMENT,
                )
                .count()
            )
            total_segments = (
                db.session.query(DocumentSegment)
-                .where(DocumentSegment.document_id == str(document.id), DocumentSegment.status != "re_segment")
+                .where(
+                    DocumentSegment.document_id == str(document.id), DocumentSegment.status != SegmentStatus.RE_SEGMENT
+                )
                .count()
            )
            # Create a dictionary with document attributes and additional fields
@@ -42,6 +42,7 @@ from libs.datetime_utils import naive_utc_now
 from libs.login import current_account_with_tenant, login_required
 from models import DatasetProcessRule, Document, DocumentSegment, UploadFile
 from models.dataset import DocumentPipelineExecutionLog
+from models.enums import IndexingStatus, SegmentStatus
 from services.dataset_service import DatasetService, DocumentService
 from services.entities.knowledge_entities.knowledge_entities import KnowledgeConfig, ProcessRule, RetrievalModel
 from services.file_service import FileService
@@ -297,6 +298,7 @@ class DatasetDocumentListApi(Resource):
        if sort == "hit_count":
            sub_query = (
                sa.select(DocumentSegment.document_id, sa.func.sum(DocumentSegment.hit_count).label("total_hit_count"))
+                .where(DocumentSegment.dataset_id == str(dataset_id))
                .group_by(DocumentSegment.document_id)
                .subquery()
            )
@@ -332,13 +334,16 @@ class DatasetDocumentListApi(Resource):
                    .where(
                        DocumentSegment.completed_at.isnot(None),
                        DocumentSegment.document_id == str(document.id),
-                        DocumentSegment.status != "re_segment",
+                        DocumentSegment.status != SegmentStatus.RE_SEGMENT,
                    )
                    .count()
                )
                total_segments = (
                    db.session.query(DocumentSegment)
-                    .where(DocumentSegment.document_id == str(document.id), DocumentSegment.status != "re_segment")
+                    .where(
+                        DocumentSegment.document_id == str(document.id),
+                        DocumentSegment.status != SegmentStatus.RE_SEGMENT,
+                    )
                    .count()
                )
                document.completed_segments = completed_segments
@@ -503,7 +508,7 @@ class DocumentIndexingEstimateApi(DocumentResource):
        document_id = str(document_id)
        document = self.get_document(dataset_id, document_id)

-        if document.indexing_status in {"completed", "error"}:
+        if document.indexing_status in {IndexingStatus.COMPLETED, IndexingStatus.ERROR}:
            raise DocumentAlreadyFinishedError()

        data_process_rule = document.dataset_process_rule
@@ -573,7 +578,7 @@ class DocumentBatchIndexingEstimateApi(DocumentResource):
        data_process_rule_dict = data_process_rule.to_dict() if data_process_rule else {}
        extract_settings = []
        for document in documents:
-            if document.indexing_status in {"completed", "error"}:
+            if document.indexing_status in {IndexingStatus.COMPLETED, IndexingStatus.ERROR}:
                raise DocumentAlreadyFinishedError()
            data_source_info = document.data_source_info_dict
            match document.data_source_type:
@@ -671,19 +676,21 @@ class DocumentBatchIndexingStatusApi(DocumentResource):
                .where(
                    DocumentSegment.completed_at.isnot(None),
                    DocumentSegment.document_id == str(document.id),
-                    DocumentSegment.status != "re_segment",
+                    DocumentSegment.status != SegmentStatus.RE_SEGMENT,
                )
                .count()
            )
            total_segments = (
                db.session.query(DocumentSegment)
-                .where(DocumentSegment.document_id == str(document.id), DocumentSegment.status != "re_segment")
+                .where(
+                    DocumentSegment.document_id == str(document.id), DocumentSegment.status != SegmentStatus.RE_SEGMENT
+                )
                .count()
            )
            # Create a dictionary with document attributes and additional fields
            document_dict = {
                "id": document.id,
-                "indexing_status": "paused" if document.is_paused else document.indexing_status,
+                "indexing_status": IndexingStatus.PAUSED if document.is_paused else document.indexing_status,
                "processing_started_at": document.processing_started_at,
                "parsing_completed_at": document.parsing_completed_at,
                "cleaning_completed_at": document.cleaning_completed_at,
@@ -720,20 +727,20 @@ class DocumentIndexingStatusApi(DocumentResource):
            .where(
                DocumentSegment.completed_at.isnot(None),
                DocumentSegment.document_id == str(document_id),
-                DocumentSegment.status != "re_segment",
+                DocumentSegment.status != SegmentStatus.RE_SEGMENT,
            )
            .count()
        )
        total_segments = (
            db.session.query(DocumentSegment)
-            .where(DocumentSegment.document_id == str(document_id), DocumentSegment.status != "re_segment")
+            .where(DocumentSegment.document_id == str(document_id), DocumentSegment.status != SegmentStatus.RE_SEGMENT)
            .count()
        )

        # Create a dictionary with document attributes and additional fields
        document_dict = {
            "id": document.id,
-            "indexing_status": "paused" if document.is_paused else document.indexing_status,
+            "indexing_status": IndexingStatus.PAUSED if document.is_paused else document.indexing_status,
            "processing_started_at": document.processing_started_at,
            "parsing_completed_at": document.parsing_completed_at,
            "cleaning_completed_at": document.cleaning_completed_at,
@@ -955,7 +962,7 @@ class DocumentProcessingApi(DocumentResource):

        match action:
            case "pause":
-                if document.indexing_status != "indexing":
+                if document.indexing_status != IndexingStatus.INDEXING:
                    raise InvalidActionError("Document not in indexing state.")

                document.paused_by = current_user.id
@@ -964,7 +971,7 @@ class DocumentProcessingApi(DocumentResource):
                db.session.commit()

            case "resume":
-                if document.indexing_status not in {"paused", "error"}:
+                if document.indexing_status not in {IndexingStatus.PAUSED, IndexingStatus.ERROR}:
                    raise InvalidActionError("Document not in paused or error state.")

                document.paused_by = None
@@ -1169,7 +1176,7 @@ class DocumentRetryApi(DocumentResource):
                    raise ArchivedDocumentImmutableError()

                # 400 if document is completed
-                if document.indexing_status == "completed":
+                if document.indexing_status == IndexingStatus.COMPLETED:
                    raise DocumentAlreadyFinishedError()
                retry_documents.append(document)
            except Exception:
@@ -24,6 +24,7 @@ from fields.hit_testing_fields import hit_testing_record_fields
 from libs.login import current_user
 from models.account import Account
 from services.dataset_service import DatasetService
+from services.entities.knowledge_entities.knowledge_entities import RetrievalModel
 from services.hit_testing_service import HitTestingService

 logger = logging.getLogger(__name__)
@@ -31,7 +32,7 @@ logger = logging.getLogger(__name__)

 class HitTestingPayload(BaseModel):
    query: str = Field(max_length=250)
-    retrieval_model: dict[str, Any] | None = None
+    retrieval_model: RetrievalModel | None = None
    external_retrieval_model: dict[str, Any] | None = None
    attachment_ids: list[str] | None = None

@@ -46,6 +46,8 @@ class PipelineTemplateDetailApi(Resource):
        type = request.args.get("type", default="built-in", type=str)
        rag_pipeline_service = RagPipelineService()
        pipeline_template = rag_pipeline_service.get_pipeline_template_detail(template_id, type)
+        if pipeline_template is None:
+            return {"error": "Pipeline template not found from upstream service."}, 404
        return pipeline_template, 200


@@ -4,6 +4,7 @@ from flask_restx import Resource
 from controllers.console import api
 from controllers.console.explore.wraps import explore_banner_enabled
 from extensions.ext_database import db
+from models.enums import BannerStatus
 from models.model import ExporleBanner


@@ -16,7 +17,7 @@ class BannerApi(Resource):
        language = request.args.get("language", "en-US")

        # Build base query for enabled banners
-        base_query = db.session.query(ExporleBanner).where(ExporleBanner.status == "enabled")
+        base_query = db.session.query(ExporleBanner).where(ExporleBanner.status == BannerStatus.ENABLED)

        # Try to get banners in the requested language
        banners = base_query.where(ExporleBanner.language == language).order_by(ExporleBanner.sort).all()
@@ -27,6 +27,7 @@ from fields.message_fields import MessageInfiniteScrollPagination, MessageListIt
 from libs import helper
 from libs.helper import UUIDStrOrEmpty
 from libs.login import current_account_with_tenant
+from models.enums import FeedbackRating
 from models.model import AppMode
 from services.app_generate_service import AppGenerateService
 from services.errors.app import MoreLikeThisDisabledError
@@ -116,7 +117,7 @@ class MessageFeedbackApi(InstalledAppResource):
                app_model=app_model,
                message_id=message_id,
                user=current_user,
-                rating=payload.rating,
+                rating=FeedbackRating(payload.rating) if payload.rating else None,
                content=payload.content,
            )
        except MessageNotExistsError:
@@ -15,6 +15,7 @@ from core.app.entities.app_invoke_entities import InvokeFrom
 from fields.conversation_fields import ResultResponse
 from fields.message_fields import MessageInfiniteScrollPagination, MessageListItem
 from libs.helper import UUIDStrOrEmpty
+from models.enums import FeedbackRating
 from models.model import App, AppMode, EndUser
 from services.errors.message import (
    FirstMessageNotExistsError,
@@ -116,7 +117,7 @@ class MessageFeedbackApi(Resource):
                app_model=app_model,
                message_id=message_id,
                user=end_user,
-                rating=payload.rating,
+                rating=FeedbackRating(payload.rating) if payload.rating else None,
                content=payload.content,
            )
        except MessageNotExistsError:
@@ -36,6 +36,7 @@ from extensions.ext_database import db
 from fields.document_fields import document_fields, document_status_fields
 from libs.login import current_user
 from models.dataset import Dataset, Document, DocumentSegment
+from models.enums import SegmentStatus
 from services.dataset_service import DatasetService, DocumentService
 from services.entities.knowledge_entities.knowledge_entities import (
    KnowledgeConfig,
@@ -622,13 +623,15 @@ class DocumentIndexingStatusApi(DatasetApiResource):
                .where(
                    DocumentSegment.completed_at.isnot(None),
                    DocumentSegment.document_id == str(document.id),
-                    DocumentSegment.status != "re_segment",
+                    DocumentSegment.status != SegmentStatus.RE_SEGMENT,
                )
                .count()
            )
            total_segments = (
                db.session.query(DocumentSegment)
-                .where(DocumentSegment.document_id == str(document.id), DocumentSegment.status != "re_segment")
+                .where(
+                    DocumentSegment.document_id == str(document.id), DocumentSegment.status != SegmentStatus.RE_SEGMENT
+                )
                .count()
            )
            # Create a dictionary with document attributes and additional fields
@@ -52,8 +52,19 @@ def handle_webhook(webhook_id: str):
        if error:
            return jsonify({"error": "Bad Request", "message": error}), 400

+        trigger_call_depth = WebhookService.extract_workflow_call_depth(
+            dict(request.headers),
+            request_method=request.method,
+            request_path=request.path,
+        )
+
        # Process webhook call (send to Celery)
-        WebhookService.trigger_workflow_execution(webhook_trigger, webhook_data, workflow)
+        WebhookService.trigger_workflow_execution(
+            webhook_trigger,
+            webhook_data,
+            workflow,
+            call_depth=trigger_call_depth,
+        )

        # Return configured response
        response_data, status_code = WebhookService.generate_webhook_response(node_config)
@@ -70,7 +81,14 @@ def handle_webhook(webhook_id: str):

@bp.route("/webhook-debug/<string:webhook_id>", methods=["GET", "POST", "PUT", "PATCH", "DELETE", "HEAD", "OPTIONS"])
 def handle_webhook_debug(webhook_id: str):
-    """Handle webhook debug calls without triggering production workflow execution."""
+    """Handle webhook debug calls without triggering production workflow execution.
+
+    The debug webhook endpoint is only for draft inspection flows. It never enqueues
+    Celery work for the published workflow; instead it dispatches an in-memory debug
+    event to an active Variable Inspector listener. Returning a clear error when no
+    listener is registered prevents a misleading 200 response for requests that are
+    effectively dropped.
+    """
    try:
        webhook_trigger, _, node_config, webhook_data, error = _prepare_webhook_execution(webhook_id, is_debug=True)
        if error:
@@ -94,11 +112,32 @@ def handle_webhook_debug(webhook_id: str):
                "method": webhook_data.get("method"),
            },
        )
-        TriggerDebugEventBus.dispatch(
+        dispatch_count = TriggerDebugEventBus.dispatch(
            tenant_id=webhook_trigger.tenant_id,
            event=event,
            pool_key=pool_key,
        )
+        if dispatch_count == 0:
+            logger.warning(
+                "Webhook debug request dropped without an active listener for webhook %s (tenant=%s, app=%s, node=%s)",
+                webhook_trigger.webhook_id,
+                webhook_trigger.tenant_id,
+                webhook_trigger.app_id,
+                webhook_trigger.node_id,
+            )
+            return (
+                jsonify(
+                    {
+                        "error": "No active debug listener",
+                        "message": (
+                            "The webhook debug URL only works while the Variable Inspector is listening. "
+                            "Use the published webhook URL to execute the workflow in Celery."
+                        ),
+                        "execution_url": webhook_trigger.webhook_url,
+                    }
+                ),
+                409,
+            )
        response_data, status_code = WebhookService.generate_webhook_response(node_config)
        return jsonify(response_data), status_code

@@ -25,6 +25,7 @@ from fields.conversation_fields import ResultResponse
 from fields.message_fields import SuggestedQuestionsResponse, WebMessageInfiniteScrollPagination, WebMessageListItem
 from libs import helper
 from libs.helper import uuid_value
+from models.enums import FeedbackRating
 from models.model import AppMode
 from services.app_generate_service import AppGenerateService
 from services.errors.app import MoreLikeThisDisabledError
@@ -157,7 +158,7 @@ class MessageFeedbackApi(WebApiResource):
                app_model=app_model,
                message_id=message_id,
                user=end_user,
-                rating=payload.rating,
+                rating=FeedbackRating(payload.rating) if payload.rating else None,
                content=payload.content,
            )
        except MessageNotExistsError:
@@ -441,7 +441,7 @@ class BaseAgentRunner(AppRunner):
                continue

            result.append(self.organize_agent_user_prompt(message))
-            agent_thoughts: list[MessageAgentThought] = message.agent_thoughts
+            agent_thoughts = message.agent_thoughts
            if agent_thoughts:
                for agent_thought in agent_thoughts:
                    tool_names_raw = agent_thought.tool
@@ -1,13 +1,36 @@
 from collections.abc import Mapping
-from typing import Any
+from typing import Any, TypedDict

 from configs import dify_config
 from constants import DEFAULT_FILE_NUMBER_LIMITS


+class SystemParametersDict(TypedDict):
+    image_file_size_limit: int
+    video_file_size_limit: int
+    audio_file_size_limit: int
+    file_size_limit: int
+    workflow_file_upload_limit: int
+
+
+class AppParametersDict(TypedDict):
+    opening_statement: str | None
+    suggested_questions: list[str]
+    suggested_questions_after_answer: dict[str, Any]
+    speech_to_text: dict[str, Any]
+    text_to_speech: dict[str, Any]
+    retriever_resource: dict[str, Any]
+    annotation_reply: dict[str, Any]
+    more_like_this: dict[str, Any]
+    user_input_form: list[dict[str, Any]]
+    sensitive_word_avoidance: dict[str, Any]
+    file_upload: dict[str, Any]
+    system_parameters: SystemParametersDict
+
+
 def get_parameters_from_feature_dict(
    *, features_dict: Mapping[str, Any], user_input_form: list[dict[str, Any]]
-) -> Mapping[str, Any]:
+) -> AppParametersDict:
    """
    Mapping from feature dict to webapp parameters
    """
@@ -8,6 +8,7 @@ from core.app.app_config.entities import (
    ModelConfig,
 )
 from core.entities.agent_entities import PlanningStrategy
+from core.rag.data_post_processor.data_post_processor import RerankingModelDict, WeightsDict
 from models.model import AppMode, AppModelConfigDict
 from services.dataset_service import DatasetService

@@ -117,8 +118,10 @@ class DatasetConfigManager:
                    score_threshold=float(score_threshold_val)
                    if dataset_configs.get("score_threshold_enabled", False) and score_threshold_val is not None
                    else None,
-                    reranking_model=reranking_model_val if isinstance(reranking_model_val, dict) else None,
-                    weights=weights_val if isinstance(weights_val, dict) else None,
+                    reranking_model=cast(RerankingModelDict, reranking_model_val)
+                    if isinstance(reranking_model_val, dict)
+                    else None,
+                    weights=cast(WeightsDict, weights_val) if isinstance(weights_val, dict) else None,
                    reranking_enabled=bool(dataset_configs.get("reranking_enabled", True)),
                    rerank_mode=dataset_configs.get("reranking_mode", "reranking_model"),
                    metadata_filtering_mode=cast(
@@ -4,6 +4,7 @@ from typing import Any, Literal

 from pydantic import BaseModel, Field

+from core.rag.data_post_processor.data_post_processor import RerankingModelDict, WeightsDict
 from dify_graph.file import FileUploadConfig
 from dify_graph.model_runtime.entities.llm_entities import LLMMode
 from dify_graph.model_runtime.entities.message_entities import PromptMessageRole
@@ -194,8 +195,8 @@ class DatasetRetrieveConfigEntity(BaseModel):
    top_k: int | None = None
    score_threshold: float | None = 0.0
    rerank_mode: str | None = "reranking_model"
-    reranking_model: dict | None = None
-    weights: dict | None = None
+    reranking_model: RerankingModelDict | None = None
+    weights: WeightsDict | None = None
    reranking_enabled: bool | None = True
    metadata_filtering_mode: Literal["disabled", "automatic", "manual"] | None = "disabled"
    metadata_model_config: ModelConfig | None = None
@@ -76,7 +76,7 @@ from dify_graph.system_variable import SystemVariable
 from extensions.ext_database import db
 from libs.datetime_utils import naive_utc_now
 from models import Account, Conversation, EndUser, Message, MessageFile
-from models.enums import CreatorUserRole, MessageStatus
+from models.enums import CreatorUserRole, MessageFileBelongsTo, MessageStatus
 from models.execution_extra_content import HumanInputContent
 from models.workflow import Workflow

@@ -939,7 +939,7 @@ class AdvancedChatAppGenerateTaskPipeline(GraphRuntimeStateSupport):
                type=file["type"],
                transfer_method=file["transfer_method"],
                url=file["remote_url"],
-                belongs_to="assistant",
+                belongs_to=MessageFileBelongsTo.ASSISTANT,
                upload_file_id=file["related_id"],
                created_by_role=CreatorUserRole.ACCOUNT
                if message.invoke_from in {InvokeFrom.EXPLORE, InvokeFrom.DEBUGGER}
@@ -40,7 +40,7 @@ from dify_graph.model_runtime.entities.message_entities import (
 from dify_graph.model_runtime.entities.model_entities import ModelPropertyKey
 from dify_graph.model_runtime.errors.invoke import InvokeBadRequestError
 from extensions.ext_database import db
-from models.enums import CreatorUserRole
+from models.enums import CreatorUserRole, MessageFileBelongsTo
 from models.model import App, AppMode, Message, MessageAnnotation, MessageFile

 if TYPE_CHECKING:
@@ -419,7 +419,7 @@ class AppRunner:
            message_id=message_id,
            type=FileType.IMAGE,
            transfer_method=FileTransferMethod.TOOL_FILE,
-            belongs_to="assistant",
+            belongs_to=MessageFileBelongsTo.ASSISTANT,
            url=f"/files/tools/{tool_file.id}",
            upload_file_id=tool_file.id,
            created_by_role=(
@@ -3,7 +3,7 @@ import time
 from collections.abc import Mapping, Sequence
 from dataclasses import dataclass
 from datetime import datetime
-from typing import Any, NewType, Union
+from typing import Any, NewType, TypedDict, Union

 from sqlalchemy import select
 from sqlalchemy.orm import Session
@@ -76,6 +76,20 @@ NodeExecutionId = NewType("NodeExecutionId", str)
 logger = logging.getLogger(__name__)


+class AccountCreatedByDict(TypedDict):
+    id: str
+    name: str
+    email: str
+
+
+class EndUserCreatedByDict(TypedDict):
+    id: str
+    user: str
+
+
+CreatedByDict = AccountCreatedByDict | EndUserCreatedByDict
+
+
@dataclass(slots=True)
 class _NodeSnapshot:
    """In-memory cache for node metadata between start and completion events."""
@@ -249,19 +263,19 @@ class WorkflowResponseConverter:
        outputs_mapping = graph_runtime_state.outputs or {}
        encoded_outputs = WorkflowRuntimeTypeConverter().to_json_encodable(outputs_mapping)

-        created_by: Mapping[str, object] | None
+        created_by: CreatedByDict | dict[str, object] = {}
        user = self._user
        if isinstance(user, Account):
-            created_by = {
-                "id": user.id,
-                "name": user.name,
-                "email": user.email,
-            }
-        else:
-            created_by = {
-                "id": user.id,
-                "user": user.session_id,
-            }
+            created_by = AccountCreatedByDict(
+                id=user.id,
+                name=user.name,
+                email=user.email,
+            )
+        elif isinstance(user, EndUser):
+            created_by = EndUserCreatedByDict(
+                id=user.id,
+                user=user.session_id,
+            )

        return WorkflowFinishStreamResponse(
            task_id=task_id,
@@ -503,7 +517,7 @@ class WorkflowResponseConverter:
        snapshot = self._pop_snapshot(event.node_execution_id)

        start_at = snapshot.start_at if snapshot else event.start_at
-        finished_at = naive_utc_now()
+        finished_at = event.finished_at or naive_utc_now()
        elapsed_time = (finished_at - start_at).total_seconds()

        inputs, inputs_truncated = self._truncate_mapping(event.inputs)
@@ -33,7 +33,7 @@ from extensions.ext_redis import get_pubsub_broadcast_channel
 from libs.broadcast_channel.channel import Topic
 from libs.datetime_utils import naive_utc_now
 from models import Account
-from models.enums import CreatorUserRole
+from models.enums import CreatorUserRole, MessageFileBelongsTo
 from models.model import App, AppMode, AppModelConfig, Conversation, EndUser, Message, MessageFile
 from services.errors.app_model_config import AppModelConfigBrokenError
 from services.errors.conversation import ConversationNotExistsError
@@ -225,7 +225,7 @@ class MessageBasedAppGenerator(BaseAppGenerator):
                    message_id=message.id,
                    type=file.type,
                    transfer_method=file.transfer_method,
-                    belongs_to="user",
+                    belongs_to=MessageFileBelongsTo.USER,
                    url=file.remote_url,
                    upload_file_id=file.related_id,
                    created_by_role=(CreatorUserRole.ACCOUNT if account_id else CreatorUserRole.END_USER),
@@ -84,6 +84,7 @@ class WorkflowAppRunner(WorkflowBasedAppRunner):
                user_id=self.application_generate_entity.user_id,
                user_from=user_from,
                invoke_from=invoke_from,
+                call_depth=self.application_generate_entity.call_depth,
                root_node_id=self._root_node_id,
            )
        elif self.application_generate_entity.single_iteration_run or self.application_generate_entity.single_loop_run:
@@ -91,6 +92,7 @@ class WorkflowAppRunner(WorkflowBasedAppRunner):
                workflow=self._workflow,
                single_iteration_run=self.application_generate_entity.single_iteration_run,
                single_loop_run=self.application_generate_entity.single_loop_run,
+                call_depth=self.application_generate_entity.call_depth,
            )
        else:
            inputs = self.application_generate_entity.inputs
@@ -120,6 +122,7 @@ class WorkflowAppRunner(WorkflowBasedAppRunner):
                user_id=self.application_generate_entity.user_id,
                user_from=user_from,
                invoke_from=invoke_from,
+                call_depth=self.application_generate_entity.call_depth,
                root_node_id=self._root_node_id,
            )

@@ -102,6 +102,7 @@ class WorkflowBasedAppRunner:
        graph_runtime_state: GraphRuntimeState,
        user_from: UserFrom,
        invoke_from: InvokeFrom,
+        call_depth: int = 0,
        workflow_id: str = "",
        tenant_id: str = "",
        user_id: str = "",
@@ -130,7 +131,7 @@ class WorkflowBasedAppRunner:
                user_from=user_from,
                invoke_from=invoke_from,
            ),
-            call_depth=0,
+            call_depth=call_depth,
        )

        # Use the provided graph_runtime_state for consistent state management
@@ -156,6 +157,7 @@ class WorkflowBasedAppRunner:
        workflow: Workflow,
        single_iteration_run: Any | None = None,
        single_loop_run: Any | None = None,
+        call_depth: int = 0,
    ) -> tuple[Graph, VariablePool, GraphRuntimeState]:
        """
        Prepare graph, variable pool, and runtime state for single node execution
@@ -189,6 +191,7 @@ class WorkflowBasedAppRunner:
                node_id=single_iteration_run.node_id,
                user_inputs=dict(single_iteration_run.inputs),
                graph_runtime_state=graph_runtime_state,
+                call_depth=call_depth,
                node_type_filter_key="iteration_id",
                node_type_label="iteration",
            )
@@ -198,6 +201,7 @@ class WorkflowBasedAppRunner:
                node_id=single_loop_run.node_id,
                user_inputs=dict(single_loop_run.inputs),
                graph_runtime_state=graph_runtime_state,
+                call_depth=call_depth,
                node_type_filter_key="loop_id",
                node_type_label="loop",
            )
@@ -214,6 +218,7 @@ class WorkflowBasedAppRunner:
        node_id: str,
        user_inputs: dict[str, Any],
        graph_runtime_state: GraphRuntimeState,
+        call_depth: int,
        node_type_filter_key: str,  # 'iteration_id' or 'loop_id'
        node_type_label: str = "node",  # 'iteration' or 'loop' for error messages
    ) -> tuple[Graph, VariablePool]:
@@ -283,7 +288,7 @@ class WorkflowBasedAppRunner:
                user_from=UserFrom.ACCOUNT,
                invoke_from=InvokeFrom.DEBUGGER,
            ),
-            call_depth=0,
+            call_depth=call_depth,
        )

        node_factory = DifyNodeFactory(
@@ -456,6 +461,7 @@ class WorkflowBasedAppRunner:
                    node_id=event.node_id,
                    node_type=event.node_type,
                    start_at=event.start_at,
+                    finished_at=event.finished_at,
                    inputs=inputs,
                    process_data=process_data,
                    outputs=outputs,
@@ -471,6 +477,7 @@ class WorkflowBasedAppRunner:
                    node_id=event.node_id,
                    node_type=event.node_type,
                    start_at=event.start_at,
+                    finished_at=event.finished_at,
                    inputs=event.node_run_result.inputs,
                    process_data=event.node_run_result.process_data,
                    outputs=event.node_run_result.outputs,
@@ -487,6 +494,7 @@ class WorkflowBasedAppRunner:
                    node_id=event.node_id,
                    node_type=event.node_type,
                    start_at=event.start_at,
+                    finished_at=event.finished_at,
                    inputs=event.node_run_result.inputs,
                    process_data=event.node_run_result.process_data,
                    outputs=event.node_run_result.outputs,
@@ -335,6 +335,7 @@ class QueueNodeSucceededEvent(AppQueueEvent):
    in_loop_id: str | None = None
    """loop id if node is in loop"""
    start_at: datetime
+    finished_at: datetime | None = None

    inputs: Mapping[str, object] = Field(default_factory=dict)
    process_data: Mapping[str, object] = Field(default_factory=dict)
@@ -390,6 +391,7 @@ class QueueNodeExceptionEvent(AppQueueEvent):
    in_loop_id: str | None = None
    """loop id if node is in loop"""
    start_at: datetime
+    finished_at: datetime | None = None

    inputs: Mapping[str, object] = Field(default_factory=dict)
    process_data: Mapping[str, object] = Field(default_factory=dict)
@@ -414,6 +416,7 @@ class QueueNodeFailedEvent(AppQueueEvent):
    in_loop_id: str | None = None
    """loop id if node is in loop"""
    start_at: datetime
+    finished_at: datetime | None = None

    inputs: Mapping[str, object] = Field(default_factory=dict)
    process_data: Mapping[str, object] = Field(default_factory=dict)
@@ -6,6 +6,7 @@ from core.app.entities.app_invoke_entities import InvokeFrom
 from core.rag.datasource.vdb.vector_factory import Vector
 from extensions.ext_database import db
 from models.dataset import Dataset
+from models.enums import CollectionBindingType
 from models.model import App, AppAnnotationSetting, Message, MessageAnnotation
 from services.annotation_service import AppAnnotationService
 from services.dataset_service import DatasetCollectionBindingService
@@ -43,7 +44,7 @@ class AnnotationReplyFeature:
            embedding_model_name = collection_binding_detail.model_name

            dataset_collection_binding = DatasetCollectionBindingService.get_dataset_collection_binding(
-                embedding_provider_name, embedding_model_name, "annotation"
+                embedding_provider_name, embedding_model_name, CollectionBindingType.ANNOTATION
            )

            dataset = Dataset(
@@ -34,6 +34,7 @@ from core.llm_generator.llm_generator import LLMGenerator
 from core.tools.signature import sign_tool_file
 from extensions.ext_database import db
 from extensions.ext_redis import redis_client
+from models.enums import MessageFileBelongsTo
 from models.model import AppMode, Conversation, MessageAnnotation, MessageFile
 from services.annotation_service import AppAnnotationService

@@ -233,7 +234,7 @@ class MessageCycleManager:
                task_id=self._application_generate_entity.task_id,
                id=message_file.id,
                type=message_file.type,
-                belongs_to=message_file.belongs_to or "user",
+                belongs_to=message_file.belongs_to or MessageFileBelongsTo.USER,
                url=url,
            )

@@ -1,3 +1,5 @@
+from typing import TypedDict
+
 from core.tools.signature import sign_tool_file
 from dify_graph.file import helpers as file_helpers
 from dify_graph.file.enums import FileTransferMethod
@@ -6,7 +8,20 @@ from models.model import MessageFile, UploadFile
 MAX_TOOL_FILE_EXTENSION_LENGTH = 10


-def prepare_file_dict(message_file: MessageFile, upload_files_map: dict[str, UploadFile]) -> dict:
+class MessageFileInfoDict(TypedDict):
+    related_id: str
+    extension: str
+    filename: str
+    size: int
+    mime_type: str
+    transfer_method: str
+    type: str
+    url: str
+    upload_file_id: str
+    remote_url: str | None
+
+
+def prepare_file_dict(message_file: MessageFile, upload_files_map: dict[str, UploadFile]) -> MessageFileInfoDict:
    """
    Prepare file dictionary for message end stream response.

@@ -268,7 +268,12 @@ class WorkflowPersistenceLayer(GraphEngineLayer):

    def _handle_node_succeeded(self, event: NodeRunSucceededEvent) -> None:
        domain_execution = self._get_node_execution(event.id)
-        self._update_node_execution(domain_execution, event.node_run_result, WorkflowNodeExecutionStatus.SUCCEEDED)
+        self._update_node_execution(
+            domain_execution,
+            event.node_run_result,
+            WorkflowNodeExecutionStatus.SUCCEEDED,
+            finished_at=event.finished_at,
+        )

    def _handle_node_failed(self, event: NodeRunFailedEvent) -> None:
        domain_execution = self._get_node_execution(event.id)
@@ -277,6 +282,7 @@ class WorkflowPersistenceLayer(GraphEngineLayer):
            event.node_run_result,
            WorkflowNodeExecutionStatus.FAILED,
            error=event.error,
+            finished_at=event.finished_at,
        )

    def _handle_node_exception(self, event: NodeRunExceptionEvent) -> None:
@@ -286,6 +292,7 @@ class WorkflowPersistenceLayer(GraphEngineLayer):
            event.node_run_result,
            WorkflowNodeExecutionStatus.EXCEPTION,
            error=event.error,
+            finished_at=event.finished_at,
        )

    def _handle_node_pause_requested(self, event: NodeRunPauseRequestedEvent) -> None:
@@ -352,13 +359,14 @@ class WorkflowPersistenceLayer(GraphEngineLayer):
        *,
        error: str | None = None,
        update_outputs: bool = True,
+        finished_at: datetime | None = None,
    ) -> None:
-        finished_at = naive_utc_now()
+        actual_finished_at = finished_at or naive_utc_now()
        snapshot = self._node_snapshots.get(domain_execution.id)
        start_at = snapshot.created_at if snapshot else domain_execution.created_at
        domain_execution.status = status
-        domain_execution.finished_at = finished_at
-        domain_execution.elapsed_time = max((finished_at - start_at).total_seconds(), 0.0)
+        domain_execution.finished_at = actual_finished_at
+        domain_execution.elapsed_time = max((actual_finished_at - start_at).total_seconds(), 0.0)

        if error:
            domain_execution.error = error
@@ -12,7 +12,7 @@ from core.rag.models.document import Document
 from extensions.ext_database import db
 from models.dataset import ChildChunk, DatasetQuery, DocumentSegment
 from models.dataset import Document as DatasetDocument
-from models.enums import CreatorUserRole
+from models.enums import CreatorUserRole, DatasetQuerySource

 _logger = logging.getLogger(__name__)

@@ -36,7 +36,7 @@ class DatasetIndexToolCallbackHandler:
        dataset_query = DatasetQuery(
            dataset_id=dataset_id,
            content=query,
-            source="app",
+            source=DatasetQuerySource.APP,
            source_app_id=self._app_id,
            created_by_role=(
                CreatorUserRole.ACCOUNT
@@ -15,6 +15,7 @@ from configs import dify_config
 from core.helper import ssrf_proxy
 from extensions.ext_database import db
 from extensions.ext_storage import storage
+from extensions.storage.storage_type import StorageType
 from models.enums import CreatorUserRole
 from models.model import MessageFile, UploadFile
 from models.tools import ToolFile
@@ -81,7 +82,7 @@ class DatasourceFileManager:

        upload_file = UploadFile(
            tenant_id=tenant_id,
-            storage_type=dify_config.STORAGE_TYPE,
+            storage_type=StorageType(dify_config.STORAGE_TYPE),
            key=filepath,
            name=present_filename,
            size=len(file_binary),
@@ -30,6 +30,7 @@ from dify_graph.model_runtime.model_providers.__base.ai_model import AIModel
 from dify_graph.model_runtime.model_providers.model_provider_factory import ModelProviderFactory
 from libs.datetime_utils import naive_utc_now
 from models.engine import db
+from models.enums import CredentialSourceType
 from models.provider import (
    LoadBalancingModelConfig,
    Provider,
@@ -546,7 +547,7 @@ class ProviderConfiguration(BaseModel):
                self._update_load_balancing_configs_with_credential(
                    credential_id=credential_id,
                    credential_record=credential_record,
-                    credential_source="provider",
+                    credential_source=CredentialSourceType.PROVIDER,
                    session=session,
                )
            except Exception:
@@ -623,7 +624,7 @@ class ProviderConfiguration(BaseModel):
                LoadBalancingModelConfig.tenant_id == self.tenant_id,
                LoadBalancingModelConfig.provider_name.in_(self._get_provider_names()),
                LoadBalancingModelConfig.credential_id == credential_id,
-                LoadBalancingModelConfig.credential_source_type == "provider",
+                LoadBalancingModelConfig.credential_source_type == CredentialSourceType.PROVIDER,
            )
            lb_configs_using_credential = session.execute(lb_stmt).scalars().all()
            try:
@@ -1043,7 +1044,7 @@ class ProviderConfiguration(BaseModel):
                self._update_load_balancing_configs_with_credential(
                    credential_id=credential_id,
                    credential_record=credential_record,
-                    credential_source="custom_model",
+                    credential_source=CredentialSourceType.CUSTOM_MODEL,
                    session=session,
                )
            except Exception:
@@ -1073,7 +1074,7 @@ class ProviderConfiguration(BaseModel):
                LoadBalancingModelConfig.tenant_id == self.tenant_id,
                LoadBalancingModelConfig.provider_name.in_(self._get_provider_names()),
                LoadBalancingModelConfig.credential_id == credential_id,
-                LoadBalancingModelConfig.credential_source_type == "custom_model",
+                LoadBalancingModelConfig.credential_source_type == CredentialSourceType.CUSTOM_MODEL,
            )
            lb_configs_using_credential = session.execute(lb_stmt).scalars().all()

@@ -1421,12 +1422,12 @@ class ProviderConfiguration(BaseModel):
            preferred_model_provider = s.execute(stmt).scalars().first()

            if preferred_model_provider:
-                preferred_model_provider.preferred_provider_type = provider_type.value
+                preferred_model_provider.preferred_provider_type = provider_type
            else:
                preferred_model_provider = TenantPreferredModelProvider(
                    tenant_id=self.tenant_id,
                    provider_name=self.provider.provider,
-                    preferred_provider_type=provider_type.value,
+                    preferred_provider_type=provider_type,
                )
                s.add(preferred_model_provider)
            s.commit()
@@ -1711,7 +1712,7 @@ class ProviderConfiguration(BaseModel):
                    provider_model_lb_configs = [
                        config
                        for config in model_setting.load_balancing_configs
-                        if config.credential_source_type != "custom_model"
+                        if config.credential_source_type != CredentialSourceType.CUSTOM_MODEL
                    ]

                    load_balancing_enabled = model_setting.load_balancing_enabled
@@ -1769,7 +1770,7 @@ class ProviderConfiguration(BaseModel):
                custom_model_lb_configs = [
                    config
                    for config in model_setting.load_balancing_configs
-                    if config.credential_source_type != "provider"
+                    if config.credential_source_type != CredentialSourceType.PROVIDER
                ]

                load_balancing_enabled = model_setting.load_balancing_enabled
@@ -40,6 +40,7 @@ from libs.datetime_utils import naive_utc_now
 from models import Account
 from models.dataset import AutomaticRulesConfig, ChildChunk, Dataset, DatasetProcessRule, DocumentSegment
 from models.dataset import Document as DatasetDocument
+from models.enums import DataSourceType, IndexingStatus, ProcessRuleMode, SegmentStatus
 from models.model import UploadFile
 from services.feature_service import FeatureService

@@ -56,7 +57,7 @@ class IndexingRunner:
        logger.exception("consume document failed")
        document = db.session.get(DatasetDocument, document_id)
        if document:
-            document.indexing_status = "error"
+            document.indexing_status = IndexingStatus.ERROR
            error_message = getattr(error, "description", str(error))
            document.error = str(error_message)
            document.stopped_at = naive_utc_now()
@@ -219,7 +220,7 @@ class IndexingRunner:
            if document_segments:
                for document_segment in document_segments:
                    # transform segment to node
-                    if document_segment.status != "completed":
+                    if document_segment.status != SegmentStatus.COMPLETED:
                        document = Document(
                            page_content=document_segment.content,
                            metadata={
@@ -382,7 +383,7 @@ class IndexingRunner:
        data_source_info = dataset_document.data_source_info_dict
        text_docs = []
        match dataset_document.data_source_type:
-            case "upload_file":
+            case DataSourceType.UPLOAD_FILE:
                if not data_source_info or "upload_file_id" not in data_source_info:
                    raise ValueError("no upload file found")
                stmt = select(UploadFile).where(UploadFile.id == data_source_info["upload_file_id"])
@@ -395,7 +396,7 @@ class IndexingRunner:
                        document_model=dataset_document.doc_form,
                    )
                    text_docs = index_processor.extract(extract_setting, process_rule_mode=process_rule["mode"])
-            case "notion_import":
+            case DataSourceType.NOTION_IMPORT:
                if (
                    not data_source_info
                    or "notion_workspace_id" not in data_source_info
@@ -417,7 +418,7 @@ class IndexingRunner:
                    document_model=dataset_document.doc_form,
                )
                text_docs = index_processor.extract(extract_setting, process_rule_mode=process_rule["mode"])
-            case "website_crawl":
+            case DataSourceType.WEBSITE_CRAWL:
                if (
                    not data_source_info
                    or "provider" not in data_source_info
@@ -445,7 +446,7 @@ class IndexingRunner:
        # update document status to splitting
        self._update_document_index_status(
            document_id=dataset_document.id,
-            after_indexing_status="splitting",
+            after_indexing_status=IndexingStatus.SPLITTING,
            extra_update_params={
                DatasetDocument.parsing_completed_at: naive_utc_now(),
            },
@@ -545,7 +546,7 @@ class IndexingRunner:
        Clean the document text according to the processing rules.
        """
        rules: AutomaticRulesConfig | dict[str, Any]
-        if processing_rule.mode == "automatic":
+        if processing_rule.mode == ProcessRuleMode.AUTOMATIC:
            rules = DatasetProcessRule.AUTOMATIC_RULES
        else:
            rules = json.loads(processing_rule.rules) if processing_rule.rules else {}
@@ -636,7 +637,7 @@ class IndexingRunner:
        # update document status to completed
        self._update_document_index_status(
            document_id=dataset_document.id,
-            after_indexing_status="completed",
+            after_indexing_status=IndexingStatus.COMPLETED,
            extra_update_params={
                DatasetDocument.tokens: tokens,
                DatasetDocument.completed_at: naive_utc_now(),
@@ -659,10 +660,10 @@ class IndexingRunner:
                    DocumentSegment.document_id == document_id,
                    DocumentSegment.dataset_id == dataset_id,
                    DocumentSegment.index_node_id.in_(document_ids),
-                    DocumentSegment.status == "indexing",
+                    DocumentSegment.status == SegmentStatus.INDEXING,
                ).update(
                    {
-                        DocumentSegment.status: "completed",
+                        DocumentSegment.status: SegmentStatus.COMPLETED,
                        DocumentSegment.enabled: True,
                        DocumentSegment.completed_at: naive_utc_now(),
                    }
@@ -703,10 +704,10 @@ class IndexingRunner:
                DocumentSegment.document_id == dataset_document.id,
                DocumentSegment.dataset_id == dataset.id,
                DocumentSegment.index_node_id.in_(document_ids),
-                DocumentSegment.status == "indexing",
+                DocumentSegment.status == SegmentStatus.INDEXING,
            ).update(
                {
-                    DocumentSegment.status: "completed",
+                    DocumentSegment.status: SegmentStatus.COMPLETED,
                    DocumentSegment.enabled: True,
                    DocumentSegment.completed_at: naive_utc_now(),
                }
@@ -725,7 +726,7 @@ class IndexingRunner:

    @staticmethod
    def _update_document_index_status(
-        document_id: str, after_indexing_status: str, extra_update_params: dict | None = None
+        document_id: str, after_indexing_status: IndexingStatus, extra_update_params: dict | None = None
    ):
        """
        Update the document indexing status.
@@ -803,7 +804,7 @@ class IndexingRunner:
        cur_time = naive_utc_now()
        self._update_document_index_status(
            document_id=dataset_document.id,
-            after_indexing_status="indexing",
+            after_indexing_status=IndexingStatus.INDEXING,
            extra_update_params={
                DatasetDocument.cleaning_completed_at: cur_time,
                DatasetDocument.splitting_completed_at: cur_time,
@@ -815,7 +816,7 @@ class IndexingRunner:
        self._update_segments_by_document(
            dataset_document_id=dataset_document.id,
            update_params={
-                DocumentSegment.status: "indexing",
+                DocumentSegment.status: SegmentStatus.INDEXING,
                DocumentSegment.indexing_at: naive_utc_now(),
            },
        )
@@ -195,7 +195,7 @@ class ProviderManager:
            preferred_provider_type_record = provider_name_to_preferred_model_provider_records_dict.get(provider_name)

            if preferred_provider_type_record:
-                preferred_provider_type = ProviderType.value_of(preferred_provider_type_record.preferred_provider_type)
+                preferred_provider_type = preferred_provider_type_record.preferred_provider_type
            elif dify_config.EDITION == "CLOUD" and system_configuration.enabled:
                preferred_provider_type = ProviderType.SYSTEM
            elif custom_configuration.provider or custom_configuration.models:
@@ -1,3 +1,5 @@
+from typing_extensions import TypedDict
+
 from core.model_manager import ModelInstance, ModelManager
 from core.rag.data_post_processor.reorder import ReorderRunner
 from core.rag.index_processor.constant.query_type import QueryType
@@ -10,6 +12,26 @@ from dify_graph.model_runtime.entities.model_entities import ModelType
 from dify_graph.model_runtime.errors.invoke import InvokeAuthorizationError


+class RerankingModelDict(TypedDict):
+    reranking_provider_name: str
+    reranking_model_name: str
+
+
+class VectorSettingDict(TypedDict):
+    vector_weight: float
+    embedding_provider_name: str
+    embedding_model_name: str
+
+
+class KeywordSettingDict(TypedDict):
+    keyword_weight: float
+
+
+class WeightsDict(TypedDict):
+    vector_setting: VectorSettingDict
+    keyword_setting: KeywordSettingDict
+
+
 class DataPostProcessor:
    """Interface for data post-processing document."""

@@ -17,8 +39,8 @@ class DataPostProcessor:
        self,
        tenant_id: str,
        reranking_mode: str,
-        reranking_model: dict | None = None,
-        weights: dict | None = None,
+        reranking_model: RerankingModelDict | None = None,
+        weights: WeightsDict | None = None,
        reorder_enabled: bool = False,
    ):
        self.rerank_runner = self._get_rerank_runner(reranking_mode, tenant_id, reranking_model, weights)
@@ -45,8 +67,8 @@ class DataPostProcessor:
        self,
        reranking_mode: str,
        tenant_id: str,
-        reranking_model: dict | None = None,
-        weights: dict | None = None,
+        reranking_model: RerankingModelDict | None = None,
+        weights: WeightsDict | None = None,
    ) -> BaseRerankRunner | None:
        if reranking_mode == RerankMode.WEIGHTED_SCORE and weights:
            runner = RerankRunnerFactory.create_rerank_runner(
@@ -79,12 +101,14 @@ class DataPostProcessor:
            return ReorderRunner()
        return None

-    def _get_rerank_model_instance(self, tenant_id: str, reranking_model: dict | None) -> ModelInstance | None:
+    def _get_rerank_model_instance(
+        self, tenant_id: str, reranking_model: RerankingModelDict | None
+    ) -> ModelInstance | None:
        if reranking_model:
            try:
                model_manager = ModelManager()
-                reranking_provider_name = reranking_model.get("reranking_provider_name")
-                reranking_model_name = reranking_model.get("reranking_model_name")
+                reranking_provider_name = reranking_model["reranking_provider_name"]
+                reranking_model_name = reranking_model["reranking_model_name"]
                if not reranking_provider_name or not reranking_model_name:
                    return None
                rerank_model_instance = model_manager.get_model_instance(
@@ -1,19 +1,20 @@
 import concurrent.futures
 import logging
 from concurrent.futures import ThreadPoolExecutor
-from typing import Any
+from typing import Any, NotRequired

 from flask import Flask, current_app
 from sqlalchemy import select
 from sqlalchemy.orm import Session, load_only
+from typing_extensions import TypedDict

 from configs import dify_config
 from core.db.session_factory import session_factory
 from core.model_manager import ModelManager
-from core.rag.data_post_processor.data_post_processor import DataPostProcessor
+from core.rag.data_post_processor.data_post_processor import DataPostProcessor, RerankingModelDict, WeightsDict
 from core.rag.datasource.keyword.keyword_factory import Keyword
 from core.rag.datasource.vdb.vector_factory import Vector
-from core.rag.embedding.retrieval import RetrievalChildChunk, RetrievalSegments
+from core.rag.embedding.retrieval import AttachmentInfoDict, RetrievalChildChunk, RetrievalSegments
 from core.rag.entities.metadata_entities import MetadataCondition
 from core.rag.index_processor.constant.doc_type import DocType
 from core.rag.index_processor.constant.index_type import IndexStructureType
@@ -35,7 +36,49 @@ from models.dataset import Document as DatasetDocument
 from models.model import UploadFile
 from services.external_knowledge_service import ExternalDatasetService

-default_retrieval_model = {
+
+class SegmentAttachmentResult(TypedDict):
+    attachment_info: AttachmentInfoDict
+    segment_id: str
+
+
+class SegmentAttachmentInfoResult(TypedDict):
+    attachment_id: str
+    attachment_info: AttachmentInfoDict
+    segment_id: str
+
+
+class ChildChunkDetail(TypedDict):
+    id: str
+    content: str
+    position: int
+    score: float
+
+
+class SegmentChildMapDetail(TypedDict):
+    max_score: float
+    child_chunks: list[ChildChunkDetail]
+
+
+class SegmentRecord(TypedDict):
+    segment: DocumentSegment
+    score: NotRequired[float]
+    child_chunks: NotRequired[list[ChildChunkDetail]]
+    files: NotRequired[list[AttachmentInfoDict]]
+
+
+class DefaultRetrievalModelDict(TypedDict):
+    search_method: RetrievalMethod
+    reranking_enable: bool
+    reranking_model: RerankingModelDict
+    reranking_mode: NotRequired[str]
+    weights: NotRequired[WeightsDict | None]
+    score_threshold: NotRequired[float]
+    top_k: int
+    score_threshold_enabled: bool
+
+
+default_retrieval_model: DefaultRetrievalModelDict = {
    "search_method": RetrievalMethod.SEMANTIC_SEARCH,
    "reranking_enable": False,
    "reranking_model": {"reranking_provider_name": "", "reranking_model_name": ""},
@@ -56,9 +99,9 @@ class RetrievalService:
        query: str,
        top_k: int = 4,
        score_threshold: float | None = 0.0,
-        reranking_model: dict | None = None,
+        reranking_model: RerankingModelDict | None = None,
        reranking_mode: str = "reranking_model",
-        weights: dict | None = None,
+        weights: WeightsDict | None = None,
        document_ids_filter: list[str] | None = None,
        attachment_ids: list | None = None,
    ):
@@ -235,7 +278,7 @@ class RetrievalService:
        query: str,
        top_k: int,
        score_threshold: float | None,
-        reranking_model: dict | None,
+        reranking_model: RerankingModelDict | None,
        all_documents: list,
        retrieval_method: RetrievalMethod,
        exceptions: list,
@@ -277,8 +320,8 @@ class RetrievalService:
                if documents:
                    if (
                        reranking_model
-                        and reranking_model.get("reranking_model_name")
-                        and reranking_model.get("reranking_provider_name")
+                        and reranking_model["reranking_model_name"]
+                        and reranking_model["reranking_provider_name"]
                        and retrieval_method == RetrievalMethod.SEMANTIC_SEARCH
                    ):
                        data_post_processor = DataPostProcessor(
@@ -288,8 +331,8 @@ class RetrievalService:
                            model_manager = ModelManager()
                            is_support_vision = model_manager.check_model_support_vision(
                                tenant_id=dataset.tenant_id,
-                                provider=reranking_model.get("reranking_provider_name") or "",
-                                model=reranking_model.get("reranking_model_name") or "",
+                                provider=reranking_model["reranking_provider_name"],
+                                model=reranking_model["reranking_model_name"],
                                model_type=ModelType.RERANK,
                            )
                            if is_support_vision:
@@ -329,7 +372,7 @@ class RetrievalService:
        query: str,
        top_k: int,
        score_threshold: float | None,
-        reranking_model: dict | None,
+        reranking_model: RerankingModelDict | None,
        all_documents: list,
        retrieval_method: str,
        exceptions: list,
@@ -349,8 +392,8 @@ class RetrievalService:
                if documents:
                    if (
                        reranking_model
-                        and reranking_model.get("reranking_model_name")
-                        and reranking_model.get("reranking_provider_name")
+                        and reranking_model["reranking_model_name"]
+                        and reranking_model["reranking_provider_name"]
                        and retrieval_method == RetrievalMethod.FULL_TEXT_SEARCH
                    ):
                        data_post_processor = DataPostProcessor(
@@ -459,7 +502,7 @@ class RetrievalService:
            segment_ids: list[str] = []
            index_node_segments: list[DocumentSegment] = []
            segments: list[DocumentSegment] = []
-            attachment_map: dict[str, list[dict[str, Any]]] = {}
+            attachment_map: dict[str, list[AttachmentInfoDict]] = {}
            child_chunk_map: dict[str, list[ChildChunk]] = {}
            doc_segment_map: dict[str, list[str]] = {}
            segment_summary_map: dict[str, str] = {}  # Map segment_id to summary content
@@ -544,12 +587,12 @@ class RetrievalService:
                            segment_summary_map[summary.chunk_id] = summary.summary_content

            include_segment_ids = set()
-            segment_child_map: dict[str, dict[str, Any]] = {}
-            records: list[dict[str, Any]] = []
+            segment_child_map: dict[str, SegmentChildMapDetail] = {}
+            records: list[SegmentRecord] = []

            for segment in segments:
                child_chunks: list[ChildChunk] = child_chunk_map.get(segment.id, [])
-                attachment_infos: list[dict[str, Any]] = attachment_map.get(segment.id, [])
+                attachment_infos: list[AttachmentInfoDict] = attachment_map.get(segment.id, [])
                ds_dataset_document: DatasetDocument | None = valid_dataset_documents.get(segment.document_id)

                if ds_dataset_document and ds_dataset_document.doc_form == IndexStructureType.PARENT_CHILD_INDEX:
@@ -560,14 +603,14 @@ class RetrievalService:
                        max_score = summary_score_map.get(segment.id, 0.0)

                        if child_chunks or attachment_infos:
-                            child_chunk_details = []
+                            child_chunk_details: list[ChildChunkDetail] = []
                            for child_chunk in child_chunks:
                                child_document: Document | None = doc_to_document_map.get(child_chunk.index_node_id)
                                if child_document:
                                    child_score = child_document.metadata.get("score", 0.0)
                                else:
                                    child_score = 0.0
-                                child_chunk_detail = {
+                                child_chunk_detail: ChildChunkDetail = {
                                    "id": child_chunk.id,
                                    "content": child_chunk.content,
                                    "position": child_chunk.position,
@@ -580,7 +623,7 @@ class RetrievalService:
                                if file_document:
                                    max_score = max(max_score, file_document.metadata.get("score", 0.0))

-                            map_detail = {
+                            map_detail: SegmentChildMapDetail = {
                                "max_score": max_score,
                                "child_chunks": child_chunk_details,
                            }
@@ -593,7 +636,7 @@ class RetrievalService:
                                    "max_score": summary_score,
                                    "child_chunks": [],
                                }
-                        record: dict[str, Any] = {
+                        record: SegmentRecord = {
                            "segment": segment,
                        }
                        records.append(record)
@@ -617,19 +660,19 @@ class RetrievalService:
                            if file_doc:
                                max_score = max(max_score, file_doc.metadata.get("score", 0.0))

-                        record = {
+                        another_record: SegmentRecord = {
                            "segment": segment,
                            "score": max_score,
                        }
-                        records.append(record)
+                        records.append(another_record)

            # Add child chunks information to records
            for record in records:
                if record["segment"].id in segment_child_map:
-                    record["child_chunks"] = segment_child_map[record["segment"].id].get("child_chunks")  # type: ignore
-                    record["score"] = segment_child_map[record["segment"].id]["max_score"]  # type: ignore
+                    record["child_chunks"] = segment_child_map[record["segment"].id]["child_chunks"]
+                    record["score"] = segment_child_map[record["segment"].id]["max_score"]
                if record["segment"].id in attachment_map:
-                    record["files"] = attachment_map[record["segment"].id]  # type: ignore[assignment]
+                    record["files"] = attachment_map[record["segment"].id]

            result: list[RetrievalSegments] = []
            for record in records:
@@ -693,9 +736,9 @@ class RetrievalService:
        query: str | None = None,
        top_k: int = 4,
        score_threshold: float | None = 0.0,
-        reranking_model: dict | None = None,
+        reranking_model: RerankingModelDict | None = None,
        reranking_mode: str = "reranking_model",
-        weights: dict | None = None,
+        weights: WeightsDict | None = None,
        document_ids_filter: list[str] | None = None,
        attachment_id: str | None = None,
    ):
@@ -807,7 +850,7 @@ class RetrievalService:
    @classmethod
    def get_segment_attachment_info(
        cls, dataset_id: str, tenant_id: str, attachment_id: str, session: Session
-    ) -> dict[str, Any] | None:
+    ) -> SegmentAttachmentResult | None:
        upload_file = session.query(UploadFile).where(UploadFile.id == attachment_id).first()
        if upload_file:
            attachment_binding = (
@@ -816,7 +859,7 @@ class RetrievalService:
                .first()
            )
            if attachment_binding:
-                attachment_info = {
+                attachment_info: AttachmentInfoDict = {
                    "id": upload_file.id,
                    "name": upload_file.name,
                    "extension": "." + upload_file.extension,
@@ -828,8 +871,10 @@ class RetrievalService:
        return None

    @classmethod
-    def get_segment_attachment_infos(cls, attachment_ids: list[str], session: Session) -> list[dict[str, Any]]:
-        attachment_infos = []
+    def get_segment_attachment_infos(
+        cls, attachment_ids: list[str], session: Session
+    ) -> list[SegmentAttachmentInfoResult]:
+        attachment_infos: list[SegmentAttachmentInfoResult] = []
        upload_files = session.query(UploadFile).where(UploadFile.id.in_(attachment_ids)).all()
        if upload_files:
            upload_file_ids = [upload_file.id for upload_file in upload_files]
@@ -843,7 +888,7 @@ class RetrievalService:
            if attachment_bindings:
                for upload_file in upload_files:
                    attachment_binding = attachment_binding_map.get(upload_file.id)
-                    attachment_info = {
+                    info: AttachmentInfoDict = {
                        "id": upload_file.id,
                        "name": upload_file.name,
                        "extension": "." + upload_file.extension,
@@ -855,7 +900,7 @@ class RetrievalService:
                        attachment_infos.append(
                            {
                                "attachment_id": attachment_binding.attachment_id,
-                                "attachment_info": attachment_info,
+                                "attachment_info": info,
                                "segment_id": attachment_binding.segment_id,
                            }
                        )
@@ -5,6 +5,7 @@ This module provides integration with Weaviate vector database for storing and r
 document embeddings used in retrieval-augmented generation workflows.
 """

+import atexit
 import datetime
 import json
 import logging
@@ -37,6 +38,32 @@ _weaviate_client: weaviate.WeaviateClient | None = None
 _weaviate_client_lock = threading.Lock()


+def _shutdown_weaviate_client() -> None:
+    """
+    Best-effort shutdown hook to close the module-level Weaviate client.
+
+    This is registered with atexit so that HTTP/gRPC resources are released
+    when the Python interpreter exits.
+    """
+    global _weaviate_client
+
+    # Ensure thread-safety when accessing the shared client instance
+    with _weaviate_client_lock:
+        client = _weaviate_client
+        _weaviate_client = None
+
+    if client is not None:
+        try:
+            client.close()
+        except Exception:
+            # Best-effort cleanup; log at debug level and ignore errors.
+            logger.debug("Failed to close Weaviate client during shutdown", exc_info=True)
+
+
+# Register the shutdown hook once per process.
+atexit.register(_shutdown_weaviate_client)
+
+
 class WeaviateConfig(BaseModel):
    """
    Configuration model for Weaviate connection settings.
@@ -85,18 +112,6 @@ class WeaviateVector(BaseVector):
        self._client = self._init_client(config)
        self._attributes = attributes

-    def __del__(self):
-        """
-        Destructor to properly close the Weaviate client connection.
-        Prevents connection leaks and resource warnings.
-        """
-        if hasattr(self, "_client") and self._client is not None:
-            try:
-                self._client.close()
-            except Exception as e:
-                # Ignore errors during cleanup as object is being destroyed
-                logger.warning("Error closing Weaviate client %s", e, exc_info=True)
-
    def _init_client(self, config: WeaviateConfig) -> weaviate.WeaviateClient:
        """
        Initializes and returns a connected Weaviate client.
@@ -1,8 +1,18 @@
 from pydantic import BaseModel
+from typing_extensions import TypedDict

 from models.dataset import DocumentSegment


+class AttachmentInfoDict(TypedDict):
+    id: str
+    name: str
+    extension: str
+    mime_type: str
+    source_url: str
+    size: int
+
+
 class RetrievalChildChunk(BaseModel):
    """Retrieval segments."""

@@ -19,5 +29,5 @@ class RetrievalSegments(BaseModel):
    segment: DocumentSegment
    child_chunks: list[RetrievalChildChunk] | None = None
    score: float | None = None
-    files: list[dict[str, str | int]] | None = None
+    files: list[AttachmentInfoDict] | None = None
    summary: str | None = None  # Summary content if retrieved via summary index
@@ -1,12 +1,38 @@
 import json
 import time
-from typing import Any, cast
+from typing import Any, NotRequired, cast

 import httpx
+from typing_extensions import TypedDict

 from extensions.ext_storage import storage


+class FirecrawlDocumentData(TypedDict):
+    title: str | None
+    description: str | None
+    source_url: str | None
+    markdown: str | None
+
+
+class CrawlStatusResponse(TypedDict):
+    status: str
+    total: int | None
+    current: int | None
+    data: list[FirecrawlDocumentData]
+
+
+class MapResponse(TypedDict):
+    success: bool
+    links: list[str]
+
+
+class SearchResponse(TypedDict):
+    success: bool
+    data: list[dict[str, Any]]
+    warning: NotRequired[str]
+
+
 class FirecrawlApp:
    def __init__(self, api_key=None, base_url=None):
        self.api_key = api_key
@@ -14,7 +40,7 @@ class FirecrawlApp:
        if self.api_key is None and self.base_url == "https://api.firecrawl.dev":
            raise ValueError("No API key provided")

-    def scrape_url(self, url, params=None) -> dict[str, Any]:
+    def scrape_url(self, url, params=None) -> FirecrawlDocumentData:
        # Documentation: https://docs.firecrawl.dev/api-reference/endpoint/scrape
        headers = self._prepare_headers()
        json_data = {
@@ -32,9 +58,7 @@ class FirecrawlApp:
            return self._extract_common_fields(data)
        elif response.status_code in {402, 409, 500, 429, 408}:
            self._handle_error(response, "scrape URL")
-            return {}  # Avoid additional exception after handling error
-        else:
-            raise Exception(f"Failed to scrape URL. Status code: {response.status_code}")
+        raise Exception(f"Failed to scrape URL. Status code: {response.status_code}")

    def crawl_url(self, url, params=None) -> str:
        # Documentation: https://docs.firecrawl.dev/api-reference/endpoint/crawl-post
@@ -51,7 +75,7 @@ class FirecrawlApp:
            self._handle_error(response, "start crawl job")
            return ""  # unreachable

-    def map(self, url: str, params: dict[str, Any] | None = None) -> dict[str, Any]:
+    def map(self, url: str, params: dict[str, Any] | None = None) -> MapResponse:
        # Documentation: https://docs.firecrawl.dev/api-reference/endpoint/map
        headers = self._prepare_headers()
        json_data: dict[str, Any] = {"url": url, "integration": "dify"}
@@ -60,14 +84,12 @@ class FirecrawlApp:
            json_data.update(params)
        response = self._post_request(self._build_url("v2/map"), json_data, headers)
        if response.status_code == 200:
-            return cast(dict[str, Any], response.json())
+            return cast(MapResponse, response.json())
        elif response.status_code in {402, 409, 500, 429, 408}:
            self._handle_error(response, "start map job")
-            return {}
-        else:
-            raise Exception(f"Failed to start map job. Status code: {response.status_code}")
+        raise Exception(f"Failed to start map job. Status code: {response.status_code}")

-    def check_crawl_status(self, job_id) -> dict[str, Any]:
+    def check_crawl_status(self, job_id) -> CrawlStatusResponse:
        headers = self._prepare_headers()
        response = self._get_request(self._build_url(f"v2/crawl/{job_id}"), headers)
        if response.status_code == 200:
@@ -77,7 +99,7 @@ class FirecrawlApp:
                if total == 0:
                    raise Exception("Failed to check crawl status. Error: No page found")
                data = crawl_status_response.get("data", [])
-                url_data_list = []
+                url_data_list: list[FirecrawlDocumentData] = []
                for item in data:
                    if isinstance(item, dict) and "metadata" in item and "markdown" in item:
                        url_data = self._extract_common_fields(item)
@@ -95,13 +117,15 @@ class FirecrawlApp:
                return self._format_crawl_status_response(
                    crawl_status_response.get("status"), crawl_status_response, []
                )
-        else:
-            self._handle_error(response, "check crawl status")
-            return {}  # unreachable
+        self._handle_error(response, "check crawl status")
+        raise RuntimeError("unreachable: _handle_error always raises")

    def _format_crawl_status_response(
-        self, status: str, crawl_status_response: dict[str, Any], url_data_list: list[dict[str, Any]]
-    ) -> dict[str, Any]:
+        self,
+        status: str,
+        crawl_status_response: dict[str, Any],
+        url_data_list: list[FirecrawlDocumentData],
+    ) -> CrawlStatusResponse:
        return {
            "status": status,
            "total": crawl_status_response.get("total"),
@@ -109,7 +133,7 @@ class FirecrawlApp:
            "data": url_data_list,
        }

-    def _extract_common_fields(self, item: dict[str, Any]) -> dict[str, Any]:
+    def _extract_common_fields(self, item: dict[str, Any]) -> FirecrawlDocumentData:
        return {
            "title": item.get("metadata", {}).get("title"),
            "description": item.get("metadata", {}).get("description"),
@@ -117,7 +141,7 @@ class FirecrawlApp:
            "markdown": item.get("markdown"),
        }

-    def _prepare_headers(self) -> dict[str, Any]:
+    def _prepare_headers(self) -> dict[str, str]:
        return {"Content-Type": "application/json", "Authorization": f"Bearer {self.api_key}"}

    def _build_url(self, path: str) -> str:
@@ -150,10 +174,10 @@ class FirecrawlApp:
            error_message = response.text or "Unknown error occurred"
        raise Exception(f"Failed to {action}. Status code: {response.status_code}. Error: {error_message}")  # type: ignore[return]

-    def search(self, query: str, params: dict[str, Any] | None = None) -> dict[str, Any]:
+    def search(self, query: str, params: dict[str, Any] | None = None) -> SearchResponse:
        # Documentation: https://docs.firecrawl.dev/api-reference/endpoint/search
        headers = self._prepare_headers()
-        json_data = {
+        json_data: dict[str, Any] = {
            "query": query,
            "limit": 5,
            "lang": "en",
@@ -170,12 +194,10 @@ class FirecrawlApp:
            json_data.update(params)
        response = self._post_request(self._build_url("v2/search"), json_data, headers)
        if response.status_code == 200:
-            response_data = response.json()
+            response_data: SearchResponse = response.json()
            if not response_data.get("success"):
                raise Exception(f"Search failed. Error: {response_data.get('warning', 'Unknown error')}")
-            return cast(dict[str, Any], response_data)
+            return response_data
        elif response.status_code in {402, 409, 500, 429, 408}:
            self._handle_error(response, "perform search")
-            return {}  # Avoid additional exception after handling error
-        else:
-            raise Exception(f"Failed to perform search. Status code: {response.status_code}")
+        raise Exception(f"Failed to perform search. Status code: {response.status_code}")
@@ -15,6 +15,7 @@ from core.rag.extractor.extractor_base import BaseExtractor
 from core.rag.models.document import Document
 from extensions.ext_database import db
 from extensions.ext_storage import storage
+from extensions.storage.storage_type import StorageType
 from libs.datetime_utils import naive_utc_now
 from models.enums import CreatorUserRole
 from models.model import UploadFile
@@ -150,7 +151,7 @@ class PdfExtractor(BaseExtractor):
                    # save file to db
                    upload_file = UploadFile(
                        tenant_id=self._tenant_id,
-                        storage_type=dify_config.STORAGE_TYPE,
+                        storage_type=StorageType(dify_config.STORAGE_TYPE),
                        key=file_key,
                        name=file_key,
                        size=len(img_bytes),
@@ -1,10 +1,11 @@
 import json
 from collections.abc import Generator
-from typing import Union
+from typing import Any, Union
 from urllib.parse import urljoin

 import httpx
 from httpx import Response
+from typing_extensions import TypedDict

 from core.rag.extractor.watercrawl.exceptions import (
    WaterCrawlAuthenticationError,
@@ -13,6 +14,27 @@ from core.rag.extractor.watercrawl.exceptions import (
 )


+class SpiderOptions(TypedDict):
+    max_depth: int
+    page_limit: int
+    allowed_domains: list[str]
+    exclude_paths: list[str]
+    include_paths: list[str]
+
+
+class PageOptions(TypedDict):
+    exclude_tags: list[str]
+    include_tags: list[str]
+    wait_time: int
+    include_html: bool
+    only_main_content: bool
+    include_links: bool
+    timeout: int
+    accept_cookies_selector: str
+    locale: str
+    actions: list[Any]
+
+
 class BaseAPIClient:
    def __init__(self, api_key, base_url):
        self.api_key = api_key
@@ -121,9 +143,9 @@ class WaterCrawlAPIClient(BaseAPIClient):
    def create_crawl_request(
        self,
        url: Union[list, str] | None = None,
-        spider_options: dict | None = None,
-        page_options: dict | None = None,
-        plugin_options: dict | None = None,
+        spider_options: SpiderOptions | None = None,
+        page_options: PageOptions | None = None,
+        plugin_options: dict[str, Any] | None = None,
    ):
        data = {
            # 'urls': url if isinstance(url, list) else [url],
@@ -176,8 +198,8 @@ class WaterCrawlAPIClient(BaseAPIClient):
    def scrape_url(
        self,
        url: str,
-        page_options: dict | None = None,
-        plugin_options: dict | None = None,
+        page_options: PageOptions | None = None,
+        plugin_options: dict[str, Any] | None = None,
        sync: bool = True,
        prefetched: bool = True,
    ):
@@ -2,16 +2,39 @@ from collections.abc import Generator
 from datetime import datetime
 from typing import Any

-from core.rag.extractor.watercrawl.client import WaterCrawlAPIClient
+from typing_extensions import TypedDict
+
+from core.rag.extractor.watercrawl.client import PageOptions, SpiderOptions, WaterCrawlAPIClient
+
+
+class WatercrawlDocumentData(TypedDict):
+    title: str | None
+    description: str | None
+    source_url: str | None
+    markdown: str | None
+
+
+class CrawlJobResponse(TypedDict):
+    status: str
+    job_id: str | None
+
+
+class WatercrawlCrawlStatusResponse(TypedDict):
+    status: str
+    job_id: str | None
+    total: int
+    current: int
+    data: list[WatercrawlDocumentData]
+    time_consuming: float


 class WaterCrawlProvider:
    def __init__(self, api_key, base_url: str | None = None):
        self.client = WaterCrawlAPIClient(api_key, base_url)

-    def crawl_url(self, url, options: dict | Any | None = None):
+    def crawl_url(self, url: str, options: dict[str, Any] | None = None) -> CrawlJobResponse:
        options = options or {}
-        spider_options = {
+        spider_options: SpiderOptions = {
            "max_depth": 1,
            "page_limit": 1,
            "allowed_domains": [],
@@ -25,7 +48,7 @@ class WaterCrawlProvider:
            spider_options["exclude_paths"] = options.get("excludes", "").split(",") if options.get("excludes") else []

        wait_time = options.get("wait_time", 1000)
-        page_options = {
+        page_options: PageOptions = {
            "exclude_tags": options.get("exclude_tags", "").split(",") if options.get("exclude_tags") else [],
            "include_tags": options.get("include_tags", "").split(",") if options.get("include_tags") else [],
            "wait_time": max(1000, wait_time),  # minimum wait time is 1 second
@@ -41,9 +64,9 @@ class WaterCrawlProvider:

        return {"status": "active", "job_id": result.get("uuid")}

-    def get_crawl_status(self, crawl_request_id):
+    def get_crawl_status(self, crawl_request_id: str) -> WatercrawlCrawlStatusResponse:
        response = self.client.get_crawl_request(crawl_request_id)
-        data = []
+        data: list[WatercrawlDocumentData] = []
        if response["status"] in ["new", "running"]:
            status = "active"
        else:
@@ -67,7 +90,7 @@ class WaterCrawlProvider:
            "time_consuming": time_consuming,
        }

-    def get_crawl_url_data(self, job_id, url) -> dict | None:
+    def get_crawl_url_data(self, job_id: str, url: str) -> WatercrawlDocumentData | None:
        if not job_id:
            return self.scrape_url(url)

@@ -82,11 +105,11 @@ class WaterCrawlProvider:

        return None

-    def scrape_url(self, url: str):
+    def scrape_url(self, url: str) -> WatercrawlDocumentData:
        response = self.client.scrape_url(url=url, sync=True, prefetched=True)
        return self._structure_data(response)

-    def _structure_data(self, result_object: dict):
+    def _structure_data(self, result_object: dict[str, Any]) -> WatercrawlDocumentData:
        if isinstance(result_object.get("result", {}), str):
            raise ValueError("Invalid result object. Expected a dictionary.")

@@ -98,7 +121,9 @@ class WaterCrawlProvider:
            "markdown": result_object.get("result", {}).get("markdown"),
        }

-    def _get_results(self, crawl_request_id: str, query_params: dict | None = None) -> Generator[dict, None, None]:
+    def _get_results(
+        self, crawl_request_id: str, query_params: dict | None = None
+    ) -> Generator[WatercrawlDocumentData, None, None]:
        page = 0
        page_size = 100

@@ -21,6 +21,7 @@ from core.rag.extractor.extractor_base import BaseExtractor
 from core.rag.models.document import Document
 from extensions.ext_database import db
 from extensions.ext_storage import storage
+from extensions.storage.storage_type import StorageType
 from libs.datetime_utils import naive_utc_now
 from models.enums import CreatorUserRole
 from models.model import UploadFile
@@ -112,7 +113,7 @@ class WordExtractor(BaseExtractor):
                        # save file to db
                        upload_file = UploadFile(
                            tenant_id=self.tenant_id,
-                            storage_type=dify_config.STORAGE_TYPE,
+                            storage_type=StorageType(dify_config.STORAGE_TYPE),
                            key=file_key,
                            name=file_key,
                            size=0,
@@ -140,7 +141,7 @@ class WordExtractor(BaseExtractor):
                    # save file to db
                    upload_file = UploadFile(
                        tenant_id=self.tenant_id,
-                        storage_type=dify_config.STORAGE_TYPE,
+                        storage_type=StorageType(dify_config.STORAGE_TYPE),
                        key=file_key,
                        name=file_key,
                        size=0,
@@ -9,6 +9,7 @@ from flask import current_app
 from sqlalchemy import delete, func, select

 from core.db.session_factory import session_factory
+from core.rag.index_processor.index_processor_base import SummaryIndexSettingDict
 from core.workflow.nodes.knowledge_index.exc import KnowledgeIndexNodeError
 from core.workflow.nodes.knowledge_index.protocols import Preview, PreviewItem, QaPreview
 from models.dataset import Dataset, Document, DocumentSegment
@@ -51,7 +52,7 @@ class IndexProcessor:
        original_document_id: str,
        chunks: Mapping[str, Any],
        batch: Any,
-        summary_index_setting: dict | None = None,
+        summary_index_setting: SummaryIndexSettingDict | None = None,
    ):
        with session_factory.create_session() as session:
            document = session.query(Document).filter_by(id=document_id).first()
@@ -131,7 +132,12 @@ class IndexProcessor:
        }

    def get_preview_output(
-        self, chunks: Any, dataset_id: str, document_id: str, chunk_structure: str, summary_index_setting: dict | None
+        self,
+        chunks: Any,
+        dataset_id: str,
+        document_id: str,
+        chunk_structure: str,
+        summary_index_setting: SummaryIndexSettingDict | None,
    ) -> Preview:
        doc_language = None
        with session_factory.create_session() as session:
@@ -7,14 +7,16 @@ import os
 import re
 from abc import ABC, abstractmethod
 from collections.abc import Mapping
-from typing import TYPE_CHECKING, Any, Optional
+from typing import TYPE_CHECKING, Any, NotRequired, Optional
 from urllib.parse import unquote, urlparse

 import httpx
+from typing_extensions import TypedDict

 from configs import dify_config
 from core.entities.knowledge_entities import PreviewDetail
 from core.helper import ssrf_proxy
+from core.rag.data_post_processor.data_post_processor import RerankingModelDict
 from core.rag.extractor.entity.extract_setting import ExtractSetting
 from core.rag.index_processor.constant.doc_type import DocType
 from core.rag.models.document import AttachmentDocument, Document
@@ -35,6 +37,13 @@ if TYPE_CHECKING:
    from core.model_manager import ModelInstance


+class SummaryIndexSettingDict(TypedDict):
+    enable: bool
+    model_name: NotRequired[str]
+    model_provider_name: NotRequired[str]
+    summary_prompt: NotRequired[str]
+
+
 class BaseIndexProcessor(ABC):
    """Interface for extract files."""

@@ -51,7 +60,7 @@ class BaseIndexProcessor(ABC):
        self,
        tenant_id: str,
        preview_texts: list[PreviewDetail],
-        summary_index_setting: dict,
+        summary_index_setting: SummaryIndexSettingDict,
        doc_language: str | None = None,
    ) -> list[PreviewDetail]:
        """
@@ -98,7 +107,7 @@ class BaseIndexProcessor(ABC):
        dataset: Dataset,
        top_k: int,
        score_threshold: float,
-        reranking_model: dict,
+        reranking_model: RerankingModelDict,
    ) -> list[Document]:
        raise NotImplementedError

@@ -14,6 +14,7 @@ from core.llm_generator.prompts import DEFAULT_GENERATOR_SUMMARY_PROMPT
 from core.model_manager import ModelInstance
 from core.provider_manager import ProviderManager
 from core.rag.cleaner.clean_processor import CleanProcessor
+from core.rag.data_post_processor.data_post_processor import RerankingModelDict
 from core.rag.datasource.keyword.keyword_factory import Keyword
 from core.rag.datasource.retrieval_service import RetrievalService
 from core.rag.datasource.vdb.vector_factory import Vector
@@ -22,7 +23,7 @@ from core.rag.extractor.entity.extract_setting import ExtractSetting
 from core.rag.extractor.extract_processor import ExtractProcessor
 from core.rag.index_processor.constant.doc_type import DocType
 from core.rag.index_processor.constant.index_type import IndexStructureType
-from core.rag.index_processor.index_processor_base import BaseIndexProcessor
+from core.rag.index_processor.index_processor_base import BaseIndexProcessor, SummaryIndexSettingDict
 from core.rag.models.document import AttachmentDocument, Document, MultimodalGeneralStructureChunk
 from core.rag.retrieval.retrieval_methods import RetrievalMethod
 from core.tools.utils.text_processing_utils import remove_leading_symbols
@@ -175,7 +176,7 @@ class ParagraphIndexProcessor(BaseIndexProcessor):
        dataset: Dataset,
        top_k: int,
        score_threshold: float,
-        reranking_model: dict,
+        reranking_model: RerankingModelDict,
    ) -> list[Document]:
        # Set search parameters.
        results = RetrievalService.retrieve(
@@ -278,7 +279,7 @@ class ParagraphIndexProcessor(BaseIndexProcessor):
        self,
        tenant_id: str,
        preview_texts: list[PreviewDetail],
-        summary_index_setting: dict,
+        summary_index_setting: SummaryIndexSettingDict,
        doc_language: str | None = None,
    ) -> list[PreviewDetail]:
        """
@@ -362,7 +363,7 @@ class ParagraphIndexProcessor(BaseIndexProcessor):
    def generate_summary(
        tenant_id: str,
        text: str,
-        summary_index_setting: dict | None = None,
+        summary_index_setting: SummaryIndexSettingDict | None = None,
        segment_id: str | None = None,
        document_language: str | None = None,
    ) -> tuple[str, LLMUsage]:
@@ -11,6 +11,7 @@ from core.db.session_factory import session_factory
 from core.entities.knowledge_entities import PreviewDetail
 from core.model_manager import ModelInstance
 from core.rag.cleaner.clean_processor import CleanProcessor
+from core.rag.data_post_processor.data_post_processor import RerankingModelDict
 from core.rag.datasource.retrieval_service import RetrievalService
 from core.rag.datasource.vdb.vector_factory import Vector
 from core.rag.docstore.dataset_docstore import DatasetDocumentStore
@@ -18,7 +19,7 @@ from core.rag.extractor.entity.extract_setting import ExtractSetting
 from core.rag.extractor.extract_processor import ExtractProcessor
 from core.rag.index_processor.constant.doc_type import DocType
 from core.rag.index_processor.constant.index_type import IndexStructureType
-from core.rag.index_processor.index_processor_base import BaseIndexProcessor
+from core.rag.index_processor.index_processor_base import BaseIndexProcessor, SummaryIndexSettingDict
 from core.rag.models.document import AttachmentDocument, ChildDocument, Document, ParentChildStructureChunk
 from core.rag.retrieval.retrieval_methods import RetrievalMethod
 from extensions.ext_database import db
@@ -215,7 +216,7 @@ class ParentChildIndexProcessor(BaseIndexProcessor):
        dataset: Dataset,
        top_k: int,
        score_threshold: float,
-        reranking_model: dict,
+        reranking_model: RerankingModelDict,
    ) -> list[Document]:
        # Set search parameters.
        results = RetrievalService.retrieve(
@@ -361,7 +362,7 @@ class ParentChildIndexProcessor(BaseIndexProcessor):
        self,
        tenant_id: str,
        preview_texts: list[PreviewDetail],
-        summary_index_setting: dict,
+        summary_index_setting: SummaryIndexSettingDict,
        doc_language: str | None = None,
    ) -> list[PreviewDetail]:
        """
@@ -15,13 +15,14 @@ from core.db.session_factory import session_factory
 from core.entities.knowledge_entities import PreviewDetail
 from core.llm_generator.llm_generator import LLMGenerator
 from core.rag.cleaner.clean_processor import CleanProcessor
+from core.rag.data_post_processor.data_post_processor import RerankingModelDict
 from core.rag.datasource.retrieval_service import RetrievalService
 from core.rag.datasource.vdb.vector_factory import Vector
 from core.rag.docstore.dataset_docstore import DatasetDocumentStore
 from core.rag.extractor.entity.extract_setting import ExtractSetting
 from core.rag.extractor.extract_processor import ExtractProcessor
 from core.rag.index_processor.constant.index_type import IndexStructureType
-from core.rag.index_processor.index_processor_base import BaseIndexProcessor
+from core.rag.index_processor.index_processor_base import BaseIndexProcessor, SummaryIndexSettingDict
 from core.rag.models.document import AttachmentDocument, Document, QAStructureChunk
 from core.rag.retrieval.retrieval_methods import RetrievalMethod
 from core.tools.utils.text_processing_utils import remove_leading_symbols
@@ -185,7 +186,7 @@ class QAIndexProcessor(BaseIndexProcessor):
        dataset: Dataset,
        top_k: int,
        score_threshold: float,
-        reranking_model: dict,
+        reranking_model: RerankingModelDict,
    ):
        # Set search parameters.
        results = RetrievalService.retrieve(
@@ -244,7 +245,7 @@ class QAIndexProcessor(BaseIndexProcessor):
        self,
        tenant_id: str,
        preview_texts: list[PreviewDetail],
-        summary_index_setting: dict,
+        summary_index_setting: SummaryIndexSettingDict,
        doc_language: str | None = None,
    ) -> list[PreviewDetail]:
        """
@@ -31,9 +31,9 @@ from core.ops.utils import measure_time
 from core.prompt.advanced_prompt_transform import AdvancedPromptTransform
 from core.prompt.entities.advanced_prompt_entities import ChatModelMessage, CompletionModelPromptTemplate
 from core.prompt.simple_prompt_transform import ModelMode
-from core.rag.data_post_processor.data_post_processor import DataPostProcessor
+from core.rag.data_post_processor.data_post_processor import DataPostProcessor, RerankingModelDict, WeightsDict
 from core.rag.datasource.keyword.jieba.jieba_keyword_table_handler import JiebaKeywordTableHandler
-from core.rag.datasource.retrieval_service import RetrievalService
+from core.rag.datasource.retrieval_service import DefaultRetrievalModelDict, RetrievalService
 from core.rag.entities.citation_metadata import RetrievalSourceMetadata
 from core.rag.entities.context_entities import DocumentContext
 from core.rag.entities.metadata_entities import Condition, MetadataCondition
@@ -83,11 +83,11 @@ from models.dataset import (
 )
 from models.dataset import Document as DatasetDocument
 from models.dataset import Document as DocumentModel
-from models.enums import CreatorUserRole
+from models.enums import CreatorUserRole, DatasetQuerySource
 from services.external_knowledge_service import ExternalDatasetService
 from services.feature_service import FeatureService

-default_retrieval_model: dict[str, Any] = {
+default_retrieval_model: DefaultRetrievalModelDict = {
    "search_method": RetrievalMethod.SEMANTIC_SEARCH,
    "reranking_enable": False,
    "reranking_model": {"reranking_provider_name": "", "reranking_model_name": ""},
@@ -666,7 +666,11 @@ class DatasetRetrieval:
                            document_ids_filter = document_ids
                        else:
                            return []
-                    retrieval_model_config = dataset.retrieval_model or default_retrieval_model
+                    retrieval_model_config: DefaultRetrievalModelDict = (
+                        cast(DefaultRetrievalModelDict, dataset.retrieval_model)
+                        if dataset.retrieval_model
+                        else default_retrieval_model
+                    )

                    # get top k
                    top_k = retrieval_model_config["top_k"]
@@ -727,8 +731,8 @@ class DatasetRetrieval:
        top_k: int,
        score_threshold: float,
        reranking_mode: str,
-        reranking_model: dict | None = None,
-        weights: dict[str, Any] | None = None,
+        reranking_model: RerankingModelDict | None = None,
+        weights: WeightsDict | None = None,
        reranking_enable: bool = True,
        message_id: str | None = None,
        metadata_filter_document_ids: dict[str, list[str]] | None = None,
@@ -1008,7 +1012,7 @@ class DatasetRetrieval:
                dataset_query = DatasetQuery(
                    dataset_id=dataset_id,
                    content=json.dumps(contents),
-                    source="app",
+                    source=DatasetQuerySource.APP,
                    source_app_id=app_id,
                    created_by_role=CreatorUserRole(user_from),
                    created_by=user_id,
@@ -1058,7 +1062,11 @@ class DatasetRetrieval:
                    all_documents.append(document)
            else:
                # get retrieval model , if the model is not setting , using default
-                retrieval_model = dataset.retrieval_model or default_retrieval_model
+                retrieval_model: DefaultRetrievalModelDict = (
+                    cast(DefaultRetrievalModelDict, dataset.retrieval_model)
+                    if dataset.retrieval_model
+                    else default_retrieval_model
+                )

                if dataset.indexing_technique == "economy":
                    # use keyword table query
@@ -1132,7 +1140,7 @@ class DatasetRetrieval:

        if retrieve_config.retrieve_strategy == DatasetRetrieveConfigEntity.RetrieveStrategy.SINGLE:
            # get retrieval model config
-            default_retrieval_model = {
+            default_retrieval_model: DefaultRetrievalModelDict = {
                "search_method": RetrievalMethod.SEMANTIC_SEARCH,
                "reranking_enable": False,
                "reranking_model": {"reranking_provider_name": "", "reranking_model_name": ""},
@@ -1141,7 +1149,11 @@ class DatasetRetrieval:
            }

            for dataset in available_datasets:
-                retrieval_model_config = dataset.retrieval_model or default_retrieval_model
+                retrieval_model_config: DefaultRetrievalModelDict = (
+                    cast(DefaultRetrievalModelDict, dataset.retrieval_model)
+                    if dataset.retrieval_model
+                    else default_retrieval_model
+                )

                # get top k
                top_k = retrieval_model_config["top_k"]
@@ -1181,8 +1193,8 @@ class DatasetRetrieval:
                hit_callbacks=[hit_callback],
                return_resource=return_resource,
                retriever_from=invoke_from.to_source(),
-                reranking_provider_name=retrieve_config.reranking_model.get("reranking_provider_name"),
-                reranking_model_name=retrieve_config.reranking_model.get("reranking_model_name"),
+                reranking_provider_name=retrieve_config.reranking_model["reranking_provider_name"],
+                reranking_model_name=retrieve_config.reranking_model["reranking_model_name"],
            )

            tools.append(tool)
@@ -1685,8 +1697,8 @@ class DatasetRetrieval:
        tenant_id: str,
        reranking_enable: bool,
        reranking_mode: str,
-        reranking_model: dict | None,
-        weights: dict[str, Any] | None,
+        reranking_model: RerankingModelDict | None,
+        weights: WeightsDict | None,
        top_k: int,
        score_threshold: float,
        query: str | None,
@@ -2,6 +2,7 @@ import concurrent.futures
 import logging

 from core.db.session_factory import session_factory
+from core.rag.index_processor.index_processor_base import SummaryIndexSettingDict
 from models.dataset import Dataset, Document, DocumentSegment, DocumentSegmentSummary
 from services.summary_index_service import SummaryIndexService
 from tasks.generate_summary_index_task import generate_summary_index_task
@@ -11,7 +12,11 @@ logger = logging.getLogger(__name__)

 class SummaryIndex:
    def generate_and_vectorize_summary(
-        self, dataset_id: str, document_id: str, is_preview: bool, summary_index_setting: dict | None = None
+        self,
+        dataset_id: str,
+        document_id: str,
+        is_preview: bool,
+        summary_index_setting: SummaryIndexSettingDict | None = None,
    ) -> None:
        if is_preview:
            with session_factory.create_session() as session:
@@ -34,7 +34,7 @@ from core.tools.workflow_as_tool.tool import WorkflowTool
 from dify_graph.file import FileType
 from dify_graph.file.models import FileTransferMethod
 from extensions.ext_database import db
-from models.enums import CreatorUserRole
+from models.enums import CreatorUserRole, MessageFileBelongsTo
 from models.model import Message, MessageFile

 logger = logging.getLogger(__name__)
@@ -352,7 +352,7 @@ class ToolEngine:
                message_id=agent_message.id,
                type=file_type,
                transfer_method=FileTransferMethod.TOOL_FILE,
-                belongs_to="assistant",
+                belongs_to=MessageFileBelongsTo.ASSISTANT,
                url=message.url,
                upload_file_id=tool_file_id,
                created_by_role=(
@@ -72,6 +72,11 @@ class ApiProviderControllerItem(TypedDict):
    controller: ApiToolProviderController


+class EmojiIconDict(TypedDict):
+    background: str
+    content: str
+
+
 class ToolManager:
    _builtin_provider_lock = Lock()
    _hardcoded_providers: dict[str, BuiltinToolProviderController] = {}
@@ -916,7 +921,7 @@ class ToolManager:
        )

    @classmethod
-    def generate_workflow_tool_icon_url(cls, tenant_id: str, provider_id: str) -> Mapping[str, str]:
+    def generate_workflow_tool_icon_url(cls, tenant_id: str, provider_id: str) -> EmojiIconDict:
        try:
            workflow_provider: WorkflowToolProvider | None = (
                db.session.query(WorkflowToolProvider)
@@ -933,7 +938,7 @@ class ToolManager:
            return {"background": "#252525", "content": "\ud83d\ude01"}

    @classmethod
-    def generate_api_tool_icon_url(cls, tenant_id: str, provider_id: str) -> Mapping[str, str]:
+    def generate_api_tool_icon_url(cls, tenant_id: str, provider_id: str) -> EmojiIconDict:
        try:
            api_provider: ApiToolProvider | None = (
                db.session.query(ApiToolProvider)
@@ -950,7 +955,7 @@ class ToolManager:
            return {"background": "#252525", "content": "\ud83d\ude01"}

    @classmethod
-    def generate_mcp_tool_icon_url(cls, tenant_id: str, provider_id: str) -> Mapping[str, str] | str:
+    def generate_mcp_tool_icon_url(cls, tenant_id: str, provider_id: str) -> EmojiIconDict | dict[str, str] | str:
        try:
            with Session(db.engine) as session:
                mcp_service = MCPToolManageService(session=session)
@@ -970,7 +975,7 @@ class ToolManager:
        tenant_id: str,
        provider_type: ToolProviderType,
        provider_id: str,
-    ) -> str | Mapping[str, str]:
+    ) -> str | EmojiIconDict | dict[str, str]:
        """
        get the tool icon

@@ -1,5 +1,4 @@
 import threading
-from typing import Any

 from flask import Flask, current_app
 from pydantic import BaseModel, Field
@@ -13,11 +12,12 @@ from core.rag.models.document import Document as RagDocument
 from core.rag.rerank.rerank_model import RerankModelRunner
 from core.rag.retrieval.retrieval_methods import RetrievalMethod
 from core.tools.utils.dataset_retriever.dataset_retriever_base_tool import DatasetRetrieverBaseTool
+from core.tools.utils.dataset_retriever.dataset_retriever_tool import DefaultRetrievalModelDict
 from dify_graph.model_runtime.entities.model_entities import ModelType
 from extensions.ext_database import db
 from models.dataset import Dataset, Document, DocumentSegment

-default_retrieval_model: dict[str, Any] = {
+default_retrieval_model: DefaultRetrievalModelDict = {
    "search_method": RetrievalMethod.SEMANTIC_SEARCH,
    "reranking_enable": False,
    "reranking_model": {"reranking_provider_name": "", "reranking_model_name": ""},
@@ -1,9 +1,10 @@
-from typing import Any, cast
+from typing import NotRequired, TypedDict, cast

 from pydantic import BaseModel, Field
 from sqlalchemy import select

 from core.app.app_config.entities import DatasetRetrieveConfigEntity, ModelConfig
+from core.rag.data_post_processor.data_post_processor import RerankingModelDict, WeightsDict
 from core.rag.datasource.retrieval_service import RetrievalService
 from core.rag.entities.citation_metadata import RetrievalSourceMetadata
 from core.rag.entities.context_entities import DocumentContext
@@ -16,7 +17,19 @@ from models.dataset import Dataset
 from models.dataset import Document as DatasetDocument
 from services.external_knowledge_service import ExternalDatasetService

-default_retrieval_model: dict[str, Any] = {
+
+class DefaultRetrievalModelDict(TypedDict):
+    search_method: RetrievalMethod
+    reranking_enable: bool
+    reranking_model: RerankingModelDict
+    reranking_mode: NotRequired[str]
+    weights: NotRequired[WeightsDict | None]
+    score_threshold: NotRequired[float]
+    top_k: int
+    score_threshold_enabled: bool
+
+
+default_retrieval_model: DefaultRetrievalModelDict = {
    "search_method": RetrievalMethod.SEMANTIC_SEARCH,
    "reranking_enable": False,
    "reranking_model": {"reranking_provider_name": "", "reranking_model_name": ""},
@@ -125,7 +138,7 @@ class DatasetRetrieverTool(DatasetRetrieverBaseTool):
            if metadata_condition and not document_ids_filter:
                return ""
            # get retrieval model , if the model is not setting , using default
-            retrieval_model: dict[str, Any] = dataset.retrieval_model or default_retrieval_model
+            retrieval_model = dataset.retrieval_model or default_retrieval_model
            retrieval_resource_list: list[RetrievalSourceMetadata] = []
            if dataset.indexing_technique == "economy":
                # use keyword table query
@@ -1,4 +1,5 @@
 import re
+from collections.abc import Mapping
 from json import dumps as json_dumps
 from json import loads as json_loads
 from json.decoder import JSONDecodeError
@@ -20,10 +21,18 @@ class InterfaceDict(TypedDict):
    operation: dict[str, Any]


+class OpenAPISpecDict(TypedDict):
+    openapi: str
+    info: dict[str, str]
+    servers: list[dict[str, Any]]
+    paths: dict[str, Any]
+    components: dict[str, Any]
+
+
 class ApiBasedToolSchemaParser:
    @staticmethod
    def parse_openapi_to_tool_bundle(
-        openapi: dict, extra_info: dict | None = None, warning: dict | None = None
+        openapi: Mapping[str, Any], extra_info: dict | None = None, warning: dict | None = None
    ) -> list[ApiToolBundle]:
        warning = warning if warning is not None else {}
        extra_info = extra_info if extra_info is not None else {}
@@ -277,7 +286,7 @@ class ApiBasedToolSchemaParser:
    @staticmethod
    def parse_swagger_to_openapi(
        swagger: dict, extra_info: dict | None = None, warning: dict | None = None
-    ) -> dict[str, Any]:
+    ) -> OpenAPISpecDict:
        warning = warning or {}
        """
        parse swagger to openapi
@@ -293,7 +302,7 @@ class ApiBasedToolSchemaParser:
        if len(servers) == 0:
            raise ToolApiSchemaError("No server found in the swagger yaml.")

-        converted_openapi: dict[str, Any] = {
+        converted_openapi: OpenAPISpecDict = {
            "openapi": "3.0.0",
            "info": {
                "title": info.get("title", "Swagger"),
@@ -3,7 +3,6 @@ from typing import Final
 TRIGGER_WEBHOOK_NODE_TYPE: Final[str] = "trigger-webhook"
 TRIGGER_SCHEDULE_NODE_TYPE: Final[str] = "trigger-schedule"
 TRIGGER_PLUGIN_NODE_TYPE: Final[str] = "trigger-plugin"
-TRIGGER_INFO_METADATA_KEY: Final[str] = "trigger_info"

 TRIGGER_NODE_TYPES: Final[frozenset[str]] = frozenset(
    {
@@ -282,6 +282,7 @@ class DifyNodeFactory(NodeFactory):
            max_text_size=dify_config.HTTP_REQUEST_NODE_MAX_TEXT_SIZE,
            ssl_verify=dify_config.HTTP_REQUEST_NODE_SSL_VERIFY,
            ssrf_default_max_retries=dify_config.SSRF_DEFAULT_MAX_RETRIES,
+            secret_key=dify_config.SECRET_KEY,
        )

        self._llm_credentials_provider, self._llm_model_factory = build_dify_model_access(self._dify_context.tenant_id)
@@ -2,6 +2,7 @@ from typing import Literal, Union

 from pydantic import BaseModel

+from core.rag.index_processor.index_processor_base import SummaryIndexSettingDict
 from core.rag.retrieval.retrieval_methods import RetrievalMethod
 from core.workflow.nodes.knowledge_index import KNOWLEDGE_INDEX_NODE_TYPE
 from dify_graph.entities.base_node_data import BaseNodeData
@@ -161,4 +162,4 @@ class KnowledgeIndexNodeData(BaseNodeData):
    chunk_structure: str
    index_chunk_variable_selector: list[str]
    indexing_technique: str | None = None
-    summary_index_setting: dict | None = None
+    summary_index_setting: SummaryIndexSettingDict | None = None
@@ -3,6 +3,7 @@ from collections.abc import Mapping
 from typing import TYPE_CHECKING, Any

 from core.rag.index_processor.index_processor import IndexProcessor
+from core.rag.index_processor.index_processor_base import SummaryIndexSettingDict
 from core.rag.summary_index.summary_index import SummaryIndex
 from core.workflow.nodes.knowledge_index import KNOWLEDGE_INDEX_NODE_TYPE
 from dify_graph.entities.graph_config import NodeConfigDict
@@ -127,7 +128,7 @@ class KnowledgeIndexNode(Node[KnowledgeIndexNodeData]):
        is_preview: bool,
        batch: Any,
        chunks: Mapping[str, Any],
-        summary_index_setting: dict | None = None,
+        summary_index_setting: SummaryIndexSettingDict | None = None,
    ):
        if not document_id:
            raise KnowledgeIndexNodeError("document_id is required.")
@@ -9,6 +9,7 @@ from collections.abc import Mapping, Sequence
 from typing import TYPE_CHECKING, Any, Literal

 from core.app.app_config.entities import DatasetRetrieveConfigEntity
+from core.rag.data_post_processor.data_post_processor import RerankingModelDict, WeightsDict
 from core.rag.retrieval.dataset_retrieval import DatasetRetrieval
 from dify_graph.entities import GraphInitParams
 from dify_graph.entities.graph_config import NodeConfigDict
@@ -201,8 +202,8 @@ class KnowledgeRetrievalNode(LLMUsageTrackingMixin, Node[KnowledgeRetrievalNodeD
        elif str(node_data.retrieval_mode) == DatasetRetrieveConfigEntity.RetrieveStrategy.MULTIPLE:
            if node_data.multiple_retrieval_config is None:
                raise ValueError("multiple_retrieval_config is required")
-            reranking_model = None
-            weights = None
+            reranking_model: RerankingModelDict | None = None
+            weights: WeightsDict | None = None
            match node_data.multiple_retrieval_config.reranking_mode:
                case "reranking_model":
                    if node_data.multiple_retrieval_config.reranking_model:
@@ -2,6 +2,7 @@ from typing import Any, Literal, Protocol

 from pydantic import BaseModel, Field

+from core.rag.data_post_processor.data_post_processor import RerankingModelDict, WeightsDict
 from dify_graph.model_runtime.entities import LLMUsage
 from dify_graph.nodes.llm.entities import ModelConfig

@@ -75,8 +76,8 @@ class KnowledgeRetrievalRequest(BaseModel):
    top_k: int = Field(default=0, description="Number of top results to return")
    score_threshold: float = Field(default=0.0, description="Minimum relevance score threshold")
    reranking_mode: str = Field(default="reranking_model", description="Reranking strategy")
-    reranking_model: dict | None = Field(default=None, description="Reranking model configuration")
-    weights: dict[str, Any] | None = Field(default=None, description="Weights for weighted score reranking")
+    reranking_model: RerankingModelDict | None = Field(default=None, description="Reranking model configuration")
+    weights: WeightsDict | None = Field(default=None, description="Weights for weighted score reranking")
    reranking_enable: bool = Field(default=True, description="Whether reranking is enabled")
    attachment_ids: list[str] | None = Field(default=None, description="List of attachment file IDs for retrieval")

@@ -1,7 +1,7 @@
 from collections.abc import Mapping
-from typing import Any, cast
+from typing import Any

-from core.trigger.constants import TRIGGER_INFO_METADATA_KEY, TRIGGER_PLUGIN_NODE_TYPE
+from core.trigger.constants import TRIGGER_PLUGIN_NODE_TYPE
 from dify_graph.constants import SYSTEM_VARIABLE_NODE_ID
 from dify_graph.entities.workflow_node_execution import WorkflowNodeExecutionStatus
 from dify_graph.enums import NodeExecutionType, WorkflowNodeExecutionMetadataKey
@@ -47,7 +47,7 @@ class TriggerEventNode(Node[TriggerEventNodeData]):

        # Get trigger data passed when workflow was triggered
        metadata: dict[WorkflowNodeExecutionMetadataKey, Any] = {
-            cast(WorkflowNodeExecutionMetadataKey, TRIGGER_INFO_METADATA_KEY): {
+            WorkflowNodeExecutionMetadataKey.TRIGGER_INFO: {
                "provider_id": self.node_data.provider_id,
                "event_name": self.node_data.event_name,
                "plugin_unique_identifier": self.node_data.plugin_unique_identifier,
@@ -0,0 +1,27 @@
+"""Helpers for workflow recursion depth propagation.
+
+The HTTP request node emits a reserved depth header pair on outbound requests,
+and ``services.trigger.webhook_service`` validates that pair when a webhook is
+received. The signature binds the propagated depth to the concrete HTTP method
+and request path so a depth value captured for one endpoint cannot be replayed
+verbatim against another path.
+"""
+
+import hashlib
+import hmac
+
+
+def build_workflow_call_depth_signature(*, secret_key: str, method: str, path: str, depth: str) -> str:
+    """Build the stable HMAC payload for workflow call-depth propagation.
+
+    Args:
+        secret_key: Shared signing key used by both sender and receiver.
+        method: Outbound or inbound HTTP method.
+        path: Request path that the signature is bound to.
+        depth: Workflow call depth value serialized as a string.
+
+    Returns:
+        Hex-encoded HMAC-SHA256 digest for the method/path/depth tuple.
+    """
+    payload = f"{method.upper()}:{path}:{depth}"
+    return hmac.new(secret_key.encode("utf-8"), payload.encode("utf-8"), hashlib.sha256).hexdigest()
@@ -2,3 +2,8 @@ SYSTEM_VARIABLE_NODE_ID = "sys"
 ENVIRONMENT_VARIABLE_NODE_ID = "env"
 CONVERSATION_VARIABLE_NODE_ID = "conversation"
 RAG_PIPELINE_VARIABLE_NODE_ID = "rag"
+
+# Reserved for internal workflow-to-workflow HTTP calls. External callers should
+# not rely on or set this header.
+WORKFLOW_CALL_DEPTH_HEADER = "X-Dify-Workflow-Call-Depth"
+WORKFLOW_CALL_DEPTH_SIGNATURE_HEADER = "X-Dify-Workflow-Call-Depth-Signature"
@@ -245,6 +245,9 @@ _END_STATE = frozenset(
 class WorkflowNodeExecutionMetadataKey(StrEnum):
    """
    Node Run Metadata Key.
+
+    Values in this enum are persisted as execution metadata and must stay in sync
+    with every node that writes `NodeRunResult.metadata`.
    """

    TOTAL_TOKENS = "total_tokens"
@@ -266,6 +269,7 @@ class WorkflowNodeExecutionMetadataKey(StrEnum):
    ERROR_STRATEGY = "error_strategy"  # node in continue on error mode return the field
    LOOP_VARIABLE_MAP = "loop_variable_map"  # single loop variable output
    DATASOURCE_INFO = "datasource_info"
+    TRIGGER_INFO = "trigger_info"
    COMPLETED_REASON = "completed_reason"  # completed reason for loop node


@@ -159,6 +159,7 @@ class ErrorHandler:
            node_id=event.node_id,
            node_type=event.node_type,
            start_at=event.start_at,
+            finished_at=event.finished_at,
            node_run_result=NodeRunResult(
                status=WorkflowNodeExecutionStatus.EXCEPTION,
                inputs=event.node_run_result.inputs,
@@ -198,6 +199,7 @@ class ErrorHandler:
            node_id=event.node_id,
            node_type=event.node_type,
            start_at=event.start_at,
+            finished_at=event.finished_at,
            node_run_result=NodeRunResult(
                status=WorkflowNodeExecutionStatus.EXCEPTION,
                inputs=event.node_run_result.inputs,
@@ -15,10 +15,13 @@ from typing import TYPE_CHECKING, final
 from typing_extensions import override

 from dify_graph.context import IExecutionContext
+from dify_graph.enums import WorkflowNodeExecutionStatus
 from dify_graph.graph import Graph
 from dify_graph.graph_engine.layers.base import GraphEngineLayer
-from dify_graph.graph_events import GraphNodeEventBase, NodeRunFailedEvent, is_node_result_event
+from dify_graph.graph_events import GraphNodeEventBase, NodeRunFailedEvent, NodeRunStartedEvent, is_node_result_event
+from dify_graph.node_events import NodeRunResult
 from dify_graph.nodes.base.node import Node
+from libs.datetime_utils import naive_utc_now

 from .ready_queue import ReadyQueue

@@ -65,6 +68,7 @@ class Worker(threading.Thread):
        self._stop_event = threading.Event()
        self._layers = layers if layers is not None else []
        self._last_task_time = time.time()
+        self._current_node_started_at: datetime | None = None

    def stop(self) -> None:
        """Signal the worker to stop processing."""
@@ -104,18 +108,15 @@ class Worker(threading.Thread):
            self._last_task_time = time.time()
            node = self._graph.nodes[node_id]
            try:
+                self._current_node_started_at = None
                self._execute_node(node)
                self._ready_queue.task_done()
            except Exception as e:
-                error_event = NodeRunFailedEvent(
-                    id=node.execution_id,
-                    node_id=node.id,
-                    node_type=node.node_type,
-                    in_iteration_id=None,
-                    error=str(e),
-                    start_at=datetime.now(),
+                self._event_queue.put(
+                    self._build_fallback_failure_event(node, e, started_at=self._current_node_started_at)
                )
-                self._event_queue.put(error_event)
+            finally:
+                self._current_node_started_at = None

    def _execute_node(self, node: Node) -> None:
        """
@@ -136,6 +137,8 @@ class Worker(threading.Thread):
                try:
                    node_events = node.run()
                    for event in node_events:
+                        if isinstance(event, NodeRunStartedEvent) and event.id == node.execution_id:
+                            self._current_node_started_at = event.start_at
                        self._event_queue.put(event)
                        if is_node_result_event(event):
                            result_event = event
@@ -149,6 +152,8 @@ class Worker(threading.Thread):
            try:
                node_events = node.run()
                for event in node_events:
+                    if isinstance(event, NodeRunStartedEvent) and event.id == node.execution_id:
+                        self._current_node_started_at = event.start_at
                    self._event_queue.put(event)
                    if is_node_result_event(event):
                        result_event = event
@@ -177,3 +182,24 @@ class Worker(threading.Thread):
            except Exception:
                # Silently ignore layer errors to prevent disrupting node execution
                continue
+
+    def _build_fallback_failure_event(
+        self, node: Node, error: Exception, *, started_at: datetime | None = None
+    ) -> NodeRunFailedEvent:
+        """Build a failed event when worker-level execution aborts before a node emits its own result event."""
+        failure_time = naive_utc_now()
+        error_message = str(error)
+        return NodeRunFailedEvent(
+            id=node.execution_id,
+            node_id=node.id,
+            node_type=node.node_type,
+            in_iteration_id=None,
+            error=error_message,
+            start_at=started_at or failure_time,
+            finished_at=failure_time,
+            node_run_result=NodeRunResult(
+                status=WorkflowNodeExecutionStatus.FAILED,
+                error=error_message,
+                error_type=type(error).__name__,
+            ),
+        )
@@ -36,16 +36,19 @@ class NodeRunRetrieverResourceEvent(GraphNodeEventBase):

 class NodeRunSucceededEvent(GraphNodeEventBase):
    start_at: datetime = Field(..., description="node start time")
+    finished_at: datetime | None = Field(default=None, description="node finish time")


 class NodeRunFailedEvent(GraphNodeEventBase):
    error: str = Field(..., description="error")
    start_at: datetime = Field(..., description="node start time")
+    finished_at: datetime | None = Field(default=None, description="node finish time")


 class NodeRunExceptionEvent(GraphNodeEventBase):
    error: str = Field(..., description="error")
    start_at: datetime = Field(..., description="node start time")
+    finished_at: datetime | None = Field(default=None, description="node finish time")


 class NodeRunRetryEvent(NodeRunStartedEvent):
@@ -406,11 +406,13 @@ class Node(Generic[NodeDataT]):
                error=str(e),
                error_type="WorkflowNodeError",
            )
+            finished_at = naive_utc_now()
            yield NodeRunFailedEvent(
                id=self.execution_id,
                node_id=self._node_id,
                node_type=self.node_type,
                start_at=self._start_at,
+                finished_at=finished_at,
                node_run_result=result,
                error=str(e),
            )
@@ -568,6 +570,7 @@ class Node(Generic[NodeDataT]):
        return self._node_data

    def _convert_node_run_result_to_graph_node_event(self, result: NodeRunResult) -> GraphNodeEventBase:
+        finished_at = naive_utc_now()
        match result.status:
            case WorkflowNodeExecutionStatus.FAILED:
                return NodeRunFailedEvent(
@@ -575,6 +578,7 @@ class Node(Generic[NodeDataT]):
                    node_id=self.id,
                    node_type=self.node_type,
                    start_at=self._start_at,
+                    finished_at=finished_at,
                    node_run_result=result,
                    error=result.error,
                )
@@ -584,6 +588,7 @@ class Node(Generic[NodeDataT]):
                    node_id=self.id,
                    node_type=self.node_type,
                    start_at=self._start_at,
+                    finished_at=finished_at,
                    node_run_result=result,
                )
            case _:
@@ -606,6 +611,7 @@ class Node(Generic[NodeDataT]):

    @_dispatch.register
    def _(self, event: StreamCompletedEvent) -> NodeRunSucceededEvent | NodeRunFailedEvent:
+        finished_at = naive_utc_now()
        match event.node_run_result.status:
            case WorkflowNodeExecutionStatus.SUCCEEDED:
                return NodeRunSucceededEvent(
@@ -613,6 +619,7 @@ class Node(Generic[NodeDataT]):
                    node_id=self._node_id,
                    node_type=self.node_type,
                    start_at=self._start_at,
+                    finished_at=finished_at,
                    node_run_result=event.node_run_result,
                )
            case WorkflowNodeExecutionStatus.FAILED:
@@ -621,6 +628,7 @@ class Node(Generic[NodeDataT]):
                    node_id=self._node_id,
                    node_type=self.node_type,
                    start_at=self._start_at,
+                    finished_at=finished_at,
                    node_run_result=event.node_run_result,
                    error=event.node_run_result.error,
                )
@@ -12,6 +12,7 @@ def build_http_request_config(
    max_text_size: int = 1 * 1024 * 1024,
    ssl_verify: bool = True,
    ssrf_default_max_retries: int = 3,
+    secret_key: str = "",
 ) -> HttpRequestNodeConfig:
    return HttpRequestNodeConfig(
        max_connect_timeout=max_connect_timeout,
@@ -21,6 +22,7 @@ def build_http_request_config(
        max_text_size=max_text_size,
        ssl_verify=ssl_verify,
        ssrf_default_max_retries=ssrf_default_max_retries,
+        secret_key=secret_key,
    )


@@ -76,6 +76,7 @@ class HttpRequestNodeConfig:
    max_text_size: int
    ssl_verify: bool
    ssrf_default_max_retries: int
+    secret_key: str = ""

    def default_timeout(self) -> "HttpRequestNodeTimeout":
        return HttpRequestNodeTimeout(
@@ -1,3 +1,11 @@
+"""HTTP request execution helpers for workflow nodes.
+
+Besides normal request assembly, this executor is responsible for propagating
+workflow recursion depth across outbound HTTP calls. The reserved call-depth
+headers are always regenerated from the current node context so user-supplied
+values cannot override or poison the propagation contract.
+"""
+
 import base64
 import json
 import secrets
@@ -10,6 +18,8 @@ from urllib.parse import urlencode, urlparse
 import httpx
 from json_repair import repair_json

+from dify_graph.call_depth import build_workflow_call_depth_signature
+from dify_graph.constants import WORKFLOW_CALL_DEPTH_HEADER, WORKFLOW_CALL_DEPTH_SIGNATURE_HEADER
 from dify_graph.file.enums import FileTransferMethod
 from dify_graph.runtime import VariablePool
 from dify_graph.variables.segments import ArrayFileSegment, FileSegment
@@ -41,6 +51,8 @@ BODY_TYPE_TO_CONTENT_TYPE = {


 class Executor:
+    """Prepare, execute, and log a workflow HTTP request node invocation."""
+
    method: Literal[
        "get",
        "head",
@@ -77,6 +89,7 @@ class Executor:
        timeout: HttpRequestNodeTimeout,
        variable_pool: VariablePool,
        http_request_config: HttpRequestNodeConfig,
+        workflow_call_depth: int = 0,
        max_retries: int | None = None,
        ssl_verify: bool | None = None,
        http_client: HttpClientProtocol,
@@ -120,6 +133,7 @@ class Executor:
        # init template
        self.variable_pool = variable_pool
        self.node_data = node_data
+        self.workflow_call_depth = workflow_call_depth
        self._initialize()

    def _initialize(self):
@@ -272,8 +286,33 @@ class Executor:
                    self.data = form_data

    def _assembling_headers(self) -> dict[str, Any]:
+        """Assemble outbound headers for the request.
+
+        Reserved workflow call-depth headers are removed case-insensitively
+        before the canonical pair is re-added from ``workflow_call_depth``.
+        The signature path mirrors Flask request matching, so URLs without an
+        explicit path are normalized to ``/`` before signing. This keeps
+        propagation deterministic even if a workflow author manually configured
+        colliding headers on the node.
+        """
        authorization = deepcopy(self.auth)
        headers = deepcopy(self.headers) or {}
+        reserved_header_names = {
+            WORKFLOW_CALL_DEPTH_HEADER.lower(),
+            WORKFLOW_CALL_DEPTH_SIGNATURE_HEADER.lower(),
+        }
+        headers = {k: v for k, v in headers.items() if k.lower() not in reserved_header_names}
+        parsed_url = urlparse(self.url)
+        signed_path = parsed_url.path or "/"
+        next_call_depth = str(self.workflow_call_depth + 1)
+        headers[WORKFLOW_CALL_DEPTH_HEADER] = next_call_depth
+        headers[WORKFLOW_CALL_DEPTH_SIGNATURE_HEADER] = build_workflow_call_depth_signature(
+            secret_key=self._http_request_config.secret_key,
+            method=self.method,
+            path=signed_path,
+            depth=next_call_depth,
+        )
+
        if self.auth.type == "api-key":
            if self.auth.config is None:
                raise AuthorizationConfigError("self.authorization config is required")
@@ -388,6 +427,12 @@ class Executor:
        return self._validate_and_parse_response(response)

    def to_log(self):
+        """Render the request in raw HTTP form for node logs.
+
+        Internal workflow call-depth headers and authentication headers are
+        masked so operational logs remain useful without exposing replayable or
+        credential-bearing values.
+        """
        url_parts = urlparse(self.url)
        path = url_parts.path or "/"

@@ -410,6 +455,12 @@ class Executor:
            if body.type == "form-data":
                headers["Content-Type"] = f"multipart/form-data; boundary={boundary}"
        for k, v in headers.items():
+            if k.lower() in {
+                WORKFLOW_CALL_DEPTH_HEADER.lower(),
+                WORKFLOW_CALL_DEPTH_SIGNATURE_HEADER.lower(),
+            }:
+                raw += f"{k}: [internal]\r\n"
+                continue
            if self.auth.type == "api-key":
                authorization_header = "Authorization"
                if self.auth.config and self.auth.config.header:
@@ -101,7 +101,7 @@ class HttpRequestNode(Node[HttpRequestNodeData]):
                timeout=self._get_request_timeout(self.node_data),
                variable_pool=self.graph_runtime_state.variable_pool,
                http_request_config=self._http_request_config,
-                max_retries=0,
+                workflow_call_depth=self.workflow_call_depth,
                ssl_verify=self.node_data.ssl_verify,
                http_client=self._http_client,
                file_manager=self._file_manager,
@@ -236,7 +236,7 @@ class IterationNode(LLMUsageTrackingMixin, Node[IterationNodeData]):
            future_to_index: dict[
                Future[
                    tuple[
-                        datetime,
+                        float,
                        list[GraphNodeEventBase],
                        object | None,
                        dict[str, Variable],
@@ -261,7 +261,7 @@ class IterationNode(LLMUsageTrackingMixin, Node[IterationNodeData]):
                try:
                    result = future.result()
                    (
-                        iter_start_at,
+                        iteration_duration,
                        events,
                        output_value,
                        conversation_snapshot,
@@ -274,8 +274,9 @@ class IterationNode(LLMUsageTrackingMixin, Node[IterationNodeData]):
                    # Yield all events from this iteration
                    yield from events

-                    # Update tokens and timing
-                    iter_run_map[str(index)] = (datetime.now(UTC).replace(tzinfo=None) - iter_start_at).total_seconds()
+                    # The worker computes duration before we replay buffered events here,
+                    # so slow downstream consumers don't inflate per-iteration timing.
+                    iter_run_map[str(index)] = iteration_duration

                    usage_accumulator[0] = self._merge_usage(usage_accumulator[0], iteration_usage)

@@ -305,7 +306,7 @@ class IterationNode(LLMUsageTrackingMixin, Node[IterationNodeData]):
        index: int,
        item: object,
        execution_context: "IExecutionContext",
-    ) -> tuple[datetime, list[GraphNodeEventBase], object | None, dict[str, Variable], LLMUsage]:
+    ) -> tuple[float, list[GraphNodeEventBase], object | None, dict[str, Variable], LLMUsage]:
        """Execute a single iteration in parallel mode and return results."""
        with execution_context:
            iter_start_at = datetime.now(UTC).replace(tzinfo=None)
@@ -327,9 +328,10 @@ class IterationNode(LLMUsageTrackingMixin, Node[IterationNodeData]):
            conversation_snapshot = self._extract_conversation_variable_snapshot(
                variable_pool=graph_engine.graph_runtime_state.variable_pool
            )
+            iteration_duration = (datetime.now(UTC).replace(tzinfo=None) - iter_start_at).total_seconds()

            return (
-                iter_start_at,
+                iteration_duration,
                events,
                output_value,
                conversation_snapshot,
@@ -256,9 +256,13 @@ def fetch_prompt_messages(
                ):
                    continue
                prompt_message_content.append(content_item)
-            if prompt_message_content:
+            if not prompt_message_content:
+                continue
+            if len(prompt_message_content) == 1 and prompt_message_content[0].type == PromptMessageContentType.TEXT:
+                prompt_message.content = prompt_message_content[0].data
+            else:
                prompt_message.content = prompt_message_content
-                filtered_prompt_messages.append(prompt_message)
+            filtered_prompt_messages.append(prompt_message)
        elif not prompt_message.is_empty():
            filtered_prompt_messages.append(prompt_message)

@@ -3,6 +3,7 @@ import logging
 import time

 import click
+from sqlalchemy import select
 from werkzeug.exceptions import NotFound

 from core.indexing_runner import DocumentIsPausedError, IndexingRunner
@@ -10,6 +11,7 @@ from events.document_index_event import document_index_created
 from extensions.ext_database import db
 from libs.datetime_utils import naive_utc_now
 from models.dataset import Document
+from models.enums import IndexingStatus

 logger = logging.getLogger(__name__)

@@ -23,19 +25,17 @@ def handle(sender, **kwargs):
    for document_id in document_ids:
        logger.info(click.style(f"Start process document: {document_id}", fg="green"))

-        document = (
-            db.session.query(Document)
-            .where(
+        document = db.session.scalar(
+            select(Document).where(
                Document.id == document_id,
                Document.dataset_id == dataset_id,
            )
-            .first()
        )

        if not document:
            raise NotFound("Document not found")

-        document.indexing_status = "parsing"
+        document.indexing_status = IndexingStatus.PARSING
        document.processing_started_at = naive_utc_now()
        documents.append(document)
        db.session.add(document)
@@ -1,6 +1,6 @@
 from typing import Any, cast

-from sqlalchemy import select
+from sqlalchemy import delete, select

 from events.app_event import app_model_config_was_updated
 from extensions.ext_database import db
@@ -31,9 +31,9 @@ def handle(sender, **kwargs):

    if removed_dataset_ids:
        for dataset_id in removed_dataset_ids:
-            db.session.query(AppDatasetJoin).where(
-                AppDatasetJoin.app_id == app.id, AppDatasetJoin.dataset_id == dataset_id
-            ).delete()
+            db.session.execute(
+                delete(AppDatasetJoin).where(AppDatasetJoin.app_id == app.id, AppDatasetJoin.dataset_id == dataset_id)
+            )

    if added_dataset_ids:
        for dataset_id in added_dataset_ids:
@@ -1,6 +1,6 @@
 from typing import cast

-from sqlalchemy import select
+from sqlalchemy import delete, select

 from core.workflow.nodes.knowledge_retrieval.entities import KnowledgeRetrievalNodeData
 from dify_graph.nodes import BuiltinNodeTypes
@@ -31,9 +31,9 @@ def handle(sender, **kwargs):

    if removed_dataset_ids:
        for dataset_id in removed_dataset_ids:
-            db.session.query(AppDatasetJoin).where(
-                AppDatasetJoin.app_id == app.id, AppDatasetJoin.dataset_id == dataset_id
-            ).delete()
+            db.session.execute(
+                delete(AppDatasetJoin).where(AppDatasetJoin.app_id == app.id, AppDatasetJoin.dataset_id == dataset_id)
+            )

    if added_dataset_ids:
        for dataset_id in added_dataset_ids:
@@ -3,6 +3,7 @@ import json
 import flask_login
 from flask import Response, request
 from flask_login import user_loaded_from_request, user_logged_in
+from sqlalchemy import select
 from werkzeug.exceptions import NotFound, Unauthorized

 from configs import dify_config
@@ -34,16 +35,15 @@ def load_user_from_request(request_from_flask_login):
        if admin_api_key and admin_api_key == auth_token:
            workspace_id = request.headers.get("X-WORKSPACE-ID")
            if workspace_id:
-                tenant_account_join = (
-                    db.session.query(Tenant, TenantAccountJoin)
+                tenant_account_join = db.session.execute(
+                    select(Tenant, TenantAccountJoin)
                    .where(Tenant.id == workspace_id)
                    .where(TenantAccountJoin.tenant_id == Tenant.id)
                    .where(TenantAccountJoin.role == "owner")
-                    .one_or_none()
-                )
+                ).one_or_none()
                if tenant_account_join:
                    tenant, ta = tenant_account_join
-                    account = db.session.query(Account).filter_by(id=ta.account_id).first()
+                    account = db.session.scalar(select(Account).where(Account.id == ta.account_id))
                    if account:
                        account.current_tenant = tenant
                        return account
@@ -70,7 +70,7 @@ def load_user_from_request(request_from_flask_login):
            end_user_id = decoded.get("end_user_id")
            if not end_user_id:
                raise Unauthorized("Invalid Authorization token.")
-            end_user = db.session.query(EndUser).where(EndUser.id == end_user_id).first()
+            end_user = db.session.scalar(select(EndUser).where(EndUser.id == end_user_id))
            if not end_user:
                raise NotFound("End user not found.")
            return end_user
@@ -80,7 +80,7 @@ def load_user_from_request(request_from_flask_login):
            decoded = PassportService().verify(auth_token)
            end_user_id = decoded.get("end_user_id")
            if end_user_id:
-                end_user = db.session.query(EndUser).where(EndUser.id == end_user_id).first()
+                end_user = db.session.scalar(select(EndUser).where(EndUser.id == end_user_id))
                if not end_user:
                    raise NotFound("End user not found.")
                return end_user
@@ -90,11 +90,11 @@ def load_user_from_request(request_from_flask_login):
        server_code = request.view_args.get("server_code") if request.view_args else None
        if not server_code:
            raise Unauthorized("Invalid Authorization token.")
-        app_mcp_server = db.session.query(AppMCPServer).where(AppMCPServer.server_code == server_code).first()
+        app_mcp_server = db.session.scalar(select(AppMCPServer).where(AppMCPServer.server_code == server_code).limit(1))
        if not app_mcp_server:
            raise NotFound("App MCP server not found.")
-        end_user = (
-            db.session.query(EndUser).where(EndUser.session_id == app_mcp_server.id, EndUser.type == "mcp").first()
+        end_user = db.session.scalar(
+            select(EndUser).where(EndUser.session_id == app_mcp_server.id, EndUser.type == "mcp").limit(1)
        )
        if not end_user:
            raise NotFound("End user not found.")
@@ -32,7 +32,7 @@ class OpenDALStorage(BaseStorage):
        kwargs = kwargs or _get_opendal_kwargs(scheme=scheme)

        if scheme == "fs":
-            root = kwargs.get("root", "storage")
+            root = kwargs.setdefault("root", "storage")
            Path(root).mkdir(parents=True, exist_ok=True)

        retry_layer = opendal.layers.RetryLayer(max_times=3, factor=2.0, jitter=True)
@@ -424,13 +424,11 @@ def _build_from_datasource_file(
    datasource_file_id = mapping.get("datasource_file_id")
    if not datasource_file_id:
        raise ValueError(f"DatasourceFile {datasource_file_id} not found")
-    datasource_file = (
-        db.session.query(UploadFile)
-        .where(
+    datasource_file = db.session.scalar(
+        select(UploadFile).where(
            UploadFile.id == datasource_file_id,
            UploadFile.tenant_id == tenant_id,
        )
-        .first()
    )

    if datasource_file is None:
@@ -177,13 +177,11 @@ class Account(UserMixin, TypeBase):

    @classmethod
    def get_by_openid(cls, provider: str, open_id: str):
-        account_integrate = (
-            db.session.query(AccountIntegrate)
-            .where(AccountIntegrate.provider == provider, AccountIntegrate.open_id == open_id)
-            .one_or_none()
-        )
+        account_integrate = db.session.execute(
+            select(AccountIntegrate).where(AccountIntegrate.provider == provider, AccountIntegrate.open_id == open_id)
+        ).scalar_one_or_none()
        if account_integrate:
-            return db.session.query(Account).where(Account.id == account_integrate.account_id).one_or_none()
+            return db.session.scalar(select(Account).where(Account.id == account_integrate.account_id))
        return None

    # check current_user.current_tenant.current_role in ['admin', 'owner']
@@ -8,6 +8,7 @@ import os
 import pickle
 import re
 import time
+from collections.abc import Sequence
 from datetime import datetime
 from json import JSONDecodeError
 from typing import Any, TypedDict, cast
@@ -30,7 +31,20 @@ from services.entities.knowledge_entities.knowledge_entities import ParentMode,
 from .account import Account
 from .base import Base, TypeBase
 from .engine import db
-from .enums import CreatorUserRole
+from .enums import (
+    CollectionBindingType,
+    CreatorUserRole,
+    DatasetMetadataType,
+    DatasetQuerySource,
+    DatasetRuntimeMode,
+    DataSourceType,
+    DocumentCreatedFrom,
+    DocumentDocType,
+    IndexingStatus,
+    ProcessRuleMode,
+    SegmentStatus,
+    SummaryStatus,
+)
 from .model import App, Tag, TagBinding, UploadFile
 from .types import AdjustedJSON, BinaryData, EnumText, LongText, StringUUID, adjusted_json_index

@@ -120,7 +134,7 @@ class Dataset(Base):
        server_default=sa.text("'only_me'"),
        default=DatasetPermissionEnum.ONLY_ME,
    )
-    data_source_type = mapped_column(String(255))
+    data_source_type = mapped_column(EnumText(DataSourceType, length=255))
    indexing_technique: Mapped[str | None] = mapped_column(String(255))
    index_struct = mapped_column(LongText, nullable=True)
    created_by = mapped_column(StringUUID, nullable=False)
@@ -137,7 +151,9 @@ class Dataset(Base):
    summary_index_setting = mapped_column(AdjustedJSON, nullable=True)
    built_in_field_enabled = mapped_column(sa.Boolean, nullable=False, server_default=sa.text("false"))
    icon_info = mapped_column(AdjustedJSON, nullable=True)
-    runtime_mode = mapped_column(sa.String(255), nullable=True, server_default=sa.text("'general'"))
+    runtime_mode = mapped_column(
+        EnumText(DatasetRuntimeMode, length=255), nullable=True, server_default=sa.text("'general'")
+    )
    pipeline_id = mapped_column(StringUUID, nullable=True)
    chunk_structure = mapped_column(sa.String(255), nullable=True)
    enable_api = mapped_column(sa.Boolean, nullable=False, server_default=sa.text("true"))
@@ -145,30 +161,25 @@ class Dataset(Base):

    @property
    def total_documents(self):
-        return db.session.query(func.count(Document.id)).where(Document.dataset_id == self.id).scalar()
+        return db.session.scalar(select(func.count(Document.id)).where(Document.dataset_id == self.id)) or 0

    @property
    def total_available_documents(self):
        return (
-            db.session.query(func.count(Document.id))
-            .where(
-                Document.dataset_id == self.id,
-                Document.indexing_status == "completed",
-                Document.enabled == True,
-                Document.archived == False,
+            db.session.scalar(
+                select(func.count(Document.id)).where(
+                    Document.dataset_id == self.id,
+                    Document.indexing_status == "completed",
+                    Document.enabled == True,
+                    Document.archived == False,
+                )
            )
-            .scalar()
+            or 0
        )

    @property
    def dataset_keyword_table(self):
-        dataset_keyword_table = (
-            db.session.query(DatasetKeywordTable).where(DatasetKeywordTable.dataset_id == self.id).first()
-        )
-        if dataset_keyword_table:
-            return dataset_keyword_table
-
-        return None
+        return db.session.scalar(select(DatasetKeywordTable).where(DatasetKeywordTable.dataset_id == self.id))

    @property
    def index_struct_dict(self):
@@ -195,64 +206,66 @@ class Dataset(Base):

    @property
    def latest_process_rule(self):
-        return (
-            db.session.query(DatasetProcessRule)
+        return db.session.scalar(
+            select(DatasetProcessRule)
            .where(DatasetProcessRule.dataset_id == self.id)
            .order_by(DatasetProcessRule.created_at.desc())
-            .first()
+            .limit(1)
        )

    @property
    def app_count(self):
        return (
-            db.session.query(func.count(AppDatasetJoin.id))
-            .where(AppDatasetJoin.dataset_id == self.id, App.id == AppDatasetJoin.app_id)
-            .scalar()
+            db.session.scalar(
+                select(func.count(AppDatasetJoin.id)).where(
+                    AppDatasetJoin.dataset_id == self.id, App.id == AppDatasetJoin.app_id
+                )
+            )
+            or 0
        )

    @property
    def document_count(self):
-        return db.session.query(func.count(Document.id)).where(Document.dataset_id == self.id).scalar()
+        return db.session.scalar(select(func.count(Document.id)).where(Document.dataset_id == self.id)) or 0

    @property
    def available_document_count(self):
        return (
-            db.session.query(func.count(Document.id))
-            .where(
-                Document.dataset_id == self.id,
-                Document.indexing_status == "completed",
-                Document.enabled == True,
-                Document.archived == False,
+            db.session.scalar(
+                select(func.count(Document.id)).where(
+                    Document.dataset_id == self.id,
+                    Document.indexing_status == "completed",
+                    Document.enabled == True,
+                    Document.archived == False,
+                )
            )
-            .scalar()
+            or 0
        )

    @property
    def available_segment_count(self):
        return (
-            db.session.query(func.count(DocumentSegment.id))
-            .where(
-                DocumentSegment.dataset_id == self.id,
-                DocumentSegment.status == "completed",
-                DocumentSegment.enabled == True,
+            db.session.scalar(
+                select(func.count(DocumentSegment.id)).where(
+                    DocumentSegment.dataset_id == self.id,
+                    DocumentSegment.status == "completed",
+                    DocumentSegment.enabled == True,
+                )
            )
-            .scalar()
+            or 0
        )

    @property
    def word_count(self):
-        return (
-            db.session.query(Document)
-            .with_entities(func.coalesce(func.sum(Document.word_count), 0))
-            .where(Document.dataset_id == self.id)
-            .scalar()
+        return db.session.scalar(
+            select(func.coalesce(func.sum(Document.word_count), 0)).where(Document.dataset_id == self.id)
        )

    @property
    def doc_form(self) -> str | None:
        if self.chunk_structure:
            return self.chunk_structure
-        document = db.session.query(Document).where(Document.dataset_id == self.id).first()
+        document = db.session.scalar(select(Document).where(Document.dataset_id == self.id).limit(1))
        if document:
            return document.doc_form
        return None
@@ -270,8 +283,8 @@ class Dataset(Base):

    @property
    def tags(self):
-        tags = (
-            db.session.query(Tag)
+        tags = db.session.scalars(
+            select(Tag)
            .join(TagBinding, Tag.id == TagBinding.tag_id)
            .where(
                TagBinding.target_id == self.id,
@@ -279,8 +292,7 @@ class Dataset(Base):
                Tag.tenant_id == self.tenant_id,
                Tag.type == "knowledge",
            )
-            .all()
-        )
+        ).all()

        return tags or []

@@ -288,8 +300,8 @@ class Dataset(Base):
    def external_knowledge_info(self):
        if self.provider != "external":
            return None
-        external_knowledge_binding = (
-            db.session.query(ExternalKnowledgeBindings).where(ExternalKnowledgeBindings.dataset_id == self.id).first()
+        external_knowledge_binding = db.session.scalar(
+            select(ExternalKnowledgeBindings).where(ExternalKnowledgeBindings.dataset_id == self.id)
        )
        if not external_knowledge_binding:
            return None
@@ -310,7 +322,7 @@ class Dataset(Base):
    @property
    def is_published(self):
        if self.pipeline_id:
-            pipeline = db.session.query(Pipeline).where(Pipeline.id == self.pipeline_id).first()
+            pipeline = db.session.scalar(select(Pipeline).where(Pipeline.id == self.pipeline_id))
            if pipeline:
                return pipeline.is_published
        return False
@@ -382,7 +394,7 @@ class DatasetProcessRule(Base):  # bug

    id = mapped_column(StringUUID, nullable=False, default=lambda: str(uuid4()))
    dataset_id = mapped_column(StringUUID, nullable=False)
-    mode = mapped_column(String(255), nullable=False, server_default=sa.text("'automatic'"))
+    mode = mapped_column(EnumText(ProcessRuleMode, length=255), nullable=False, server_default=sa.text("'automatic'"))
    rules = mapped_column(LongText, nullable=True)
    created_by = mapped_column(StringUUID, nullable=False)
    created_at: Mapped[datetime] = mapped_column(DateTime, nullable=False, server_default=func.current_timestamp())
@@ -428,12 +440,12 @@ class Document(Base):
    tenant_id = mapped_column(StringUUID, nullable=False)
    dataset_id = mapped_column(StringUUID, nullable=False)
    position: Mapped[int] = mapped_column(sa.Integer, nullable=False)
-    data_source_type: Mapped[str] = mapped_column(String(255), nullable=False)
+    data_source_type: Mapped[str] = mapped_column(EnumText(DataSourceType, length=255), nullable=False)
    data_source_info = mapped_column(LongText, nullable=True)
    dataset_process_rule_id = mapped_column(StringUUID, nullable=True)
    batch: Mapped[str] = mapped_column(String(255), nullable=False)
    name: Mapped[str] = mapped_column(String(255), nullable=False)
-    created_from: Mapped[str] = mapped_column(String(255), nullable=False)
+    created_from: Mapped[str] = mapped_column(EnumText(DocumentCreatedFrom, length=255), nullable=False)
    created_by = mapped_column(StringUUID, nullable=False)
    created_api_request_id = mapped_column(StringUUID, nullable=True)
    created_at: Mapped[datetime] = mapped_column(DateTime, nullable=False, server_default=func.current_timestamp())
@@ -467,7 +479,9 @@ class Document(Base):
    stopped_at: Mapped[datetime | None] = mapped_column(DateTime, nullable=True)

    # basic fields
-    indexing_status = mapped_column(String(255), nullable=False, server_default=sa.text("'waiting'"))
+    indexing_status = mapped_column(
+        EnumText(IndexingStatus, length=255), nullable=False, server_default=sa.text("'waiting'")
+    )
    enabled: Mapped[bool] = mapped_column(sa.Boolean, nullable=False, server_default=sa.text("true"))
    disabled_at: Mapped[datetime | None] = mapped_column(DateTime, nullable=True)
    disabled_by = mapped_column(StringUUID, nullable=True)
@@ -478,7 +492,7 @@ class Document(Base):
    updated_at: Mapped[datetime] = mapped_column(
        DateTime, nullable=False, server_default=func.current_timestamp(), onupdate=func.current_timestamp()
    )
-    doc_type = mapped_column(String(40), nullable=True)
+    doc_type = mapped_column(EnumText(DocumentDocType, length=40), nullable=True)
    doc_metadata = mapped_column(AdjustedJSON, nullable=True)
    doc_form = mapped_column(String(255), nullable=False, server_default=sa.text("'text_model'"))
    doc_language = mapped_column(String(255), nullable=True)
@@ -521,10 +535,8 @@ class Document(Base):
        if self.data_source_info:
            if self.data_source_type == "upload_file":
                data_source_info_dict: dict[str, Any] = json.loads(self.data_source_info)
-                file_detail = (
-                    db.session.query(UploadFile)
-                    .where(UploadFile.id == data_source_info_dict["upload_file_id"])
-                    .one_or_none()
+                file_detail = db.session.scalar(
+                    select(UploadFile).where(UploadFile.id == data_source_info_dict["upload_file_id"])
                )
                if file_detail:
                    return {
@@ -557,24 +569,23 @@ class Document(Base):

    @property
    def dataset(self):
-        return db.session.query(Dataset).where(Dataset.id == self.dataset_id).one_or_none()
+        return db.session.scalar(select(Dataset).where(Dataset.id == self.dataset_id))

    @property
    def segment_count(self):
-        return db.session.query(DocumentSegment).where(DocumentSegment.document_id == self.id).count()
+        return (
+            db.session.scalar(select(func.count(DocumentSegment.id)).where(DocumentSegment.document_id == self.id)) or 0
+        )

    @property
    def hit_count(self):
-        return (
-            db.session.query(DocumentSegment)
-            .with_entities(func.coalesce(func.sum(DocumentSegment.hit_count), 0))
-            .where(DocumentSegment.document_id == self.id)
-            .scalar()
+        return db.session.scalar(
+            select(func.coalesce(func.sum(DocumentSegment.hit_count), 0)).where(DocumentSegment.document_id == self.id)
        )

    @property
    def uploader(self):
-        user = db.session.query(Account).where(Account.id == self.created_by).first()
+        user = db.session.scalar(select(Account).where(Account.id == self.created_by))
        return user.name if user else None

    @property
@@ -588,14 +599,13 @@ class Document(Base):
    @property
    def doc_metadata_details(self) -> list[DocMetadataDetailItem] | None:
        if self.doc_metadata:
-            document_metadatas = (
-                db.session.query(DatasetMetadata)
+            document_metadatas = db.session.scalars(
+                select(DatasetMetadata)
                .join(DatasetMetadataBinding, DatasetMetadataBinding.metadata_id == DatasetMetadata.id)
                .where(
                    DatasetMetadataBinding.dataset_id == self.dataset_id, DatasetMetadataBinding.document_id == self.id
                )
-                .all()
-            )
+            ).all()
            metadata_list: list[DocMetadataDetailItem] = []
            for metadata in document_metadatas:
                metadata_dict: DocMetadataDetailItem = {
@@ -791,7 +801,7 @@ class DocumentSegment(Base):
    enabled: Mapped[bool] = mapped_column(sa.Boolean, nullable=False, server_default=sa.text("true"))
    disabled_at: Mapped[datetime | None] = mapped_column(DateTime, nullable=True)
    disabled_by = mapped_column(StringUUID, nullable=True)
-    status: Mapped[str] = mapped_column(String(255), server_default=sa.text("'waiting'"))
+    status: Mapped[str] = mapped_column(EnumText(SegmentStatus, length=255), server_default=sa.text("'waiting'"))
    created_by = mapped_column(StringUUID, nullable=False)
    created_at: Mapped[datetime] = mapped_column(DateTime, nullable=False, server_default=func.current_timestamp())
    updated_by = mapped_column(StringUUID, nullable=True)
@@ -826,7 +836,7 @@ class DocumentSegment(Base):
        )

    @property
-    def child_chunks(self) -> list[Any]:
+    def child_chunks(self) -> Sequence[Any]:
        if not self.document:
            return []
        process_rule = self.document.dataset_process_rule
@@ -835,16 +845,13 @@ class DocumentSegment(Base):
            if rules_dict:
                rules = Rule.model_validate(rules_dict)
                if rules.parent_mode and rules.parent_mode != ParentMode.FULL_DOC:
-                    child_chunks = (
-                        db.session.query(ChildChunk)
-                        .where(ChildChunk.segment_id == self.id)
-                        .order_by(ChildChunk.position.asc())
-                        .all()
-                    )
+                    child_chunks = db.session.scalars(
+                        select(ChildChunk).where(ChildChunk.segment_id == self.id).order_by(ChildChunk.position.asc())
+                    ).all()
                    return child_chunks or []
        return []

-    def get_child_chunks(self) -> list[Any]:
+    def get_child_chunks(self) -> Sequence[Any]:
        if not self.document:
            return []
        process_rule = self.document.dataset_process_rule
@@ -853,12 +860,9 @@ class DocumentSegment(Base):
            if rules_dict:
                rules = Rule.model_validate(rules_dict)
                if rules.parent_mode:
-                    child_chunks = (
-                        db.session.query(ChildChunk)
-                        .where(ChildChunk.segment_id == self.id)
-                        .order_by(ChildChunk.position.asc())
-                        .all()
-                    )
+                    child_chunks = db.session.scalars(
+                        select(ChildChunk).where(ChildChunk.segment_id == self.id).order_by(ChildChunk.position.asc())
+                    ).all()
                    return child_chunks or []
        return []

@@ -1007,15 +1011,15 @@ class ChildChunk(Base):

    @property
    def dataset(self):
-        return db.session.query(Dataset).where(Dataset.id == self.dataset_id).first()
+        return db.session.scalar(select(Dataset).where(Dataset.id == self.dataset_id))

    @property
    def document(self):
-        return db.session.query(Document).where(Document.id == self.document_id).first()
+        return db.session.scalar(select(Document).where(Document.id == self.document_id))

    @property
    def segment(self):
-        return db.session.query(DocumentSegment).where(DocumentSegment.id == self.segment_id).first()
+        return db.session.scalar(select(DocumentSegment).where(DocumentSegment.id == self.segment_id))


 class AppDatasetJoin(TypeBase):
@@ -1061,7 +1065,7 @@ class DatasetQuery(TypeBase):
    )
    dataset_id: Mapped[str] = mapped_column(StringUUID, nullable=False)
    content: Mapped[str] = mapped_column(LongText, nullable=False)
-    source: Mapped[str] = mapped_column(String(255), nullable=False)
+    source: Mapped[str] = mapped_column(EnumText(DatasetQuerySource, length=255), nullable=False)
    source_app_id: Mapped[str | None] = mapped_column(StringUUID, nullable=True)
    created_by_role: Mapped[CreatorUserRole] = mapped_column(EnumText(CreatorUserRole, length=255), nullable=False)
    created_by: Mapped[str] = mapped_column(StringUUID, nullable=False)
@@ -1076,7 +1080,7 @@ class DatasetQuery(TypeBase):
            if isinstance(queries, list):
                for query in queries:
                    if query["content_type"] == QueryType.IMAGE_QUERY:
-                        file_info = db.session.query(UploadFile).filter_by(id=query["content"]).first()
+                        file_info = db.session.scalar(select(UploadFile).where(UploadFile.id == query["content"]))
                        if file_info:
                            query["file_info"] = {
                                "id": file_info.id,
@@ -1141,7 +1145,7 @@ class DatasetKeywordTable(TypeBase):
                super().__init__(object_hook=object_hook, *args, **kwargs)

        # get dataset
-        dataset = db.session.query(Dataset).filter_by(id=self.dataset_id).first()
+        dataset = db.session.scalar(select(Dataset).where(Dataset.id == self.dataset_id))
        if not dataset:
            return None
        if self.data_source_type == "database":
@@ -1206,7 +1210,9 @@ class DatasetCollectionBinding(TypeBase):
    )
    provider_name: Mapped[str] = mapped_column(String(255), nullable=False)
    model_name: Mapped[str] = mapped_column(String(255), nullable=False)
-    type: Mapped[str] = mapped_column(String(40), server_default=sa.text("'dataset'"), nullable=False)
+    type: Mapped[str] = mapped_column(
+        EnumText(CollectionBindingType, length=40), server_default=sa.text("'dataset'"), nullable=False
+    )
    collection_name: Mapped[str] = mapped_column(String(64), nullable=False)
    created_at: Mapped[datetime] = mapped_column(
        DateTime, nullable=False, server_default=func.current_timestamp(), init=False
@@ -1433,7 +1439,7 @@ class DatasetMetadata(TypeBase):
    )
    tenant_id: Mapped[str] = mapped_column(StringUUID, nullable=False)
    dataset_id: Mapped[str] = mapped_column(StringUUID, nullable=False)
-    type: Mapped[str] = mapped_column(String(255), nullable=False)
+    type: Mapped[str] = mapped_column(EnumText(DatasetMetadataType, length=255), nullable=False)
    name: Mapped[str] = mapped_column(String(255), nullable=False)
    created_at: Mapped[datetime] = mapped_column(
        DateTime, nullable=False, server_default=sa.func.current_timestamp(), init=False
@@ -1535,7 +1541,7 @@ class PipelineCustomizedTemplate(TypeBase):

    @property
    def created_user_name(self):
-        account = db.session.query(Account).where(Account.id == self.created_by).first()
+        account = db.session.scalar(select(Account).where(Account.id == self.created_by))
        if account:
            return account.name
        return ""
@@ -1570,7 +1576,7 @@ class Pipeline(TypeBase):
    )

    def retrieve_dataset(self, session: Session):
-        return session.query(Dataset).where(Dataset.pipeline_id == self.id).first()
+        return session.scalar(select(Dataset).where(Dataset.pipeline_id == self.id))


 class DocumentPipelineExecutionLog(TypeBase):
@@ -1660,7 +1666,9 @@ class DocumentSegmentSummary(Base):
    summary_index_node_id: Mapped[str] = mapped_column(String(255), nullable=True)
    summary_index_node_hash: Mapped[str] = mapped_column(String(255), nullable=True)
    tokens: Mapped[int | None] = mapped_column(sa.Integer, nullable=True)
-    status: Mapped[str] = mapped_column(String(32), nullable=False, server_default=sa.text("'generating'"))
+    status: Mapped[str] = mapped_column(
+        EnumText(SummaryStatus, length=32), nullable=False, server_default=sa.text("'generating'")
+    )
    error: Mapped[str] = mapped_column(LongText, nullable=True)
    enabled: Mapped[bool] = mapped_column(sa.Boolean, nullable=False, server_default=sa.text("true"))
    disabled_at: Mapped[datetime | None] = mapped_column(DateTime, nullable=True)
@@ -11,6 +11,13 @@ class CreatorUserRole(StrEnum):
    ACCOUNT = "account"
    END_USER = "end_user"

+    @classmethod
+    def _missing_(cls, value):
+        if value == "end-user":
+            return cls.END_USER
+        else:
+            return super()._missing_(value)
+

 class WorkflowRunTriggeredFrom(StrEnum):
    DEBUGGING = "debugging"
@@ -96,3 +103,223 @@ class ConversationStatus(StrEnum):
    """Conversation Status Enum"""

    NORMAL = "normal"
+
+
+class DataSourceType(StrEnum):
+    """Data Source Type for Dataset and Document"""
+
+    UPLOAD_FILE = "upload_file"
+    NOTION_IMPORT = "notion_import"
+    WEBSITE_CRAWL = "website_crawl"
+    LOCAL_FILE = "local_file"
+    ONLINE_DOCUMENT = "online_document"
+
+
+class ProcessRuleMode(StrEnum):
+    """Dataset Process Rule Mode"""
+
+    AUTOMATIC = "automatic"
+    CUSTOM = "custom"
+    HIERARCHICAL = "hierarchical"
+
+
+class IndexingStatus(StrEnum):
+    """Document Indexing Status"""
+
+    WAITING = "waiting"
+    PARSING = "parsing"
+    CLEANING = "cleaning"
+    SPLITTING = "splitting"
+    INDEXING = "indexing"
+    PAUSED = "paused"
+    COMPLETED = "completed"
+    ERROR = "error"
+
+
+class DocumentCreatedFrom(StrEnum):
+    """Document Created From"""
+
+    WEB = "web"
+    API = "api"
+    RAG_PIPELINE = "rag-pipeline"
+
+
+class ConversationFromSource(StrEnum):
+    """Conversation / Message from_source"""
+
+    API = "api"
+    CONSOLE = "console"
+
+
+class FeedbackFromSource(StrEnum):
+    """MessageFeedback from_source"""
+
+    USER = "user"
+    ADMIN = "admin"
+
+
+class FeedbackRating(StrEnum):
+    """MessageFeedback rating"""
+
+    LIKE = "like"
+    DISLIKE = "dislike"
+
+
+class InvokeFrom(StrEnum):
+    """How a conversation/message was invoked"""
+
+    SERVICE_API = "service-api"
+    WEB_APP = "web-app"
+    TRIGGER = "trigger"
+    EXPLORE = "explore"
+    DEBUGGER = "debugger"
+    PUBLISHED_PIPELINE = "published"
+    VALIDATION = "validation"
+
+    @classmethod
+    def value_of(cls, value: str) -> "InvokeFrom":
+        return cls(value)
+
+    def to_source(self) -> str:
+        source_mapping = {
+            InvokeFrom.WEB_APP: "web_app",
+            InvokeFrom.DEBUGGER: "dev",
+            InvokeFrom.EXPLORE: "explore_app",
+            InvokeFrom.TRIGGER: "trigger",
+            InvokeFrom.SERVICE_API: "api",
+        }
+        return source_mapping.get(self, "dev")
+
+
+class DocumentDocType(StrEnum):
+    """Document doc_type classification"""
+
+    BOOK = "book"
+    WEB_PAGE = "web_page"
+    PAPER = "paper"
+    SOCIAL_MEDIA_POST = "social_media_post"
+    WIKIPEDIA_ENTRY = "wikipedia_entry"
+    PERSONAL_DOCUMENT = "personal_document"
+    BUSINESS_DOCUMENT = "business_document"
+    IM_CHAT_LOG = "im_chat_log"
+    SYNCED_FROM_NOTION = "synced_from_notion"
+    SYNCED_FROM_GITHUB = "synced_from_github"
+    OTHERS = "others"
+
+
+class TagType(StrEnum):
+    """Tag type"""
+
+    KNOWLEDGE = "knowledge"
+    APP = "app"
+
+
+class DatasetMetadataType(StrEnum):
+    """Dataset metadata value type"""
+
+    STRING = "string"
+    NUMBER = "number"
+    TIME = "time"
+
+
+class SegmentStatus(StrEnum):
+    """Document segment status"""
+
+    WAITING = "waiting"
+    INDEXING = "indexing"
+    COMPLETED = "completed"
+    ERROR = "error"
+    PAUSED = "paused"
+    RE_SEGMENT = "re_segment"
+
+
+class DatasetRuntimeMode(StrEnum):
+    """Dataset runtime mode"""
+
+    GENERAL = "general"
+    RAG_PIPELINE = "rag_pipeline"
+
+
+class CollectionBindingType(StrEnum):
+    """Dataset collection binding type"""
+
+    DATASET = "dataset"
+    ANNOTATION = "annotation"
+
+
+class DatasetQuerySource(StrEnum):
+    """Dataset query source"""
+
+    HIT_TESTING = "hit_testing"
+    APP = "app"
+
+
+class TidbAuthBindingStatus(StrEnum):
+    """TiDB auth binding status"""
+
+    CREATING = "CREATING"
+    ACTIVE = "ACTIVE"
+
+
+class MessageFileBelongsTo(StrEnum):
+    """MessageFile belongs_to"""
+
+    USER = "user"
+    ASSISTANT = "assistant"
+
+
+class CredentialSourceType(StrEnum):
+    """Load balancing credential source type"""
+
+    PROVIDER = "provider"
+    CUSTOM_MODEL = "custom_model"
+
+
+class PaymentStatus(StrEnum):
+    """Provider order payment status"""
+
+    WAIT_PAY = "wait_pay"
+    PAID = "paid"
+    FAILED = "failed"
+    REFUNDED = "refunded"
+
+
+class BannerStatus(StrEnum):
+    """ExporleBanner status"""
+
+    ENABLED = "enabled"
+    DISABLED = "disabled"
+
+
+class SummaryStatus(StrEnum):
+    """Document segment summary status"""
+
+    NOT_STARTED = "not_started"
+    GENERATING = "generating"
+    COMPLETED = "completed"
+    ERROR = "error"
+    TIMEOUT = "timeout"
+
+
+class MessageChainType(StrEnum):
+    """Message chain type"""
+
+    SYSTEM = "system"
+
+
+class ProviderQuotaType(StrEnum):
+    PAID = "paid"
+    """hosted paid quota"""
+
+    FREE = "free"
+    """third-party free quota"""
+
+    TRIAL = "trial"
+    """hosted trial quota"""
+
+    @staticmethod
+    def value_of(value: str) -> "ProviderQuotaType":
+        for member in ProviderQuotaType:
+            if member.value == value:
+                return member
+        raise ValueError(f"No matching enum found for value '{value}'")
--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
Yanli 盐粒	5574802631	fix: pass webhook request metadata explicitly	2026-03-20 04:51:37 +08:00
Yanli 盐粒	b8cedefd7d	fix: normalize signed call depth for root webhook URLs	2026-03-20 04:27:22 +08:00
Yanli 盐粒	4ecba5858b	fix: propagate workflow call depth through HTTP recursion	2026-03-20 04:03:00 +08:00
tmimmanuelandGitHub	5b9cb55c45	refactor: use EnumText for MessageFeedback and MessageFile columns (#33738 )	2026-03-20 01:13:26 +09:00
Tim RenandGitHub	7d19825659	fix(tests): correct keyword arguments in tool provider test constructors (#33767 )	2026-03-20 00:16:44 +09:00
github-actions[bot]GitHubclaude[bot] <41898282+claude[bot]@users.noreply.github.com>yyh	11e1787100	chore(i18n): sync translations with en-US (#33749 ) Co-authored-by: claude[bot] <41898282+claude[bot]@users.noreply.github.com> Co-authored-by: yyh <92089059+lyzno1@users.noreply.github.com>	2026-03-19 22:03:07 +08:00
yyhandGitHub	70a68f0a86	refactor: simplify the scroll area API for sidebar layouts (#33761 )	2026-03-19 19:54:16 +08:00
FFXN GitHubgemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>Copilot Autofix powered by AI	bb1a6f8a57	fix: Add dataset_id filters to the hit_count's subqueries (#33757 ) Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>	2026-03-19 19:56:31 +09:00
Coding On Star GitHub CodingOnStarautofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	4df602684b	test(workflow): add unit tests for workflow components (#33741 ) Co-authored-by: CodingOnStar <hanxujiang@dify.com> Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	2026-03-19 18:35:16 +08:00
盐粒 YanliandGitHub	df0ded210f	fix: preserve timing metrics in parallel iteration (#33216 )	2026-03-19 18:05:52 +08:00
Sean SunandGitHub	2b8823f38d	fix: use RetrievalModel type for retrieval_model field in HitTestingPayload (#33750 )	2026-03-19 18:58:23 +09:00
QuantumGhost GitHubautofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	c93289e93c	fix(api): add `trigger_info` to WorkflowNodeExecutionMetadataKey (#33753 ) Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	2026-03-19 17:56:49 +08:00
yyhandGitHub	8bbaa862f2	style(scroll-bar): align design (#33751 )	2026-03-19 17:51:55 +08:00
-LAN-GitHub Stephen Zhou	7019395a28	chore: add pytest XML and branch coverage reports (#33730 ) Co-authored-by: Stephen Zhou <38493346+hyoban@users.noreply.github.com>	2026-03-19 17:08:34 +08:00
yyhandGitHub	ef9803f8b9	refactor(web): migrate auth toast calls to ui toast (#33744 )	2026-03-19 16:15:07 +08:00
RickDamon GitHub QuantumGhost	a2c4345c00	fix: pass default root to OpenDAL Operator for fs scheme (#33678 ) Co-authored-by: QuantumGhost <obelisk.reg+git@gmail.com>	2026-03-19 15:30:13 +08:00
Renzo GitHubautofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	8a22cc06c9	refactor: migrate db.session.query to select in infra layer (#33694 ) Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	2026-03-19 15:48:29 +09:00
yyhandGitHub	79e5253410	refactor(web): sidebar app list to scroll area component (#33733 )	2026-03-19 14:33:55 +08:00
Bowen LiangandGitHub	942087cbdb	feat(confirm): input app name for app deletion confirmation (#33660 )	2026-03-19 14:25:29 +08:00
tmimmanuelandGitHub	e4f1d3c63a	refactor: use EnumText(StorageType) for UploadFile.storage_type (#33728 )	2026-03-19 15:15:32 +09:00
WillandGitHub	0166cbf122	fix: compatible with emoji/object icons in plugin card icon resolver (#33732 )	2026-03-19 14:06:55 +08:00
Stephen ZhouandGitHub	77b8012fd8	ci: use codecov (#33723 )	2026-03-19 13:24:59 +08:00
yyhandGitHub	f9615b30ab	feat(web): add base ui scroll area primitive (#33727 )	2026-03-19 13:04:04 +08:00
Stephen ZhouandGitHub	454786a9dc	chore: update deps (#33706 )	2026-03-19 10:10:27 +08:00
BitTobyandGitHub	9ff0d9df88	refactor(api): type WaterCrawl API responses with TypedDict (#33700 )	2026-03-19 10:35:44 +09:00
dependabot[bot]GitHubdependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>	9367020bfd	chore(deps): bump pypdf from 6.8.0 to 6.9.1 in /api (#33698 ) Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>	2026-03-19 04:12:54 +09:00
BitTobyandGitHub	b2a388b7bf	refactor(api): type Firecrawl API responses with TypedDict (#33691 )	2026-03-19 04:00:06 +09:00
dependabot[bot]GitHubdependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>	146f8fac45	chore(deps): bump ujson from 5.9.0 to 5.12.0 in /api (#33683 ) Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>	2026-03-19 03:55:49 +09:00
tmimmanuel GitHubautofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	29577cac14	refactor: EnumText for preferred_provider_type MessageChain, Banner (#33696 ) Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	2026-03-19 03:53:04 +09:00
Stephen ZhouandGitHub	4254392221	fix: leaked set timeout (#33692 )	2026-03-18 23:46:22 +08:00
Stephen ZhouandGitHub	5ee0633bac	ci: revert agent reporter (#33685 )	2026-03-18 22:15:48 +08:00
Stephen ZhouandGitHub	37ffa262ae	refactor: move to std-semver (#33682 )	2026-03-18 22:00:25 +08:00
QuantumGhostandGitHub	49256770c9	chore: bump version to 1.13.2 (#33681 )	2026-03-18 21:53:16 +08:00
BitToby GitHubautofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	25ab5e46b3	refactor(api): type default_retrieval_model with DefaultRetrievalModelDict in core/rag (#33676 ) Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	2026-03-18 22:47:51 +09:00
QuantumGhost GitHubgemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	29c70736dc	fix(api): Preserving the content transform logic in fetch_prompt_messages (#33666 ) Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	2026-03-18 19:41:25 +08:00
Desel72andGitHub	7c99c9f3e8	fix: sync workflow description and name to MCP server on update (#33637 )	2026-03-18 19:04:45 +08:00
wangxiaolei GitHub QuantumGhost Copilot Autofix powered by AIautofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	a87b928079	feat: remove weaviate client __del__ method (#33593 ) Co-authored-by: QuantumGhost <obelisk.reg+git@gmail.com> Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com> Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	2026-03-18 17:39:59 +08:00
yyhandGitHub	93f9546353	refactor(web): migrate core toast call sites to base ui toast (#33643 )	2026-03-18 16:53:55 +08:00
Coding On Star GitHub CodingOnStarautofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	db4deb1d6b	test(workflow): reorganize specs into __tests__ and align with shared test infrastructure (#33625 ) Co-authored-by: CodingOnStar <hanxujiang@dify.com> Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	2026-03-18 16:40:28 +08:00
wangxiaolei GitHubautofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	387e5a345f	fix(api): make CreatorUserRole accept both `end-user` and `end_user` (#33638 ) Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	2026-03-18 14:54:12 +08:00
-LAN-andGitHub	116cc22019	fix: clarify webhook debug endpoint behavior (#33597 )	2026-03-18 14:28:33 +08:00
FFXN GitHubgemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>Copilot Autofix powered by AIautofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	dc69f65b4b	fix: add responding error information when obtain pipeline template detail failed (#33628 ) Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com> Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	2026-03-18 13:31:45 +08:00
yyhandGitHub	296b7044af	refactor: route next/navigation through compat re-export (#33636 )	2026-03-18 12:57:03 +08:00
tmimmanuelandGitHub	04c0bf61fa	refactor: use EnumText in provider models (#33634 )	2026-03-18 13:27:40 +09:00
BitTobyandGitHub	3454224ff9	refactor(api): replace dict with SummaryIndexSettingDict TypedDict in core/rag (#33633 )	2026-03-18 13:26:49 +09:00
yyhandGitHub	69d1ccb7a7	refactor: route next/link through compat re-export (#33632 )	2026-03-18 12:14:59 +08:00
Coding On Star GitHub CodingOnStar	6100acb780	refactor(web): move component tests into sibling __tests__ directories (#33623 ) Co-authored-by: CodingOnStar <hanxujiang@dify.com>	2026-03-18 10:49:30 +08:00
BitToby GitHubautofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	485da15a4d	refactor(api): replace dict/Mapping with TypedDict in core/rag retrieval_service.py (#33615 ) Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	2026-03-18 11:49:09 +09:00
yyhandGitHub	d7f70f3c0f	refactor: route low-cost next modules through compat re-exports (#33622 )	2026-03-18 10:37:29 +08:00
yyh GitHub CodingOnStarautofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>Coding On Stardependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>-LAN-statxc	bbe975c6bc	feat: enhance model plugin workflow checks and model provider management UX (#33289 ) Signed-off-by: yyh <yuanyouhuilyz@gmail.com> Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: CodingOnStar <hanxujiang@dify.com> Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com> Co-authored-by: Coding On Star <447357187@qq.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: -LAN- <laipz8200@outlook.com> Co-authored-by: statxc <tyleradams93226@gmail.com>	2026-03-18 10:16:15 +08:00
wangxiaoleiandGitHub	aa4a9877f5	fix: fix max_retries is hardcode (#33619 )	2026-03-18 10:08:16 +08:00
yyhandGitHub	a073d60c12	fix(toast): refine stacked hover and focus styles (#33620 )	2026-03-18 10:01:12 +08:00
Stephen ZhouandGitHub	fa542f6628	chore: update deps (#33618 )	2026-03-18 09:41:04 +08:00
yyhandGitHub	801f8b6e64	fix: tighten toast typing and restore focus visibility (#33591 )	2026-03-18 09:17:53 +08:00
yyhandGitHub	4c5f7efeed	chore(deps): bump next packages to 16.1.7 (#33616 )	2026-03-18 09:04:03 +08:00
tmimmanuel GitHubautofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	3870b2ad2d	refactor: use EnumText for dataset and replace string literals 4 (#33606 ) Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	2026-03-18 09:18:08 +09:00
BitTobyandGitHub	0bc6c3a73e	refactor(api): replace dict/Mapping with TypedDict in core/tools (#33610 )	2026-03-18 08:00:27 +09:00
dependabot[bot]GitHubdependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>	6de605bc27	chore(deps): bump pyasn1 from 0.6.2 to 0.6.3 in /api (#33611 ) Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>	2026-03-18 07:54:44 +09:00
statxcandGitHub	236ae372be	refactor(api): replace dict/Mapping with TypedDict in core/app (#33601 )	2026-03-18 00:38:39 +09:00
tmimmanuel GitHub Claude Opus 4.6	3f1a127735	refactor: replace sa.String with EnumText for mapped_columns (#33547 ) Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-17 23:30:23 +09:00
Renzo GitHubautofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	7757bb5089	refactor(api): Query API to select function_1 (#33565 ) Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	2026-03-17 23:29:16 +09:00
github-actions[bot]GitHubclaude[bot] <41898282+claude[bot]@users.noreply.github.com>yyh	076b297b18	chore(i18n): sync translations with en-US (#33586 ) Co-authored-by: claude[bot] <41898282+claude[bot]@users.noreply.github.com> Co-authored-by: yyh <92089059+lyzno1@users.noreply.github.com>	2026-03-17 19:42:31 +08:00
github-actions[bot]GitHubclaude[bot] <41898282+claude[bot]@users.noreply.github.com>yyh	da2049be76	chore(i18n): sync translations with en-US (#33585 ) Co-authored-by: claude[bot] <41898282+claude[bot]@users.noreply.github.com> Co-authored-by: yyh <92089059+lyzno1@users.noreply.github.com>	2026-03-17 19:34:41 +08:00
yyhandGitHub	7c6d0bedc0	feat(web): add base ui toast (#33569 )	2026-03-17 18:56:36 +08:00
yyh GitHubautofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	3db1ba36e0	refactor(web): number inputs to use Base UI NumberField (#33539 ) Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>	2026-03-17 18:45:02 +08:00