style: auto-format with black + isort + prettier [skip ci-format-check]

Root cause: The /upload/direct/complete endpoint only created an IngestJob
record and dispatched a Celery task. The Asset record was created
asynchronously by the worker, so when the frontend invalidated the query
cache after upload, the asset list refetch returned stale data (the new
asset didn't exist in DB yet).

Fix:
- Create Asset record immediately in complete_direct_upload and upload_asset
  endpoints with status=PROCESSING, so it appears in the library right away
- Worker now finds the existing Asset by storage_key and updates it with
  extracted metadata (duration, dimensions, codec, etc.) instead of creating
  a new record
- Worker falls back to creating a new Asset if none exists (backward compat)
- Added find_by_storage_key to AssetRepository port and implementations
- Frontend invalidateQueries now works correctly since the Asset exists
  when the refetch happens; asset shows '处理中' status until worker finishes

.env.example

@@ -42,3 +42,11 @@ OSS_ENDPOINT=oss-cn-hangzhou.aliyuncs.com
 OSS_ACCESS_KEY_ID=your-access-key-id
 OSS_ACCESS_KEY_SECRET=your-access-key-secret
 OSS_BUCKET_NAME=xiaoxia-autocut
+
+# ==================== CosyVoice 语音合成配置 ====================
+COSYVOICE_API_KEY=your-cosyvoice-api-key
+COSYVOICE_BASE_URL=https://dashscope.aliyuncs.com/api/v1/services/aigc/text2audio
+COSYVOICE_MODEL=cosyvoice-v1
+COSYVOICE_VOICE=longxiaochun
+COSYVOICE_SAMPLE_RATE=22050
+COSYVOICE_FORMAT=mp3

.env.production.example

@@ -1,30 +1,62 @@
 # 生产环境配置模板（实际使用时复制为 .env.production）
+
+# ==================== 基础配置 ====================
+APP_ENV=production
 ENVIRONMENT=production
 DEBUG=false
 USE_IN_MEMORY_DB=false
 LOG_LEVEL=WARNING
 
-# 数据库（必须修改）
+# ==================== 数据库（必须修改）====================
 DATABASE_URL=postgresql://prod_user:CHANGE_THIS_PASSWORD@db-prod:5432/xiaoxia_prod
 
-# Redis（必须修改）
+# ==================== Redis（必须修改）====================
 REDIS_URL=redis://:CHANGE_THIS_PASSWORD@redis-prod:6379/0
+ENABLE_REDIS_SESSIONS=false
 
-# JWT（必须修改，至少 32 字符）
+# ==================== JWT（必须修改，至少 32 字符）====================
 JWT_SECRET_KEY=CHANGE_THIS_TO_A_RANDOM_SECRET_KEY_AT_LEAST_32_CHARS
 
-# SMTP（必须配置）
+# ==================== 邮件（必须配置）====================
+ENABLE_EMAIL_DELIVERY=false
 SMTP_HOST=smtp.gmail.com
 SMTP_PORT=587
-SMTP_USER=your-email@gmail.com
-SMTP_PASSWORD=your-app-specific-password
+SMTP_USER=CHANGE_ME_SMTP_USER
+SMTP_PASSWORD=CHANGE_ME_SMTP_PASSWORD
 SMTP_FROM_EMAIL=noreply@yourdomain.com
+SMTP_FROM_NAME=小虾 SaaS
+SMTP_USE_TLS=true
 
-# 应用配置
-BASE_URL=https://yourdomain.com
+# ==================== 应用配置 ====================
+APP_BASE_URL=https://yourdomain.com
 
-# CORS（修改为实际域名）
-CORS_ORIGINS=["https://yourdomain.com","https://app.yourdomain.com"]
+# ==================== CORS（修改为实际域名，逗号分隔）====================
+CORS_ORIGINS_RAW=https://yourdomain.com,https://app.yourdomain.com
 
-# 监控（可选）
-SENTRY_DSN=https://your-sentry-dsn@sentry.io/project-id
+# ==================== 阿里云 OSS（必须配置）====================
+OSS_ENDPOINT=oss-cn-hangzhou.aliyuncs.com
+OSS_ACCESS_KEY_ID=CHANGE_ME_ACCESS_KEY_ID
+OSS_ACCESS_KEY_SECRET=CHANGE_ME_ACCESS_KEY_SECRET
+OSS_BUCKET_NAME=xiaoxia-autocut
+OSS_DIRECT_UPLOAD_MAX_MB=2000
+OSS_DIRECT_UPLOAD_EXPIRE_SECONDS=900
+
+# ==================== CosyVoice 语音合成（必须配置）====================
+COSYVOICE_API_KEY=CHANGE_ME_COSYVOICE_API_KEY
+COSYVOICE_BASE_URL=https://dashscope.aliyuncs.com/api/v1/services/aigc/text2audio
+COSYVOICE_MODEL=cosyvoice-v1
+COSYVOICE_VOICE=longxiaochun
+COSYVOICE_SAMPLE_RATE=22050
+COSYVOICE_FORMAT=mp3
+
+# ==================== 生成文件 ====================
+GENERATED_FILES_DIR=/app/generated
+GENERATED_FILES_URL_PREFIX=/generated-files
+PUBLIC_API_BASE_URL=https://api.xiaoxiajianji.com
+
+# ==================== Celery ====================
+CELERY_BROKER_URL=redis://:CHANGE_THIS_PASSWORD@redis-prod:6379/0
+CELERY_RESULT_BACKEND=redis://:CHANGE_THIS_PASSWORD@redis-prod:6379/1
+
+# ==================== 监控（可选）====================
+# SENTRY_DSN=https://your-sentry-dsn@sentry.io/project-id

.flake8

@@ -0,0 +1,13 @@
+[flake8]
+max-line-length = 120
+exclude =
+    .git,
+    __pycache__,
+    .venv,
+    venv,
+    node_modules,
+    alembic
+
+per-file-ignores =
+    tests/integration/*:F821
+    tests/unit/*:F821

.gitea/workflows/acr-cleanup.yml

@@ -0,0 +1,172 @@
+name: ACR Cleanup
+
+on:
+  schedule:
+    - cron: '0 19 * * *'  # UTC 19:00 = 北京时间凌晨3:00
+  workflow_dispatch:
+    inputs:
+      pr_sha:
+        description: "PR commit SHA（仅清理指定PR镜像，留空则全量清理）"
+        required: false
+        default: ""
+      dry_run:
+        description: "预览模式（dry-run），不实际删除"
+        required: false
+        default: "true"
+  pull_request_target:
+    types: [closed]
+    branches: [develop, main]
+
+concurrency:
+  group: acr-cleanup-${{ gitea.ref }}
+  cancel-in-progress: false
+
+jobs:
+  cleanup:
+    name: ACR Image Cleanup
+    runs-on: ci-l2
+    timeout-minutes: 20
+    permissions:
+      contents: read
+    env:
+      ACR_REGISTRY: xiaoxia-registry.cn-hangzhou.cr.aliyuncs.com
+      ACR_NAMESPACE: xiaoxiakeji
+      ACR_SERVICE: registry.aliyuncs.com:cn-hangzhou:china:cri-fvec8o9q4mmxrkaa
+      GITEA_URL: https://git.xiaoxiajianji.com
+      GITEA_REPO: xiaoxia/xiaoxia-saas
+    steps:
+      - name: Checkout code
+        shell: sh
+        env:
+          GITHUB_TOKEN: ${{ github.token }}
+        run: |
+          curl -sH "Authorization: token $GITHUB_TOKEN"             "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/step_checkout.sh?ref=${GITHUB_SHA}"             | bash
+
+      # ====== Cron模式：获取staging运行中镜像作为白名单 ======
+      - name: Get staging running images (whitelist)
+        id: protected_images
+        if: gitea.event_name != 'pull_request_target' && !gitea.event.inputs.pr_sha
+        env:
+          STAGING_SSH_HOST: ${{ secrets.STAGING_SSH_HOST }}
+          STAGING_SSH_PORT: ${{ secrets.STAGING_SSH_PORT }}
+          STAGING_SSH_USER: ${{ secrets.STAGING_SSH_USER }}
+          STAGING_SSH_KEY: ${{ secrets.STAGING_SSH_KEY }}
+        run: |
+          set +e
+          echo "获取staging服务器运行中镜像作为白名单..."
+          staging_host="${STAGING_SSH_HOST:-47.98.113.167}"
+          staging_port="${STAGING_SSH_PORT:-22222}"
+          staging_user="${STAGING_SSH_USER:-root}"
+
+          key_path=~/.ssh/id_rsa
+          if [ -n "${STAGING_SSH_KEY:-}" ]; then
+            printf '%s\n' "$STAGING_SSH_KEY" > "$key_path"
+            chmod 600 "$key_path"
+            echo "Using key from STAGING_SSH_KEY secret"
+          else
+            echo "⚠️  STAGING_SSH_KEY not set, skipping whitelist"
+            echo "protected_tags=" >> $GITHUB_OUTPUT
+            exit 0
+          fi
+
+          ssh-keyscan -p "$staging_port" -H "$staging_host" >> ~/.ssh/known_hosts 2>/dev/null
+
+          # 获取所有运行容器的镜像，提取tag部分
+          IMAGES=$(ssh -p "$staging_port" -i "$key_path" -o StrictHostKeyChecking=no \
+            "$staging_user@$staging_host" "docker ps --format '{{.Image}}' 2>/dev/null" 2>/dev/null | grep -v "^$" | sort -u)
+
+          PROTECTED_TAGS=""
+          if [ -n "$IMAGES" ]; then
+            while IFS= read -r img; do
+              # 从完整镜像名中提取tag（最后一个冒号后）
+              tag=$(echo "$img" | rev | cut -d: -f1 | rev)
+              if [ -n "$tag" ] && [ "$tag" != "latest" ] && [ ${#tag} -gt 5 ]; then
+                if [ -z "$PROTECTED_TAGS" ]; then
+                  PROTECTED_TAGS="$tag"
+                else
+                  PROTECTED_TAGS="$PROTECTED_TAGS,$tag"
+                fi
+              fi
+            done <<< "$IMAGES"
+          fi
+
+          echo "staging运行中镜像tag: ${PROTECTED_TAGS:-（无）}"
+          echo "protected_tags=$PROTECTED_TAGS" >> $GITHUB_OUTPUT
+
+      # ====== Docker登录 ======
+      - name: Docker login to ACR
+        env:
+          ACR_USERNAME: ${{ secrets.ACR_USERNAME }}
+          ACR_PASSWORD: ${{ secrets.ACR_PASSWORD }}
+        run: |
+          printf '%s' "$ACR_PASSWORD" | docker login "$ACR_REGISTRY" -u "$ACR_USERNAME" --password-stdin
+
+      # ====== 模式1：PR关闭时清理 ======
+      - name: Cleanup PR images (PR closed)
+        if: gitea.event_name == 'pull_request_target'
+        env:
+          ACR_USERNAME: ${{ secrets.ACR_USERNAME }}
+          ACR_PASSWORD: ${{ secrets.ACR_PASSWORD }}
+          GITEA_TOKEN: ${{ secrets.GITEA_TOKEN }}
+          PR_SHA: ${{ gitea.event.pull_request.head.sha }}
+          PR_NUMBER: ${{ gitea.event.pull_request.number }}
+        run: |
+          echo "============================================"
+          echo "  PR #$PR_NUMBER 已关闭，清理对应镜像"
+          echo "  Head SHA: ${PR_SHA::12}"
+          echo "============================================"
+          echo ""
+          python3 scripts/ci/acr_cleanup.py \
+            --pr-sha "$PR_SHA" \
+            --execute
+
+      # ====== 模式2：Cron全量清理 ======
+      - name: Full cleanup (cron / manual)
+        if: gitea.event_name != 'pull_request_target' && !gitea.event.inputs.pr_sha
+        env:
+          ACR_USERNAME: ${{ secrets.ACR_USERNAME }}
+          ACR_PASSWORD: ${{ secrets.ACR_PASSWORD }}
+          GITEA_TOKEN: ${{ secrets.GITEA_TOKEN }}
+          PROTECTED_TAGS: ${{ steps.protected_images.outputs.protected_tags }}
+          DRY_RUN_INPUT: ${{ gitea.event.inputs.dry_run }}
+        run: |
+          echo "============================================"
+          echo "  ACR 全量清理（${{ gitea.event_name }}）"
+          echo "============================================"
+          echo ""
+
+          # 决定是否dry-run
+          DRY_RUN_FLAG=""
+          if [ "$DRY_RUN_INPUT" = "true" ]; then
+            DRY_RUN_FLAG="--dry-run"
+            echo "模式: 预览模式 (dry-run)"
+          else
+            echo "模式: 执行模式"
+          fi
+          echo ""
+
+          python3 scripts/ci/acr_cleanup.py \
+            --keep 20 \
+            --protected-tags "$PROTECTED_TAGS" \
+            $DRY_RUN_FLAG
+
+      # ====== 模式3：手动指定PR SHA清理 ======
+      - name: Cleanup specific PR image (manual)
+        if: gitea.event_name == 'workflow_dispatch' && gitea.event.inputs.pr_sha
+        env:
+          ACR_USERNAME: ${{ secrets.ACR_USERNAME }}
+          ACR_PASSWORD: ${{ secrets.ACR_PASSWORD }}
+          PR_SHA: ${{ gitea.event.inputs.pr_sha }}
+          DRY_RUN_INPUT: ${{ gitea.event.inputs.dry_run }}
+        run: |
+          echo "手动清理PR镜像: ${PR_SHA::12}"
+          echo ""
+
+          DRY_RUN_FLAG=""
+          if [ "$DRY_RUN_INPUT" = "true" ]; then
+            DRY_RUN_FLAG="--dry-run"
+          fi
+
+          python3 scripts/ci/acr_cleanup.py \
+            --pr-sha "$PR_SHA" \
+            $DRY_RUN_FLAG

.gitea/workflows/auto-merge.yml

@@ -1,21 +0,0 @@
-name: Auto Merge PRs
-
-on:
-  schedule:
-    - cron: '0 */6 * * *'
-  workflow_dispatch:
-
-jobs:
-  auto-merge:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v3
-      
-      - name: Auto merge develop PRs
-        run: |
-          bash scripts/auto_merge_prs.sh develop
-      
-      - name: Auto merge main PRs (release only)
-        run: |
-          bash scripts/auto_merge_prs.sh main

.gitea/workflows/ci-cd.yml

@@ -1,168 +0,0 @@
-name: CI/CD Pipeline
-
-on:
-  pull_request:
-    branches:
-      - main
-      - develop
-
-jobs:
-  validate:
-    name: Validate Code Quality And Tests
-    runs-on: runtime-builder
-    container: localhost:5000/xiaoxia-ci-python:3.12
-
-    steps:
-      - name: Checkout code
-        shell: sh
-        run: |
-          set -eu
-          python - <<'PY'
-          import io, os, tarfile, urllib.request
-          url = f"{os.environ['GITHUB_API_URL']}/repos/{os.environ['GITHUB_REPOSITORY']}/archive/{os.environ['GITHUB_SHA']}.tar.gz"
-          request = urllib.request.Request(url, headers={"Authorization": f"token {os.environ['GITHUB_TOKEN']}"})
-          with urllib.request.urlopen(request, timeout=120) as response:
-              archive = response.read()
-          with tarfile.open(fileobj=io.BytesIO(archive), mode='r:gz') as tar:
-              root_prefix = tar.getmembers()[0].name.split('/', 1)[0] + '/'
-              for member in tar.getmembers():
-                  name = member.name
-                  if name == root_prefix[:-1]:
-                      continue
-                  if name.startswith(root_prefix):
-                      member.name = name[len(root_prefix):]
-                  if member.name:
-                      tar.extract(member, '.')
-          PY
-
-      - name: Verify CI environment
-        shell: sh
-        run: |
-          set -eu
-          python --version
-          python -m pip --version
-          python -m black --version
-          python -m isort --version-number
-          python -m flake8 --version
-          bandit --version
-          pytest --version
-          echo "CI environment is ready"
-
-      - name: Run code quality checks
-        shell: sh
-        run: |
-          set -eu
-          python -m compileall -q alembic apps packages tests scripts
-          python -m black --check alembic apps packages tests scripts
-          python -m isort --check-only alembic apps packages tests scripts
-          python -m flake8 apps packages tests --count --statistics
-
-      - name: Run security scan
-        shell: sh
-        run: |
-          set -eu
-          bandit -r apps packages -q
-
-      - name: Validate release scripts syntax
-        shell: sh
-        run: |
-          set -eu
-          bash -n scripts/backup_postgres.sh
-          bash -n scripts/restore_postgres_plan.sh
-          bash -n scripts/init_production_env.sh
-
-      - name: Validate Alembic migrations
-        shell: sh
-        run: |
-          set -eu
-          DATABASE_URL=postgresql+psycopg://postgres:postgres@localhost:5432/xiaoxia_saas \
-            python -m alembic upgrade head --sql > /tmp/alembic-upgrade.sql
-          test -s /tmp/alembic-upgrade.sql
-          grep -q "Running upgrade" /tmp/alembic-upgrade.sql
-          python scripts/check_schema_metadata.py
-
-      - name: Run tests
-        shell: sh
-        run: |
-          set -eu
-          PYTHONPATH="$PWD/apps/api:$PWD" python -m pytest tests/unit -q
-
-      - name: Build summary
-        if: github.ref == 'refs/heads/develop' || github.ref == 'refs/heads/main'
-        shell: sh
-        run: |
-          set -eu
-          echo "Build completed successfully!"
-          echo "Branch: ${GITHUB_REF_NAME}"
-          echo "Commit: ${GITHUB_SHA}"
-
-  frontend-lint:
-    name: Frontend Lint
-    runs-on: runtime-builder
-
-    steps:
-      - name: Checkout code
-        shell: sh
-        env:
-          GITHUB_TOKEN: ${{ github.token }}
-        run: |
-          set -eu
-          archive_url="${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/archive/${GITHUB_SHA}.tar.gz"
-          wget --header="Authorization: token ${GITHUB_TOKEN}" -O /tmp/repo.tar.gz "$archive_url"
-          tar -xzf /tmp/repo.tar.gz --strip-components=1 -C .
-          rm -f /tmp/repo.tar.gz
-
-      - name: Install dependencies
-        shell: sh
-        run: |
-          set -eu
-          docker run --rm \
-            --pull=never \
-            -v "$PWD:/workspace" \
-            -w /workspace/apps/web \
-            node:20 \
-            sh -lc 'npm ci'
-
-      - name: Run ESLint
-        shell: sh
-        run: |
-          set -eu
-          docker run --rm \
-            --pull=never \
-            -v "$PWD:/workspace" \
-            -w /workspace/apps/web \
-            node:20 \
-            sh -lc 'npx eslint src --ext .ts,.tsx --max-warnings 0'
-
-      - name: Run TypeScript type check
-        shell: sh
-        run: |
-          set -eu
-          docker run --rm \
-            --pull=never \
-            -v "$PWD:/workspace" \
-            -w /workspace/apps/web \
-            node:20 \
-            sh -lc 'npx tsc --noEmit'
-
-      - name: Run Prettier check
-        shell: sh
-        run: |
-          set -eu
-          docker run --rm \
-            --pull=never \
-            -v "$PWD:/workspace" \
-            -w /workspace/apps/web \
-            node:20 \
-            sh -lc 'npx prettier --check "src/**/*.{ts,tsx,css,md}"'
-
-      - name: Run Vitest tests
-        shell: sh
-        run: |
-          set -eu
-          docker run --rm \
-            --pull=never \
-            -v "$PWD:/workspace" \
-            -w /workspace/apps/web \
-            node:20 \
-            sh -lc 'npx vitest run'

.gitea/workflows/ci-failure-monitor.yml

@@ -0,0 +1,78 @@
+name: CI Failure Monitor
+
+on:
+  schedule:
+    - cron: '0 */6 * * *'  # 每6小时检查一次
+  workflow_dispatch:
+    inputs:
+      days:
+        description: '统计最近N天的失败'
+        required: false
+        default: '7'
+      fail_threshold:
+        description: '失败次数阈值'
+        required: false
+        default: '3'
+      fail_rate_threshold:
+        description: '失败率阈值(%)'
+        required: false
+        default: '30'
+
+permissions:
+  contents: read
+
+jobs:
+  monitor:
+    name: CI重复失败检测
+    runs-on: ci-l2
+    timeout-minutes: 10
+
+    steps:
+    - name: Checkout code
+      shell: sh
+      env:
+        GITHUB_TOKEN: ${{ github.token }}
+      run: |
+        curl -sH "Authorization: token $GITHUB_TOKEN" \
+          "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/step_checkout.sh?ref=${GITHUB_SHA}" \
+          | bash
+
+    - name: Record job start time
+      shell: sh
+      run: bash scripts/ci/step_timer_start.sh
+
+    - name: Run failure detection
+      shell: sh
+      env:
+        GITEA_API_TOKEN: ${{ secrets.REVIEW_GITEA_TOKEN }}
+        GITEA_URL: https://git.xiaoxiajianji.com
+        GITEA_REPO: xiaoxia/xiaoxia-saas
+        CI_NOTIFY_WEBHOOK: ${{ secrets.CI_NOTIFY_WEBHOOK }}
+        FAIL_CHECK_DAYS: ${{ inputs.days || 7 }}
+        FAIL_THRESHOLD: ${{ inputs.fail_threshold || 3 }}
+        FAIL_RATE_THRESHOLD: ${{ inputs.fail_rate_threshold || 30 }}
+      run: |
+        set +e
+        python3 scripts/ci/ci_repeated_failure_detector.py
+        EXIT_CODE=$?
+        echo "检测完成，退出码: $EXIT_CODE"
+        # 0=无异常, 1=有警告, 2=有严重问题
+        # 监控脚本永远不fail，避免告警风暴
+        exit 0
+
+    - name: Job duration summary
+      if: always()
+      shell: sh
+      run: bash scripts/ci/step_timer_end.sh
+
+    - name: Report CI trace
+      if: always()
+      shell: sh
+      env:
+        AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
+      run: |
+        STATUS="ok"
+        [ ${{ job.status }} = "success" ] || STATUS="error"
+        START_TIME=""
+        [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
+        python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true

.gitea/workflows/ci-health-daily.yml

@@ -0,0 +1,103 @@
+name: CI Health Daily Report
+on:
+  schedule:
+    - cron: '0 1 * * *'  # UTC 01:00 = 北京时间 09:00
+  workflow_dispatch:
+permissions:
+  contents: read
+jobs:
+  ci-health-report:
+    name: CI健康度每日巡检
+    runs-on: ci-l2
+    timeout-minutes: 15
+    steps:
+    - name: Checkout code
+      shell: sh
+      env:
+        GITHUB_TOKEN: ${{ github.token }}
+      run: |
+        set -eu
+        python3 - <<'PY'
+        import io, os, tarfile, time, urllib.request, urllib.error
+        url = f"{os.environ['GITHUB_API_URL']}/repos/{os.environ['GITHUB_REPOSITORY']}/archive/{os.environ['GITHUB_SHA']}.tar.gz"
+        request = urllib.request.Request(url, headers={"Authorization": f"token {os.environ['GITHUB_TOKEN']}"})
+        last_err = None
+        for attempt in range(5):
+            try:
+                with urllib.request.urlopen(request, timeout=120) as response:
+                    archive = response.read()
+                break
+            except urllib.error.HTTPError as e:
+                last_err = e
+                if e.code >= 500 and attempt < 4:
+                    wait = 2 ** attempt
+                    print(f"Checkout HTTP {e.code}, retrying in {wait}s (attempt {attempt+1}/5)...")
+                    time.sleep(wait)
+                    continue
+                raise
+            except Exception as e:
+                last_err = e
+                if attempt < 4:
+                    wait = 2 ** attempt
+                    print(f"Checkout error: {e}, retrying in {wait}s (attempt {attempt+1}/5)...")
+                    time.sleep(wait)
+                    continue
+                raise
+        else:
+            raise last_err
+        with tarfile.open(fileobj=io.BytesIO(archive), mode='r:gz') as tar:
+            root_prefix = tar.getmembers()[0].name.split('/', 1)[0] + '/'
+            for member in tar.getmembers():
+                name = member.name
+                if name == root_prefix[:-1]:
+                    continue
+                if name.startswith(root_prefix):
+                    member.name = name[len(root_prefix):]
+                if member.name:
+                    tar.extract(member, '.')
+        PY
+
+    - name: Generate CI Dashboard HTML
+      shell: sh
+      env:
+        GITEA_TOKEN: ${{ github.token }}
+      run: |
+        set +e
+        echo "=== 生成 CI 健康度 HTML 看板 ==="
+        echo "时间: $(date '+%Y-%m-%d %H:%M:%S')"
+        echo ""
+        python3 scripts/ci/ci_dashboard.py --days 7 --html --html-output ci_dashboard.html
+        EXIT_CODE=$?
+        if [ $EXIT_CODE -eq 0 ] && [ -f ci_dashboard.html ]; then
+            HTML_SIZE=$(wc -c < ci_dashboard.html)
+            echo ""
+            echo "✅ HTML 看板生成成功 (${HTML_SIZE} bytes)"
+            echo "路径: $(pwd)/ci_dashboard.html"
+            # 输出文件内容前几行，方便在 Actions 日志中确认
+            echo ""
+            echo "--- 看板预览 (前 5 行) ---"
+            head -5 ci_dashboard.html
+            echo "...（完整内容见产物文件）"
+        else
+            echo "❌ HTML 看板生成失败 (exit code: $EXIT_CODE)"
+        fi
+        echo ""
+        # 永远成功，看板生成失败不影响主流程
+        exit 0
+
+    - name: Run CI health check and report
+      shell: sh
+      env:
+        GITEA_TOKEN: ${{ github.token }}
+        CI_NOTIFY_WEBHOOK: ${{ secrets.CI_NOTIFY_WEBHOOK }}
+      run: |
+        set +e
+        echo "=== CI健康度每日巡检 ==="
+        echo "时间: $(date '+%Y-%m-%d %H:%M:%S')"
+        echo ""
+        python3 scripts/ci/ci_health_report.py --limit 30
+        EXIT_CODE=$?
+        echo ""
+        echo "巡检完成 (exit code: $EXIT_CODE)"
+        # 永远成功，不影响CI状态（通知失败不应该标红）
+        exit 0

.gitea/workflows/ci-trigger-monitor.yml

@@ -0,0 +1,52 @@
+name: CI Trigger Monitor
+
+on:
+  schedule:
+    - cron: '*/5 * * * *'  # 每5分钟检查一次
+  workflow_dispatch:
+    inputs:
+      stale_threshold:
+        description: 'CI未触发告警阈值（分钟）'
+        required: false
+        default: '5'
+
+permissions:
+  contents: read
+
+jobs:
+  monitor:
+    name: Monitor CI Trigger Reliability
+    runs-on: ci-l2
+    timeout-minutes: 5
+    steps:
+      - name: Checkout code
+        shell: sh
+        env:
+          GITHUB_TOKEN: ${{ github.token }}
+        run: |
+          curl -sH "Authorization: token $GITHUB_TOKEN"             "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/step_checkout.sh?ref=${GITHUB_SHA}"             | bash
+      
+      - name: Check CI trigger status for all open PRs
+        env:
+          GITEA_API_TOKEN: ${{ secrets.REVIEW_GITEA_TOKEN }}
+          GITEA_URL: https://git.xiaoxiajianji.com
+          GITEA_REPO: xiaoxia/xiaoxia-saas
+          CI_NOTIFY_WEBHOOK: ${{ secrets.CI_NOTIFY_WEBHOOK }}
+          STALE_THRESHOLD_MIN: ${{ inputs.stale_threshold || 5 }}
+        run: |
+          set +e
+          python3 scripts/ci_trigger_monitor.py
+          # 监控脚本永远不fail，避免告警风暴
+          exit 0
+      - name: Report CI trace
+        if: always()
+        shell: sh
+        env:
+          AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
+        run: |
+          STATUS="ok"
+          [ ${{ job.status }} = "success" ] || STATUS="error"
+          START_TIME=""
+          [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
+          python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true
+

.gitea/workflows/code-review.yml

@@ -0,0 +1,79 @@
+name: AI Code Review
+
+on:
+  pull_request:
+    types:
+      - opened
+      - synchronize
+      - reopened
+
+# 同一个 PR 只跑一个 review，新的取消旧的
+concurrency:
+  group: code-review-${{ gitea.repository }}-${{ gitea.event.pull_request.number }}
+  cancel-in-progress: true
+
+jobs:
+  code-review:
+    name: AI Code Review
+    runs-on: ci-l2
+    # 跳过草稿 PR
+    if: ${{ !gitea.event.pull_request.draft }}
+
+    steps:
+      # actions/checkout 由 runner 在宿主机层面处理，不受容器网络影响
+      - name: Checkout code
+        shell: sh
+        env:
+          GITHUB_TOKEN: ${{ github.token }}
+        run: |
+          curl -sH "Authorization: token $GITHUB_TOKEN"             "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/step_checkout.sh?ref=${GITHUB_SHA}"             | bash
+
+      - name: Install dependencies
+        run: |
+          # 确保 python3-pip 可用（兼容不同基础镜像）
+          if ! python3 -m pip --version >/dev/null 2>&1; then
+            apt-get update -qq && apt-get install -y -qq python3-pip python3-venv >/dev/null 2>&1
+          fi
+          # 部分镜像 ensurepip 方式兜底
+          if ! python3 -m pip --version >/dev/null 2>&1; then
+            python3 -m ensurepip --upgrade 2>/dev/null || curl -sS https://bootstrap.pypa.io/get-pip.py | python3
+          fi
+          python3 -m pip install --upgrade pip
+          python3 -m pip install requests
+
+      - name: Run AI Code Review
+        env:
+          # Gitea 配置（自动从运行环境获取）
+          GITEA_API_URL: ${{ gitea.server_url }}
+          GITEA_TOKEN: ${{ secrets.REVIEW_GITEA_TOKEN }}
+          REPO_NAME: ${{ gitea.repository }}
+          PR_NUMBER: ${{ gitea.event.pull_request.number }}
+          PR_HEAD_SHA: ${{ gitea.event.pull_request.head.sha }}
+          # LLM 提供商: coze (扣子原生Bot) / openai (OpenAI兼容)
+          LLM_PROVIDER: "coze"
+          # 扣子模式配置（默认国内站 api.coze.cn）
+          LLM_BASE_URL: ${{ secrets.LLM_BASE_URL }}
+          LLM_API_KEY: ${{ secrets.LLM_API_KEY }}
+          COZE_BOT_ID: ${{ secrets.COZE_BOT_ID }}
+          LLM_MODEL: ${{ secrets.LLM_MODEL }}
+          # 可选参数
+          MAX_DIFF_CHARS: "30000"
+          LLM_TIMEOUT: "120"
+        run: |
+          python3 scripts/ci_code_review.py
+        # 注意：脚本退出码决定job状态
+        # - 有阻塞级问题 → exit 1 → job失败 → 门禁拦截
+        # - 无阻塞级问题/LLM异常 → exit 0 → 通过（fail-open）
+
+      - name: Report CI trace
+        if: always()
+        shell: sh
+        env:
+          AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
+        run: |
+          STATUS="ok"
+          [ ${{ job.status }} = "success" ] || STATUS="error"
+          START_TIME=""
+          [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
+          python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true
+

.gitea/workflows/daily-check.yml

@@ -0,0 +1,624 @@
+name: Daily Health Check
+# 注意：使用 curl step_checkout.sh 方式以兼容 docker runner
+
+on:
+  schedule:
+    - cron: '0 19 * * *'  # UTC 19:00 = 北京时间凌晨 3:00
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+jobs:
+  # ── 1. 生产环境冒烟测试 ─────────────────────────────────────────────
+  production-smoke:
+    name: Production Smoke Test
+    runs-on: ci-l2
+    timeout-minutes: 8
+    outputs:
+      report: ${{ steps.smoke.outputs.report }}
+
+    steps:
+      - name: Checkout code
+        shell: sh
+        env:
+          GITHUB_TOKEN: ${{ github.token }}
+        run: |
+          curl -sH "Authorization: token $GITHUB_TOKEN" \
+            "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/step_checkout.sh?ref=${GITHUB_SHA}" \
+            | bash
+      - name: Production health check & smoke test
+        id: smoke
+        shell: bash
+        env:
+          SMOKE_ENV: production
+          EXISTING_TOKEN: ${{ secrets.PROD_E2E_TOKEN }}
+          MODULES: health,assets,generation,subscription,nginx
+        run: |
+          set +e
+          START_TIME=$(date +%s)
+          chmod +x tests/e2e/api_smoke_test.sh
+          BASE_URL="https://api.xiaoxiajianji.com" \
+          WEB_URL="https://saas.xiaoxiajianji.com" \
+          SMOKE_ENV="${SMOKE_ENV}" \
+          EXISTING_TOKEN="${EXISTING_TOKEN}" \
+          MODULES="${MODULES}" \
+          CLEANUP_ENABLED=0 \
+          PERF_CHECK_ENABLED=1 \
+          PERF_WARN_THRESHOLD_MS=500 \
+          PERF_FAIL_THRESHOLD_MS=5000 \
+          bash tests/e2e/api_smoke_test.sh 2>&1 | tee /tmp/prod-smoke.log
+          SMOKE_EXIT=${PIPESTATUS[0]}
+          END_TIME=$(date +%s)
+          ELAPSED=$((END_TIME - START_TIME))
+
+          echo ""
+          echo "========== 生产冒烟测试报告 =========="
+          echo "环境: https://api.xiaoxiajianji.com"
+          echo "耗时: ${ELAPSED}s"
+          # 提取通过/失败数
+          grep "测试完成:" /tmp/prod-smoke.log || true
+          if [ "$SMOKE_EXIT" -eq 0 ]; then
+            echo "结果: PASS"
+            echo "report=PASS" >> "${GITHUB_OUTPUT}"
+          else
+            echo "结果: FAIL"
+            grep "失败用例:" /tmp/prod-smoke.log || true
+            echo "report=FAIL" >> "${GITHUB_OUTPUT}"
+          fi
+          echo "======================================"
+          exit $SMOKE_EXIT
+
+      - name: Report CI trace
+        if: always()
+        shell: sh
+        env:
+          AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
+        run: |
+          STATUS="ok"
+          [ ${{ job.status }} = "success" ] || STATUS="error"
+          START_TIME=""
+          [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
+          python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true
+
+  # ── 2. Staging API 集成测试 ─────────────────────────────────────────
+  staging-api-tests:
+    name: Staging API Integration Tests
+    runs-on: ci-l2
+    timeout-minutes: 10
+    outputs:
+      report: ${{ steps.report.outputs.report }}
+
+    steps:
+      - name: Checkout code
+        shell: sh
+        env:
+          GITHUB_TOKEN: ${{ github.token }}
+        run: |
+          curl -sH "Authorization: token $GITHUB_TOKEN" \
+            "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/step_checkout.sh?ref=${GITHUB_SHA}" \
+            | bash
+      - name: Run API smoke test on staging
+        id: smoke
+        shell: bash
+        env:
+          STAGING_TEST_USER: ${{ secrets.STAGING_TEST_USER }}
+          STAGING_TEST_PASSWORD: ${{ secrets.STAGING_TEST_PASSWORD }}
+        run: |
+          set +e
+          START_TIME=$(date +%s)
+          chmod +x tests/e2e/api_smoke_test.sh
+          CONTAINER_NAME="ci-test-$$"
+          docker create --name "$CONTAINER_NAME" \
+            -e BASE_URL=https://staging-api.xiaoxiajianji.com \
+            -e WEB_URL=https://staging.xiaoxiajianji.com \
+            -e TEST_USER="$STAGING_TEST_USER" \
+            -e TEST_PASSWORD="$STAGING_TEST_PASSWORD" \
+            -e CLEANUP_ENABLED=1 \
+            -e PERF_CHECK_ENABLED=1 \
+            -e PERF_WARN_THRESHOLD_MS=500 \
+            -e PERF_FAIL_THRESHOLD_MS=3000 \
+            -w /workspace \
+            git.xiaoxiajianji.com/xiaoxia/base/playwright:v1.45.0-jammy \
+            bash tests/e2e/api_smoke_test.sh 2>&1
+          docker cp . "$CONTAINER_NAME:/workspace"
+          docker start -a "$CONTAINER_NAME" 2>&1 | tee /tmp/staging-api-smoke.log
+          SMOKE_EXIT=${PIPESTATUS[0]}
+          docker rm "$CONTAINER_NAME" > /dev/null 2>&1 || true
+          END_TIME=$(date +%s)
+          ELAPSED=$((END_TIME - START_TIME))
+
+          echo ""
+          echo "========== Staging API 冒烟测试报告 =========="
+          echo "环境: https://staging-api.xiaoxiajianji.com"
+          echo "耗时: ${ELAPSED}s"
+          grep "测试完成:" /tmp/staging-api-smoke.log || true
+          if [ "$SMOKE_EXIT" -eq 0 ]; then
+            echo "结果: PASS"
+            echo "api_report=PASS" >> "${GITHUB_OUTPUT}"
+          else
+            echo "结果: FAIL"
+            grep "失败用例:" /tmp/staging-api-smoke.log || true
+            echo "api_report=FAIL" >> "${GITHUB_OUTPUT}"
+          fi
+          echo "=============================================="
+          exit $SMOKE_EXIT
+
+      - name: Run Staging API Integration Tests (Playwright)
+        id: e2e_api
+        shell: bash
+        run: |
+          set +e
+          START_TIME=$(date +%s)
+          CONTAINER_NAME="ci-test-$$"
+          docker create --name "$CONTAINER_NAME" \
+            -e E2E_BASE_URL=https://staging.xiaoxiajianji.com \
+            -e E2E_API_BASE=https://staging-api.xiaoxiajianji.com/api/v1 \
+            -w /workspace/apps/web \
+            git.xiaoxiajianji.com/xiaoxia/base/playwright:v1.45.0-jammy \
+            sh -lc "npm ci && npx playwright test --reporter=line e2e/test_auth.spec.ts e2e/test_asset.spec.ts e2e/test_project.spec.ts" 2>&1
+          docker cp . "$CONTAINER_NAME:/workspace"
+          docker start -a "$CONTAINER_NAME" 2>&1 | tee /tmp/staging-api-e2e.log
+          EXIT_CODE=${PIPESTATUS[0]}
+          docker rm "$CONTAINER_NAME" > /dev/null 2>&1 || true
+          END_TIME=$(date +%s)
+          ELAPSED=$((END_TIME - START_TIME))
+
+          echo ""
+          echo "========== Staging API 集成测试报告 =========="
+          echo "环境: https://staging-api.xiaoxiajianji.com"
+          echo "耗时: ${ELAPSED}s"
+          grep -E "passed|failed|timed out" /tmp/staging-api-e2e.log || true
+          if [ "$EXIT_CODE" -eq 0 ]; then
+            echo "结果: PASS"
+            echo "int_report=PASS" >> "${GITHUB_OUTPUT}"
+          else
+            echo "结果: FAIL"
+            echo "int_report=FAIL" >> "${GITHUB_OUTPUT}"
+          fi
+          echo "=============================================="
+          exit $EXIT_CODE
+
+      - name: Set report output
+        id: report
+        shell: sh
+        run: |
+          if [ "${{ steps.smoke.outputs.api_report }}" = "PASS" ] && [ "${{ steps.e2e_api.outputs.int_report }}" = "PASS" ]; then
+            echo "report=PASS" >> "${GITHUB_OUTPUT}"
+          else
+            echo "report=FAIL" >> "${GITHUB_OUTPUT}"
+          fi
+
+      - name: Report CI trace
+        if: always()
+        shell: sh
+        env:
+          AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
+        run: |
+          STATUS="ok"
+          [ ${{ job.status }} = "success" ] || STATUS="error"
+          START_TIME=""
+          [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
+          python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true
+
+  # ── 3. Staging 浏览器 E2E ──────────────────────────────────────────
+  staging-e2e:
+    name: Staging Browser E2E
+    runs-on: ci-l2
+    timeout-minutes: 15
+    outputs:
+      report: ${{ steps.e2e.outputs.report }}
+
+    steps:
+      - name: Checkout code
+        shell: sh
+        env:
+          GITHUB_TOKEN: ${{ github.token }}
+        run: |
+          curl -sH "Authorization: token $GITHUB_TOKEN" \
+            "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/step_checkout.sh?ref=${GITHUB_SHA}" \
+            | bash
+      - name: Run Playwright E2E on staging
+        id: e2e
+        shell: bash
+        run: |
+          set +e
+          START_TIME=$(date +%s)
+          CONTAINER_NAME="ci-test-$$"
+          docker create --name "$CONTAINER_NAME" --ipc=host \
+            -e E2E_BASE_URL=https://staging.xiaoxiajianji.com \
+            -e E2E_API_BASE=https://staging-api.xiaoxiajianji.com/api/v1 \
+            -e E2E_BROWSER_CHANNEL=chromium \
+            -e PLAYWRIGHT_HEADLESS=1 \
+            -w /workspace/apps/web \
+            git.xiaoxiajianji.com/xiaoxia/base/playwright:v1.45.0-jammy \
+            sh -lc 'npm ci && npx playwright test --reporter=line --project=chromium e2e/auth.spec.ts e2e/auth-guard.spec.ts e2e/core-upload.spec.ts e2e/core-generation.spec.ts e2e/core-titles.spec.ts' 2>&1
+          docker cp . "$CONTAINER_NAME:/workspace"
+          docker start -a "$CONTAINER_NAME" 2>&1 | tee /tmp/staging-e2e.log
+          EXIT_CODE=${PIPESTATUS[0]}
+          docker rm "$CONTAINER_NAME" > /dev/null 2>&1 || true
+          END_TIME=$(date +%s)
+          ELAPSED=$((END_TIME - START_TIME))
+
+          echo ""
+          echo "========== Staging E2E 测试报告 =========="
+          echo "环境: https://staging.xiaoxiajianji.com"
+          echo "耗时: ${ELAPSED}s"
+          grep -E "passed|failed|timed out" /tmp/staging-e2e.log || true
+          if [ "$EXIT_CODE" -eq 0 ]; then
+            echo "结果: PASS"
+            echo "report=PASS" >> "${GITHUB_OUTPUT}"
+          else
+            echo "结果: FAIL"
+            echo "report=FAIL" >> "${GITHUB_OUTPUT}"
+          fi
+          echo "=========================================="
+          exit $EXIT_CODE
+
+      - name: Report CI trace
+        if: always()
+        shell: sh
+        env:
+          AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
+        run: |
+          STATUS="ok"
+          [ ${{ job.status }} = "success" ] || STATUS="error"
+          START_TIME=""
+          [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
+          python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true
+
+  # ── 4. 性能基线巡检 ────────────────────────────────────────────────
+  performance-check:
+    name: Performance Baseline Check
+    runs-on: ci-l2
+    timeout-minutes: 8
+    outputs:
+      report: ${{ steps.report.outputs.report }}
+
+    steps:
+      - name: Run performance baseline checks
+        id: perf
+        shell: sh
+        env:
+          STAGING_TEST_USER: ${{ secrets.STAGING_TEST_USER }}
+          STAGING_TEST_PASSWORD: ${{ secrets.STAGING_TEST_PASSWORD }}
+        run: |
+          set +e
+          START_TIME=$(date +%s)
+          echo "=========================================="
+          echo "  性能基线巡检 - Staging API"
+          echo "  目标: https://staging-api.xiaoxiajianji.com"
+          echo "=========================================="
+          echo ""
+
+          TOTAL=0
+          PASS=0
+          FAIL=0
+          WARN=0
+          WARN_LIST=""
+          FAIL_LIST=""
+
+          # 核心接口配置: 名称|路径|方法|阈值(ms)|失败阈值(ms)
+          # 核心接口(core): 500ms
+          # 普通接口(normal): 1000ms
+          # 重操作接口(heavy): 3000ms
+          ENDPOINTS="
+          登录|/api/v1/auth/login|POST|500|3000
+          获取当前用户|/api/v1/auth/me|GET|500|3000
+          项目列表|/api/v1/projects|GET|500|3000
+          素材列表|/api/v1/assets|GET|500|3000
+          模板列表|/api/v1/templates|GET|500|3000
+          剪辑计划列表|/api/v1/edit-plans|GET|500|3000
+          生成任务列表|/api/v1/generation/tasks|GET|500|3000
+          订阅信息|/api/v1/subscription/current|GET|500|3000
+          音色列表|/api/v1/voices|GET|1000|5000
+          健康检查|/health|GET|200|1000
+          "
+
+          # 先登录获取 token
+          echo "--- 准备: 获取测试 Token ---"
+          LOGIN_BODY="{\"email\":\"$STAGING_TEST_USER\",\"password\":\"$STAGING_TEST_PASSWORD\"}"
+          AUTH_RESP=$(curl -s -w "\n%{http_code}" -X POST \
+            -H "Content-Type: application/json" \
+            -d "$LOGIN_BODY" \
+            "https://staging-api.xiaoxiajianji.com/api/v1/auth/login" \
+            --max-time 10 2>&1)
+          AUTH_CODE=$(echo "$AUTH_RESP" | tail -1)
+          AUTH_BODY=$(echo "$AUTH_RESP" | sed '$d')
+
+          if [ "$AUTH_CODE" = "200" ]; then
+            TOKEN=$(echo "$AUTH_BODY" | python3 -c "import sys,json; d=json.load(sys.stdin); print(d.get('access_token',''))" 2>/dev/null)
+            if [ -n "$TOKEN" ]; then
+              echo "Token 获取成功"
+            else
+              echo "Token 解析失败，部分接口可能无法测试"
+              TOKEN=""
+            fi
+          else
+            echo "登录失败 (HTTP $AUTH_CODE)，部分接口将跳过鉴权测试"
+            TOKEN=""
+          fi
+
+          echo ""
+          echo "--- 开始性能测试 ---"
+          echo ""
+
+          echo "$ENDPOINTS" | while IFS='|' read -r name path method warn_ms fail_ms; do
+            [ -z "$name" ] && continue
+            TOTAL=$((TOTAL + 1))
+
+            # 构建 curl 命令
+            CURL_ARGS="-s -o /dev/null -w '%{http_code} %{time_total}' --max-time 30"
+            if [ "$method" = "POST" ]; then
+              CURL_ARGS="$CURL_ARGS -X POST -H 'Content-Type: application/json' -d \"$LOGIN_BODY\""
+            fi
+            if [ -n "$TOKEN" ] && [ "$name" != "健康检查" ]; then
+              CURL_ARGS="$CURL_ARGS -H 'Authorization: Bearer $TOKEN'"
+            fi
+
+            # 执行请求
+            RESP=$(eval curl $CURL_ARGS "https://staging-api.xiaoxiajianji.com${path}" 2>&1)
+            HTTP_CODE=$(echo "$RESP" | awk '{print $1}')
+            TIME_TOTAL=$(echo "$RESP" | awk '{print $2}')
+            ELAPSED_MS=$(python3 -c "print(int(float('${TIME_TOTAL:-0}') * 1000))" 2>/dev/null || echo "0")
+
+            if [ "$HTTP_CODE" -ge 500 ] 2>/dev/null; then
+              FAIL=$((FAIL + 1))
+              FAIL_LIST="$FAIL_LIST\n  ❌ $name - HTTP $HTTP_CODE (${ELAPSED_MS}ms)"
+              echo "❌ $name - HTTP $HTTP_CODE - ${ELAPSED_MS}ms (FAIL)"
+            elif [ "$ELAPSED_MS" -ge "$fail_ms" ] 2>/dev/null; then
+              FAIL=$((FAIL + 1))
+              FAIL_LIST="$FAIL_LIST\n  ❌ $name - ${ELAPSED_MS}ms > ${fail_ms}ms"
+              echo "❌ $name - ${ELAPSED_MS}ms > ${fail_ms}ms (FAIL)"
+            elif [ "$ELAPSED_MS" -ge "$warn_ms" ] 2>/dev/null; then
+              WARN=$((WARN + 1))
+              WARN_LIST="$WARN_LIST\n  ⚠️  $name - ${ELAPSED_MS}ms > ${warn_ms}ms"
+              echo "⚠️  $name - ${ELAPSED_MS}ms (WARN, threshold: ${warn_ms}ms)"
+              PASS=$((PASS + 1))
+            else
+              PASS=$((PASS + 1))
+              echo "✅ $name - ${ELAPSED_MS}ms (OK, threshold: ${warn_ms}ms)"
+            fi
+          done
+
+          # 由于 while 在子 shell 中执行，用文件传递结果
+          # 重新跑一次用文件计数方式
+          echo ""
+          echo "--- 汇总性能数据 ---"
+
+          END_TIME=$(date +%s)
+          ELAPSED=$((END_TIME - START_TIME))
+
+          echo ""
+          echo "========== 性能基线巡检报告 =========="
+          echo "环境: https://staging-api.xiaoxiajianji.com"
+          echo "耗时: ${ELAPSED}s"
+          echo "======================================"
+
+      - name: Generate performance report
+        id: report
+        shell: sh
+        env:
+          STAGING_TEST_USER: ${{ secrets.STAGING_TEST_USER }}
+          STAGING_TEST_PASSWORD: ${{ secrets.STAGING_TEST_PASSWORD }}
+        run: |
+          set +e
+          echo ""
+          echo "=========================================="
+          echo "  性能基线巡检 - 详细报告"
+          echo "=========================================="
+
+          TOTAL=0
+          PASS=0
+          FAIL=0
+          WARN=0
+          RESULTS=""
+          START_TIME=$(date +%s)
+
+          LOGIN_BODY="{\"email\":\"$STAGING_TEST_USER\",\"password\":\"$STAGING_TEST_PASSWORD\"}"
+          # 先登录获取 token
+          AUTH_RESP=$(curl -s -w "\n%{http_code}" -X POST \
+            -H "Content-Type: application/json" \
+            -d "$LOGIN_BODY" \
+            "https://staging-api.xiaoxiajianji.com/api/v1/auth/login" \
+            --max-time 10 2>&1)
+          AUTH_CODE=$(echo "$AUTH_RESP" | tail -1)
+          AUTH_BODY=$(echo "$AUTH_RESP" | sed '$d')
+          TOKEN=""
+          if [ "$AUTH_CODE" = "200" ]; then
+            TOKEN=$(echo "$AUTH_BODY" | python3 -c "import sys,json; d=json.load(sys.stdin); print(d.get('access_token',''))" 2>/dev/null || echo "")
+          fi
+
+          run_perf_test() {
+            local name="$1" path="$2" method="$3" warn_ms="$4" fail_ms="$5"
+            TOTAL=$((TOTAL + 1))
+
+            local CURL_ARGS="-s -o /dev/null -w '%{http_code} %{time_total}' --max-time 30"
+            if [ "$method" = "POST" ]; then
+              CURL_ARGS="$CURL_ARGS -X POST -H 'Content-Type: application/json' -d \"$LOGIN_BODY\""
+            fi
+            if [ -n "$TOKEN" ] && [ "$name" != "健康检查" ]; then
+              CURL_ARGS="$CURL_ARGS -H 'Authorization: Bearer $TOKEN'"
+            fi
+
+            local RESP=$(eval curl $CURL_ARGS "https://staging-api.xiaoxiajianji.com${path}" 2>&1)
+            local HTTP_CODE=$(echo "$RESP" | awk '{print $1}')
+            local TIME_TOTAL=$(echo "$RESP" | awk '{print $2}')
+            local ELAPSED_MS=$(python3 -c "print(int(float('${TIME_TOTAL:-0}') * 1000))" 2>/dev/null || echo "0")
+
+            if echo "$HTTP_CODE" | grep -q "^[5]"; then
+              FAIL=$((FAIL + 1))
+              RESULTS="$RESULTS\n  ❌ $name - HTTP $HTTP_CODE (${ELAPSED_MS}ms)"
+              echo "❌ $name - HTTP $HTTP_CODE - ${ELAPSED_MS}ms [FAIL]"
+              return 1
+            elif [ "$ELAPSED_MS" -ge "$fail_ms" ] 2>/dev/null; then
+              FAIL=$((FAIL + 1))
+              RESULTS="$RESULTS\n  ❌ $name - ${ELAPSED_MS}ms > ${fail_ms}ms [FAIL]"
+              echo "❌ $name - ${ELAPSED_MS}ms > ${fail_ms}ms [FAIL]"
+              return 1
+            elif [ "$ELAPSED_MS" -ge "$warn_ms" ] 2>/dev/null; then
+              WARN=$((WARN + 1))
+              PASS=$((PASS + 1))
+              RESULTS="$RESULTS\n  ⚠️  $name - ${ELAPSED_MS}ms (阈值: ${warn_ms}ms) [WARN]"
+              echo "⚠️  $name - ${ELAPSED_MS}ms > 阈值 ${warn_ms}ms [WARN]"
+              return 0
+            else
+              PASS=$((PASS + 1))
+              RESULTS="$RESULTS\n  ✅ $name - ${ELAPSED_MS}ms (阈值: ${warn_ms}ms) [OK]"
+              echo "✅ $name - ${ELAPSED_MS}ms (阈值: ${warn_ms}ms) [OK]"
+              return 0
+            fi
+          }
+
+          echo ""
+          echo "=== 核心接口 (阈值: 500ms / 3000ms) ==="
+          run_perf_test "登录" "/api/v1/auth/login" "POST" 500 3000 || true
+          run_perf_test "获取当前用户" "/api/v1/auth/me" "GET" 500 3000 || true
+          run_perf_test "项目列表" "/api/v1/projects" "GET" 500 3000 || true
+          run_perf_test "素材列表" "/api/v1/assets" "GET" 500 3000 || true
+          run_perf_test "模板列表" "/api/v1/templates" "GET" 500 3000 || true
+          run_perf_test "剪辑计划列表" "/api/v1/edit-plans" "GET" 500 3000 || true
+          run_perf_test "生成任务列表" "/api/v1/generation/tasks" "GET" 500 3000 || true
+          run_perf_test "订阅信息" "/api/v1/subscription/current" "GET" 500 3000 || true
+
+          echo ""
+          echo "=== 普通接口 (阈值: 1000ms / 5000ms) ==="
+          run_perf_test "音色列表" "/api/v1/voices" "GET" 1000 5000 || true
+
+          echo ""
+          echo "=== 基础接口 (阈值: 200ms / 1000ms) ==="
+          run_perf_test "健康检查" "/health" "GET" 200 1000 || true
+
+          END_TIME=$(date +%s)
+          ELAPSED=$((END_TIME - START_TIME))
+
+          echo ""
+          echo "========== 性能基线巡检报告 =========="
+          echo "环境: https://staging-api.xiaoxiajianji.com"
+          echo "总接口: ${TOTAL}"
+          echo "通过: ${PASS}"
+          echo "失败: ${FAIL}"
+          echo "警告: ${WARN}"
+          echo "耗时: ${ELAPSED}s"
+          echo "======================================"
+
+          # 写入结果文件供 report job 使用
+          echo "${TOTAL}" > /tmp/perf_total
+          echo "${PASS}" > /tmp/perf_pass
+          echo "${FAIL}" > /tmp/perf_fail
+          echo "${WARN}" > /tmp/perf_warn
+          echo "${ELAPSED}" > /tmp/perf_elapsed
+
+          if [ "$FAIL" -gt 0 ]; then
+            echo "report=FAIL" >> "${GITHUB_OUTPUT}"
+            echo "perf_detail=fail:${FAIL}:warn:${WARN}" >> "${GITHUB_OUTPUT}"
+            exit 1
+          else
+            echo "report=PASS" >> "${GITHUB_OUTPUT}"
+            if [ "$WARN" -gt 0 ]; then
+              echo "perf_detail=pass:warn:${WARN}" >> "${GITHUB_OUTPUT}"
+            else
+              echo "perf_detail=pass" >> "${GITHUB_OUTPUT}"
+            fi
+            exit 0
+          fi
+
+      - name: Report CI trace
+        if: always()
+        shell: sh
+        env:
+          AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
+        run: |
+          STATUS="ok"
+          [ ${{ job.status }} = "success" ] || STATUS="error"
+          START_TIME=""
+          [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
+          python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true
+
+  # ── 5. 每日巡检汇总报告 ────────────────────────────────────────────
+  daily-report:
+    name: Daily Check Report
+    runs-on: ci-l2
+    timeout-minutes: 2
+    if: always()
+    needs:
+      - production-smoke
+      - staging-api-tests
+      - staging-e2e
+      - performance-check
+
+    steps:
+      - name: Print summary report
+        shell: sh
+        run: |
+          echo ""
+          echo "╔══════════════════════════════════════════════════════╗"
+          echo "║              每日巡检报告                             ║"
+          echo "╠══════════════════════════════════════════════════════╣"
+
+          # 获取各 job 状态
+          PROD_STATUS="${{ needs.production-smoke.result }}"
+          STAGING_API_STATUS="${{ needs.staging-api-tests.result }}"
+          STAGING_E2E_STATUS="${{ needs.staging-e2e.result }}"
+          PERF_STATUS="${{ needs.performance-check.result }}"
+
+          format_result() {
+            if [ "$1" = "success" ]; then
+              echo "✅ PASS"
+            elif [ "$1" = "failure" ]; then
+              echo "❌ FAIL"
+            elif [ "$1" = "skipped" ]; then
+              echo "⏭️  SKIP"
+            else
+              echo "❓ UNKNOWN ($1)"
+            fi
+          }
+
+          echo "║"
+          echo "║  生产冒烟测试:   $(format_result "$PROD_STATUS")"
+          echo "║  Staging API:    $(format_result "$STAGING_API_STATUS")"
+          echo "║  Staging E2E:    $(format_result "$STAGING_E2E_STATUS")"
+          echo "║  性能基线巡检:   $(format_result "$PERF_STATUS")"
+          echo "║"
+          echo "║  巡检时间: $(date '+%Y-%m-%d %H:%M:%S UTC')"
+          echo "║"
+
+          # 判断整体状态
+          ALL_PASS=true
+          FAILED_ITEMS=""
+          for status_name in "$PROD_STATUS:生产冒烟" "$STAGING_API_STATUS:Staging API" "$STAGING_E2E_STATUS:Staging E2E" "$PERF_STATUS:性能基线"; do
+            STATUS=$(echo "$status_name" | cut -d: -f1)
+            NAME=$(echo "$status_name" | cut -d: -f2)
+            if [ "$STATUS" != "success" ] && [ "$STATUS" != "skipped" ]; then
+              ALL_PASS=false
+              FAILED_ITEMS="$FAILED_ITEMS $NAME"
+            fi
+          done
+
+          echo "╠══════════════════════════════════════════════════════╣"
+          if [ "$ALL_PASS" = "true" ]; then
+            echo "║  整体状态: ✅ 全部通过                                 ║"
+          else
+            echo "║  整体状态: ❌ 存在失败                                 ║"
+            echo "║  失败项: ${FAILED_ITEMS}                              ║"
+          fi
+          echo "╚══════════════════════════════════════════════════════╝"
+          echo ""
+
+          # 如果有失败项，以非零退出码结束（方便 Gitea 标记流水线失败）
+          if [ "$ALL_PASS" = "false" ]; then
+            echo "⚠️  部分巡检项失败，请检查上方日志获取详细信息。"
+            # 不 exit 1，因为我们用了 always()，保持 report job 成功，
+            # 但其他失败的 job 已经让整体流水线标记为失败
+          fi
+      - name: Report CI trace
+        if: always()
+        shell: sh
+        env:
+          AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
+        run: |
+          STATUS="ok"
+          [ ${{ job.status }} = "success" ] || STATUS="error"
+          START_TIME=""
+          [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
+          python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true

.gitea/workflows/deploy.yml

@@ -1,219 +0,0 @@
-name: Deploy
-
-on:
-  push:
-    tags:
-      - 'v*'
-
-jobs:
-  deploy-staging:
-    name: Deploy Staging
-    runs-on: runtime-builder
-    if: github.ref_name == 'main' || github.ref_name == 'develop' || startsWith(github.ref_name, 'feature/')
-
-    steps:
-    - name: Checkout code
-      shell: sh
-      env:
-        GITHUB_TOKEN: ${{ github.token }}
-      run: |
-        set -eu
-        archive_url="${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/archive/${GITHUB_SHA}.tar.gz"
-        wget --header="Authorization: token ${GITHUB_TOKEN}" -O /tmp/repo.tar.gz "$archive_url"
-        tar -xzf /tmp/repo.tar.gz --strip-components=1 -C .
-        rm -f /tmp/repo.tar.gz
-
-    - name: Build staging web artifact
-      shell: sh
-      run: |
-        set -eu
-        docker run --rm \
-          --pull=never \
-          -v "$PWD:/workspace" \
-          -w /workspace/apps/web \
-          node:20 \
-          sh -lc 'npm ci && npm run build'
-        docker build --pull=false \
-          -f infra/docker/web-artifact.Dockerfile \
-          -t "xiaoxia-saas-web:staging-${GITHUB_SHA}" \
-          .
-        test -f apps/web/dist/index.html
-
-    - name: Package staging release artifact
-      shell: sh
-      run: |
-        set -eu
-        rm -rf dist/staging-artifacts
-        mkdir -p dist/staging-artifacts
-        tar --exclude=.git --exclude=apps/web/node_modules --exclude=./dist \
-          -czf dist/staging-artifacts/xiaoxia-staging-${GITHUB_SHA}.tar.gz .
-        docker save -o "dist/staging-artifacts/xiaoxia-web-staging-${GITHUB_SHA}.tar" "xiaoxia-saas-web:staging-${GITHUB_SHA}"
-
-    - name: Upload staging artifact to business host
-      shell: sh
-      env:
-        STAGING_SSH_HOST: ${{ secrets.STAGING_SSH_HOST }}
-        STAGING_SSH_USER: ${{ secrets.STAGING_SSH_USER }}
-        STAGING_SSH_KEY: ${{ secrets.STAGING_SSH_KEY }}
-      run: |
-        set -eu
-        staging_host="${STAGING_SSH_HOST:-47.98.113.167}"
-        [ -z "$staging_host" ] && staging_host="47.98.113.167"
-        staging_user="${STAGING_SSH_USER:-root}"
-        mkdir -p ~/.ssh
-        key_path="/root/.ssh/xiaoxia_runtime_builder"
-        echo "Using SSH key: $key_path"
-        ls -la "$key_path"
-        test -f "$key_path" || { echo "ERROR: SSH key not found"; exit 1; }
-        ssh-keyscan -H "$staging_host" >> ~/.ssh/known_hosts
-        ssh -i "$key_path" "$staging_user@$staging_host" "mkdir -p /var/lib/xiaoxia-saas-staging/artifacts"
-        scp -i "$key_path" "dist/staging-artifacts/xiaoxia-staging-${GITHUB_SHA}.tar.gz" \
-          "$staging_user@$staging_host:/var/lib/xiaoxia-saas-staging/artifacts/xiaoxia-staging-${GITHUB_SHA}.tar.gz"
-        scp -i "$key_path" "dist/staging-artifacts/xiaoxia-web-staging-${GITHUB_SHA}.tar" \
-          "$staging_user@$staging_host:/var/lib/xiaoxia-saas-staging/artifacts/xiaoxia-web-staging-${GITHUB_SHA}.tar"
-
-    - name: Deploy staging stack on business host
-      shell: sh
-      env:
-        STAGING_SSH_HOST: ${{ secrets.STAGING_SSH_HOST }}
-        STAGING_SSH_USER: ${{ secrets.STAGING_SSH_USER }}
-        STAGING_SSH_KEY: ${{ secrets.STAGING_SSH_KEY }}
-      run: |
-        set -eu
-        staging_host="${STAGING_SSH_HOST:-47.98.113.167}"
-        [ -z "$staging_host" ] && staging_host="47.98.113.167"
-        staging_user="${STAGING_SSH_USER:-root}"
-        key_path="/root/.ssh/xiaoxia_runtime_builder"
-        echo "Using SSH key: $key_path"
-        ls -la "$key_path"
-        test -f "$key_path" || { echo "ERROR: SSH key not found"; exit 1; }
-
-  build-production-runtime-images:
-    name: Build Production Runtime Images
-    runs-on: runtime-builder
-    if: github.ref_type == 'tag'
-
-    steps:
-    - name: Checkout code
-      shell: sh
-      env:
-        GITHUB_TOKEN: ${{ github.token }}
-      run: |
-        set -eu
-        archive_url="${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/archive/${GITHUB_SHA}.tar.gz"
-        wget --header="Authorization: token ${GITHUB_TOKEN}" -O /tmp/repo.tar.gz "$archive_url"
-        tar -xzf /tmp/repo.tar.gz --strip-components=1 -C .
-        rm -f /tmp/repo.tar.gz
-
-    - name: Build runtime image artifact
-      shell: sh
-      run: |
-        set -eu
-        chmod +x scripts/build_release_images.sh
-        scripts/build_release_images.sh "${GITHUB_REF_NAME}"
-
-    - name: Build production web artifact
-      shell: sh
-      run: |
-        set -eu
-        docker run --rm \
-          --pull=never \
-          -v "$PWD:/workspace" \
-          -w /workspace/apps/web \
-          node:20 \
-          sh -lc 'npm ci && npm run build'
-        docker build --pull=false \
-          -f infra/docker/web-artifact.Dockerfile \
-          -t "xiaoxia-saas-web:${GITHUB_REF_NAME}" \
-          .
-        test -f apps/web/dist/index.html
-
-    - name: Package release source artifact
-      shell: sh
-      run: |
-        set -eu
-        mkdir -p dist/release-artifacts
-        tar --exclude=.git --exclude=apps/web/node_modules --exclude=./dist \
-          -czf "dist/release-artifacts/xiaoxia-release-${GITHUB_REF_NAME}.tar.gz" .
-        docker save -o "dist/release-artifacts/xiaoxia-web-${GITHUB_REF_NAME}.tar" "xiaoxia-saas-web:${GITHUB_REF_NAME}"
-
-    - name: Upload runtime image and release artifacts
-      shell: sh
-      env:
-        PRODUCTION_SSH_HOST: ${{ secrets.PRODUCTION_SSH_HOST }}
-        PRODUCTION_SSH_USER: ${{ secrets.PRODUCTION_SSH_USER }}
-        PRODUCTION_SSH_KEY: ${{ secrets.PRODUCTION_SSH_KEY }}
-      run: |
-        set -eu
-        production_host="${PRODUCTION_SSH_HOST:-47.98.113.167}"
-        [ -z "$production_host" ] && production_host="47.98.113.167"
-        production_user="${PRODUCTION_SSH_USER:-root}"
-        mkdir -p ~/.ssh
-        key_path="/root/.ssh/xiaoxia_runtime_builder"
-        echo "Using SSH key: $key_path"
-        ls -la "$key_path"
-        test -f "$key_path" || { echo "ERROR: SSH key not found"; exit 1; }
-        ssh-keyscan -H "$production_host" >> ~/.ssh/known_hosts
-        ssh -i "$key_path" -o BatchMode=yes -o StrictHostKeyChecking=no "$production_user@$production_host" "echo SSH_OK" || { echo "SSH FAILED"; exit 1; }
-        scp -i "$key_path" "dist/release-images/xiaoxia-runtime-images-${GITHUB_REF_NAME}.tar" \
-          "$production_user@$production_host:/var/lib/xiaoxia-saas-production/runtime-images-${GITHUB_REF_NAME}.tar"
-        scp -i "$key_path" "dist/release-artifacts/xiaoxia-release-${GITHUB_REF_NAME}.tar.gz" \
-          "$production_user@$production_host:/var/lib/xiaoxia-saas-production/release-${GITHUB_REF_NAME}.tar.gz"
-        scp -i "$key_path" "dist/release-artifacts/xiaoxia-web-${GITHUB_REF_NAME}.tar" \
-          "$production_user@$production_host:/var/lib/xiaoxia-saas-production/web-${GITHUB_REF_NAME}.tar"
-
-  deploy-production:
-    name: Deploy Production
-    runs-on: runtime-builder
-    needs: build-production-runtime-images
-
-    steps:
-    - name: Deploy production over SSH
-      shell: sh
-      env:
-        PRODUCTION_SSH_HOST: ${{ secrets.PRODUCTION_SSH_HOST }}
-        PRODUCTION_SSH_USER: ${{ secrets.PRODUCTION_SSH_USER }}
-        PRODUCTION_SSH_KEY: ${{ secrets.PRODUCTION_SSH_KEY }}
-      run: |
-        set -eu
-        production_host="${PRODUCTION_SSH_HOST:-47.98.113.167}"
-        [ -z "$production_host" ] && production_host="47.98.113.167"
-        production_user="${PRODUCTION_SSH_USER:-root}"
-        mkdir -p ~/.ssh
-        key_path="/root/.ssh/xiaoxia_runtime_builder"
-        echo "Using SSH key: $key_path"
-        ls -la "$key_path"
-        test -f "$key_path" || { echo "ERROR: SSH key not found"; exit 1; }
-        ssh-keyscan -H "$production_host" >> ~/.ssh/known_hosts
-          echo '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' | base64 -d | ssh -i "$key_path" "$production_user@$production_host" "RELEASE_VERSION='${GITHUB_REF_NAME}' sh"
-
-  production-e2e:
-    name: Production Browser E2E
-    runs-on: runtime-builder
-    needs: deploy-production
-
-    steps:
-    - name: Checkout code
-      shell: sh
-      env:
-        GITHUB_TOKEN: ${{ github.token }}
-      run: |
-        set -eu
-        archive_url="${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/archive/${GITHUB_SHA}.tar.gz"
-        wget --header="Authorization: token ${GITHUB_TOKEN}" -O /tmp/repo.tar.gz "$archive_url"
-        tar -xzf /tmp/repo.tar.gz --strip-components=1 -C .
-        rm -f /tmp/repo.tar.gz
-
-    - name: Run production browser E2E
-      shell: sh
-      run: |
-        set -eu
-        docker run --rm \
-          -e E2E_BASE_URL=https://saas.xiaoxiajianji.com \
-          -e E2E_API_BASE=https://api.xiaoxiajianji.com/api/v1 \
-          -e E2E_BROWSER_CHANNEL=chromium \
-          -v "$PWD:/workspace" \
-          -w /workspace/apps/web \
-          mcr.microsoft.com/playwright:v1.45.0-jammy \
-          sh -lc 'npm ci && npx playwright test --reporter=line --project=chromium e2e/core-upload.spec.ts e2e/core-generation.spec.ts e2e/core-titles.spec.ts'
-

.gitea/workflows/pr-auto-scan.yml

@@ -0,0 +1,56 @@
+name: PR Auto Scan
+# 定时扫描所有open PR，对CI全绿的触发审批/合并
+# 作为短作业模式的兜底，防止事件驱动遗漏
+on:
+  schedule:
+    - cron: "*/5 * * * *"  # 每5分钟扫描一次
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+jobs:
+  auto-scan:
+    name: Auto Scan Open PRs
+    runs-on: ci-check
+    timeout-minutes: 5
+    if: github.repository == 'xiaoxia/xiaoxia-saas'
+    steps:
+      - name: Checkout code
+        shell: sh
+        env:
+          GITHUB_TOKEN: ${{ github.token }}
+        run: |
+          curl -sH "Authorization: token $GITHUB_TOKEN"             "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/pr_auto_scan.py?ref=develop"             -o /tmp/pr_auto_scan.py
+          python3 /tmp/pr_auto_scan.py --help > /dev/null 2>&1 || {
+            # fallback: checkout
+            echo "使用checkout方式"
+            curl -sH "Authorization: token $GITHUB_TOKEN"               "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/step_checkout.sh?ref=develop"               | bash
+          }
+
+      - name: Scan and auto process PRs
+        shell: bash
+        env:
+          GITHUB_TOKEN: ${{ github.token }}
+          REVIEW_TOKEN: ${{ secrets.REVIEW_GITEA_TOKEN }}
+          MERGE_TOKEN: ${{ secrets.REVIEW_GITEA_TOKEN }}
+        run: |
+          set -eu
+          echo "=== 扫描所有open PR并自动处理 ==="
+          echo "时间: $(date)"
+          echo
+
+          python3 /tmp/pr_auto_scan.py             --token "$REVIEW_TOKEN"             --repo "$GITHUB_REPOSITORY"             --base develop             --approve             --merge             --dry-run false
+
+          echo ""
+          echo "✅ 扫描完成"
+
+      - name: Report CI trace
+        if: always()
+        shell: sh
+        env:
+          AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
+        run: |
+          STATUS="ok"
+          [ ${{ job.status }} = "success" ] || STATUS="error"
+          python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "" || true

.gitea/workflows/pr-automation.yml

@@ -0,0 +1,117 @@
+name: PR Automation
+
+on:
+  pull_request:
+    types: [synchronize, opened, ready_for_review, review_requested]
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+concurrency:
+  group: pr-automation-${{ gitea.event.pull_request.number }}
+  cancel-in-progress: true
+
+jobs:
+  auto-approve:
+    name: Auto Approve on CI Green
+    runs-on: ci-check
+    if: github.event_name == 'pull_request' && !github.event.pull_request.draft
+    timeout-minutes: 3  # 长等待模式：等CI全绿后自动合并，不遗漏任何PR
+    steps:
+      - name: Checkout code
+        shell: sh
+        env:
+          GITHUB_TOKEN: ${{ github.token }}
+        run: |
+          curl -sH "Authorization: token $GITHUB_TOKEN"             "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/step_checkout.sh?ref=${GITHUB_SHA}"             | bash
+
+      - name: "🔍 脚本语法自检"
+        shell: bash
+        run: |
+          ERROR=0
+          for f in scripts/ci/*.sh; do [ -f "$f" ] && bash -n "$f" 2>&1 || ERROR=$((ERROR+1)); done
+          for f in scripts/ci/*.py; do [ -f "$f" ] && python3 -m py_compile "$f" 2>&1 || ERROR=$((ERROR+1)); done
+          if [ "$ERROR" -ne 0 ]; then echo "❌ 语法自检失败 ($ERROR个)"; exit 1; fi
+          echo "✅ 脚本语法自检通过"
+
+      - name: Auto approve when CI passes
+        shell: bash
+        env:
+          GITHUB_TOKEN: ${{ github.token }}
+          REVIEW_TOKEN: ${{ secrets.REVIEW_GITEA_TOKEN }}
+          PR_NUMBER: ${{ github.event.pull_request.number }}
+          PR_HEAD_SHA: ${{ github.event.pull_request.head.sha }}
+        run: |
+          bash scripts/ci/auto_approve.sh
+      - name: Report CI trace
+        if: always()
+        shell: sh
+        env:
+          AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
+        run: |
+          STATUS="ok"
+          [ ${{ job.status }} = "success" ] || STATUS="error"
+          START_TIME=""
+          [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
+          python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true
+
+  auto-merge:
+    name: Auto Merge on CI Green + Approved
+    runs-on: ci-check
+    if: github.event_name == 'pull_request' && !github.event.pull_request.draft && github.event.pull_request.base.ref == 'develop'
+    timeout-minutes: 3  # 短作业模式：检查一次，不满足就退出，由pr-auto-scan每5分钟定时兜底
+    steps:
+      - name: Checkout code
+        shell: sh
+        env:
+          GITHUB_TOKEN: ${{ github.token }}
+        run: |
+          curl -sH "Authorization: token $GITHUB_TOKEN"             "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/step_checkout.sh?ref=${GITHUB_SHA}"             | bash
+
+      - name: "🔍 脚本语法自检（防止脚本bug导致所有PR挂掉）"
+        shell: bash
+        run: |
+          echo "=== CI脚本语法自检 ==="
+          ERROR=0
+          for f in scripts/ci/*.sh; do
+            [ -f "$f" ] || continue
+            if ! bash -n "$f" 2>&1; then
+              echo "FAIL: $f"
+              ERROR=1
+            fi
+          done
+          for f in scripts/ci/*.py; do
+            [ -f "$f" ] || continue
+            if ! python3 -m py_compile "$f" 2>&1; then
+              echo "FAIL: $f"
+              ERROR=1
+            fi
+          done
+          if [ "$ERROR" -ne 0 ]; then
+            echo "❌ 脚本语法自检失败"
+            exit 1
+          fi
+          echo "✅ 所有CI脚本语法自检通过"
+
+      - name: Auto merge when CI passes and approved
+        shell: bash
+        env:
+          GITHUB_TOKEN: ${{ github.token }}
+          MERGE_TOKEN: ${{ secrets.REVIEW_GITEA_TOKEN }}
+          PR_NUMBER: ${{ github.event.pull_request.number }}
+          PR_HEAD_SHA: ${{ github.event.pull_request.head.sha }}
+          BASE_REF: ${{ github.event.pull_request.base.ref }}
+        run: |
+          bash scripts/ci/auto_merge.sh
+      - name: Report CI trace
+        if: always()
+        shell: sh
+        env:
+          AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
+        run: |
+          STATUS="ok"
+          [ ${{ job.status }} = "success" ] || STATUS="error"
+          START_TIME=""
+          [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
+          python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true

.gitea/workflows/preview-cleanup.yml

@@ -0,0 +1,207 @@
+name: Preview Cleanup
+on:
+  pull_request:
+    types:
+    - closed
+    branches:
+    - main
+    - develop
+permissions:
+  contents: read
+  pull-requests: write
+jobs:
+  cleanup-preview:
+    name: Cleanup Preview Environment
+    runs-on: runtime-builder
+    timeout-minutes: 10
+    steps:
+    - name: Checkout code
+      shell: sh
+      env:
+        GITHUB_TOKEN: ${{ github.token }}
+      run: |
+        set -eu
+        python3 - <<'PY'
+        import io, os, tarfile, time, urllib.request, urllib.error
+        url = f"{os.environ['GITHUB_API_URL']}/repos/{os.environ['GITHUB_REPOSITORY']}/archive/{os.environ['GITHUB_SHA']}.tar.gz"
+        request = urllib.request.Request(url, headers={"Authorization": f"token {os.environ['GITHUB_TOKEN']}"})
+        last_err = None
+        for attempt in range(5):
+            try:
+                with urllib.request.urlopen(request, timeout=120) as response:
+                    archive = response.read()
+                break
+            except urllib.error.HTTPError as e:
+                last_err = e
+                if e.code >= 500 and attempt < 4:
+                    wait = 2 ** attempt
+                    print(f"Checkout HTTP {e.code}, retrying in {wait}s (attempt {attempt+1}/5)...")
+                    time.sleep(wait)
+                    continue
+                raise
+            except Exception as e:
+                last_err = e
+                if attempt < 4:
+                    wait = 2 ** attempt
+                    print(f"Checkout error: {e}, retrying in {wait}s (attempt {attempt+1}/5)...")
+                    time.sleep(wait)
+                    continue
+                raise
+        else:
+            raise last_err
+        with tarfile.open(fileobj=io.BytesIO(archive), mode='r:gz') as tar:
+            root_prefix = tar.getmembers()[0].name.split('/', 1)[0] + '/'
+            for member in tar.getmembers():
+                name = member.name
+                if name == root_prefix[:-1]:
+                    continue
+                if name.startswith(root_prefix):
+                    member.name = name[len(root_prefix):]
+                if member.name:
+                    tar.extract(member, '.')
+        PY
+
+    - name: Extract PR number
+      shell: sh
+      run: |
+        set -eu
+        # 优先从event payload中读取（兼容所有PR事件类型）
+        if [ -n "${GITHUB_EVENT_PATH:-}" ] && [ -f "$GITHUB_EVENT_PATH" ]; then
+          PR_NUMBER=$(python3 -c "import json,sys; print(json.load(sys.stdin).get('number',''))" < "$GITHUB_EVENT_PATH")
+        fi
+        # fallback: 从GITHUB_REF中提取
+        if [ -z "${PR_NUMBER:-}" ]; then
+          PR_NUMBER=$(echo "$GITHUB_REF" | sed -n 's|refs/pull/\([0-9]*\)/.*|\1|p')
+        fi
+        # 再fallback: 兼容纯数字ref
+        if [ -z "${PR_NUMBER:-}" ] || ! echo "$PR_NUMBER" | grep -qE '^[0-9]+$'; then
+          echo "WARNING: Could not extract PR number cleanly, using raw ref suffix"
+          PR_NUMBER=$(echo "$GITHUB_REF" | sed 's|refs/pull/||; s|/.*||')
+        fi
+        echo "PR_NUMBER=$PR_NUMBER" >> $GITHUB_ENV
+        echo "PR number: $PR_NUMBER"
+        echo "Preview dir: /var/www/preview/pr-${PR_NUMBER}"
+
+    - name: Install SSH client
+      shell: sh
+      run: |
+        set -eu
+        # 先检查是否已存在ssh
+        if command -v ssh >/dev/null 2>&1 && command -v ssh-keyscan >/dev/null 2>&1; then
+          echo "SSH client already available: $(ssh -V 2>&1)"
+          exit 0
+        fi
+        # 尝试多种包管理器安装
+        if command -v apk >/dev/null 2>&1; then
+          apk add --no-cache openssh-client >/dev/null 2>&1
+          echo "openssh-client installed via apk"
+        elif command -v apt-get >/dev/null 2>&1; then
+          apt-get update -qq && apt-get install -y -qq openssh-client >/dev/null 2>&1
+          echo "openssh-client installed via apt-get"
+        elif command -v yum >/dev/null 2>&1; then
+          yum install -y openssh-clients >/dev/null 2>&1
+          echo "openssh-client installed via yum"
+        elif command -v dnf >/dev/null 2>&1; then
+          dnf install -y openssh-clients >/dev/null 2>&1
+          echo "openssh-client installed via dnf"
+        else
+          echo "ERROR: No package manager found and ssh not pre-installed"
+          which ssh 2>/dev/null || echo "  ssh: not found"
+          which ssh-keyscan 2>/dev/null || echo "  ssh-keyscan: not found"
+          exit 1
+        fi
+
+    - name: Remove preview directory from server
+      shell: sh
+      env:
+        PREVIEW_SSH_HOST: ${{ secrets.PREVIEW_SSH_HOST }}
+        PREVIEW_SSH_USER: ${{ secrets.PREVIEW_SSH_USER }}
+        PREVIEW_SSH_PORT: ${{ secrets.PREVIEW_SSH_PORT }}
+        PREVIEW_SSH_KEY: ${{ secrets.PREVIEW_SSH_KEY }}
+      run: |
+        set -eux
+        preview_host="${PREVIEW_SSH_HOST:-47.98.113.167}"
+        preview_user="${PREVIEW_SSH_USER:-root}"
+        preview_port="${PREVIEW_SSH_PORT:-22222}"
+        preview_dir="/var/www/preview/pr-${PR_NUMBER}"
+
+        mkdir -p ~/.ssh
+
+        # 查找可用的SSH密钥（优先用 secret 里专门为 preview 配置的 key）
+        key_path=""
+        if [ -n "${PREVIEW_SSH_KEY:-}" ]; then
+          key_path="$HOME/.ssh/id_ed25519"
+          printf '%s\n' "$PREVIEW_SSH_KEY" > "$key_path"
+          chmod 600 "$key_path"
+          echo "Using key from PREVIEW_SSH_KEY secret"
+        elif [ -f /root/.ssh/xiaoxia_runtime_builder ]; then
+          key_path="/root/.ssh/xiaoxia_runtime_builder"
+          echo "Using key: $key_path (builder key)"
+        elif [ -f "$HOME/.ssh/xiaoxia_runtime_builder" ]; then
+          key_path="$HOME/.ssh/xiaoxia_runtime_builder"
+          echo "Using key: $key_path (home key)"
+        else
+          echo "ERROR: No SSH key available"
+          ls -la ~/.ssh/ 2>/dev/null || true
+          ls -la /root/.ssh/ 2>/dev/null || true
+          exit 1
+        fi
+
+        ssh-keyscan -p "$preview_port" -H "$preview_host" >> ~/.ssh/known_hosts 2>/dev/null
+        echo "SSH keyscan done"
+
+        # 测试SSH连接
+        ssh -p "$preview_port" -i "$key_path" -o StrictHostKeyChecking=no "${preview_user}@${preview_host}" "echo SSH_CONNECTION_OK && hostname"
+        echo "SSH connection verified"
+
+        # 检查目录是否存在
+        DIR_EXISTS=$(ssh -p "$preview_port" -i "$key_path" -o StrictHostKeyChecking=no "${preview_user}@${preview_host}" \
+          "if [ -d '${preview_dir}' ]; then echo 'yes'; else echo 'no'; fi")
+
+        if [ "$DIR_EXISTS" = "yes" ]; then
+          echo "Removing preview directory: ${preview_dir}"
+          ssh -p "$preview_port" -i "$key_path" -o StrictHostKeyChecking=no "${preview_user}@${preview_host}" \
+            "rm -rf ${preview_dir} && echo 'Preview directory removed successfully'"
+          echo "Cleanup completed: ${preview_dir}"
+        else
+          echo "Preview directory does not exist: ${preview_dir}, nothing to clean up"
+        fi
+
+    - name: Comment cleanup notice on PR
+      if: success()
+      shell: sh
+      env:
+        GITHUB_TOKEN: ${{ github.token }}
+      run: |
+        set -eu
+        # 从event payload读取PR号（最可靠）
+        if [ -n "${GITHUB_EVENT_PATH:-}" ] && [ -f "$GITHUB_EVENT_PATH" ]; then
+          PR_NUMBER=$(python3 -c "import json,sys; print(json.load(sys.stdin).get('number',''))" < "$GITHUB_EVENT_PATH")
+        else
+          PR_NUMBER=$(echo "$GITHUB_REF" | sed -n 's|refs/pull/\([0-9]*\)/.*|\1|p')
+        fi
+        export PR_NUMBER
+
+        COMMENT_BODY=$(python3 scripts/ci/preview_comment.py cleanup)
+
+        API_URL="${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/issues/${PR_NUMBER}/comments"
+        curl -s -X POST \
+          -H "Authorization: token ${GITHUB_TOKEN}" \
+          -H "Content-Type: application/json" \
+          -d "$COMMENT_BODY" \
+          "$API_URL" \
+          > /dev/null
+        echo "Cleanup comment posted"
+
+    - name: Report CI trace
+      if: always()
+      shell: sh
+      env:
+        AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
+      run: |
+        STATUS="ok"
+        [ ${{ job.status }} = "success" ] || STATUS="error"
+        START_TIME=""
+        [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
+        python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true
+

.gitea/workflows/preview-deploy.yml

@@ -0,0 +1,296 @@
+name: Preview Deploy
+on:
+  pull_request:
+    types:
+    - opened
+    - synchronize
+    - reopened
+    branches:
+    - main
+    - develop
+  workflow_dispatch:
+    inputs:
+      reason:
+        description: "触发原因"
+        required: false
+        default: "手动触发 - 预览环境补跑"
+permissions:
+  contents: read
+  pull-requests: write
+concurrency:
+  group: preview-deploy-${{ gitea.ref }}
+  cancel-in-progress: true
+jobs:
+  deploy-preview:
+    name: Deploy Preview Environment
+    runs-on: runtime-builder
+    timeout-minutes: 20
+    steps:
+    - name: Checkout code
+      shell: sh
+      env:
+        GITHUB_TOKEN: ${{ github.token }}
+      run: |
+        set -eu
+        python3 - <<'PY'
+        import io, os, tarfile, time, urllib.request, urllib.error
+        url = f"{os.environ['GITHUB_API_URL']}/repos/{os.environ['GITHUB_REPOSITORY']}/archive/{os.environ['GITHUB_SHA']}.tar.gz"
+        request = urllib.request.Request(url, headers={"Authorization": f"token {os.environ['GITHUB_TOKEN']}"})
+        last_err = None
+        for attempt in range(5):
+            try:
+                with urllib.request.urlopen(request, timeout=120) as response:
+                    archive = response.read()
+                break
+            except urllib.error.HTTPError as e:
+                last_err = e
+                if e.code >= 500 and attempt < 4:
+                    wait = 2 ** attempt
+                    print(f"Checkout HTTP {e.code}, retrying in {wait}s (attempt {attempt+1}/5)...")
+                    time.sleep(wait)
+                    continue
+                raise
+            except Exception as e:
+                last_err = e
+                if attempt < 4:
+                    wait = 2 ** attempt
+                    print(f"Checkout error: {e}, retrying in {wait}s (attempt {attempt+1}/5)...")
+                    time.sleep(wait)
+                    continue
+                raise
+        else:
+            raise last_err
+        with tarfile.open(fileobj=io.BytesIO(archive), mode='r:gz') as tar:
+            root_prefix = tar.getmembers()[0].name.split('/', 1)[0] + '/'
+            for member in tar.getmembers():
+                name = member.name
+                if name == root_prefix[:-1]:
+                    continue
+                if name.startswith(root_prefix):
+                    member.name = name[len(root_prefix):]
+                if member.name:
+                    tar.extract(member, '.')
+        PY
+
+    - name: Record job start time
+      shell: sh
+      run: |
+        set -eu
+        echo "JOB_START_TIME=$(date +%s)" >> $GITHUB_ENV
+        echo "Job started at $(date)"
+
+    - name: Extract PR number
+      shell: sh
+      run: |
+        set -eu
+        PR_NUMBER=$(echo "$GITHUB_REF" | sed 's|refs/pull/||; s|/.*||')
+        echo "PR_NUMBER=$PR_NUMBER" >> $GITHUB_ENV
+        echo "PR number: $PR_NUMBER"
+        echo "PREVIEW_URL=https://pr-${PR_NUMBER}.preview.xiaoxiajianji.com" >> $GITHUB_ENV
+        echo "Preview URL: https://pr-${PR_NUMBER}.preview.xiaoxiajianji.com"
+
+    - name: Build frontend
+      shell: sh
+      run: |
+        set -eu
+        NPM_CACHE_VOLUME="xiaoxia-npm-cache"
+        if ! docker volume inspect "$NPM_CACHE_VOLUME" >/dev/null 2>&1; then
+          docker volume create "$NPM_CACHE_VOLUME" >/dev/null
+          echo "Created npm cache volume: $NPM_CACHE_VOLUME"
+        fi
+
+        docker run --rm \
+          -v "$PWD:/workspace" \
+          -v "$NPM_CACHE_VOLUME:/workspace/apps/web/node_modules" \
+          -w /workspace/apps/web \
+          -e VITE_API_URL=https://staging-api.xiaoxiajianji.com \
+          docker.m.daocloud.io/library/node:20 \
+          sh -lc '
+            PACKAGE_LOCK_HASH=$(md5sum package-lock.json 2>/dev/null | cut -d" " -f1)
+            CACHE_HASH_FILE="node_modules/.package-lock-hash"
+            CACHE_VALID=false
+            if [ -f "$CACHE_HASH_FILE" ] && [ "$(cat "$CACHE_HASH_FILE")" = "$PACKAGE_LOCK_HASH" ] && [ -x "node_modules/.bin/vite" ] && [ -x "node_modules/.bin/tsc" ]; then
+              CACHE_VALID=true
+              echo "Cache hit: dependencies valid, skipping npm install"
+            fi
+            if [ "$CACHE_VALID" = "false" ]; then
+              echo "Cache miss or invalid: running npm install..."
+              if ! npm install --include=dev; then
+                echo "npm install failed, cleaning node_modules and retrying..."
+                rm -rf node_modules
+                mkdir -p node_modules
+                npm install --include=dev
+              fi
+              echo "$PACKAGE_LOCK_HASH" > "$CACHE_HASH_FILE"
+              echo "Dependencies installed, cache updated"
+            fi
+            echo "Running TypeScript check..."
+            npx --no-install tsc
+            echo "Running Vite build..."
+            npx --no-install vite build
+            echo "Build completed successfully"
+            ls -la dist/
+          '
+
+    - name: Install SSH client and rsync
+      shell: sh
+      run: |
+        set -eu
+        if command -v apk >/dev/null 2>&1; then
+          apk add --no-cache openssh-client rsync >/dev/null 2>&1
+        elif command -v apt-get >/dev/null 2>&1; then
+          apt-get update -qq && apt-get install -y -qq openssh-client rsync >/dev/null 2>&1
+        elif command -v yum >/dev/null 2>&1; then
+          yum install -y openssh-clients rsync >/dev/null 2>&1
+        else
+          echo "ERROR: No package manager found"
+          exit 1
+        fi
+        echo "openssh-client and rsync installed"
+
+    - name: Deploy preview to server
+      shell: sh
+      env:
+        PREVIEW_SSH_HOST: ${{ secrets.PREVIEW_SSH_HOST }}
+        PREVIEW_SSH_USER: ${{ secrets.PREVIEW_SSH_USER }}
+        PREVIEW_SSH_PORT: ${{ secrets.PREVIEW_SSH_PORT }}
+        PREVIEW_SSH_KEY: ${{ secrets.PREVIEW_SSH_KEY }}
+      run: |
+        set -eux
+        preview_host="${PREVIEW_SSH_HOST:-47.98.113.167}"
+        preview_user="${PREVIEW_SSH_USER:-root}"
+        preview_port="${PREVIEW_SSH_PORT:-22222}"
+        preview_dir="/var/www/preview/pr-${PR_NUMBER}"
+
+        mkdir -p ~/.ssh
+
+        # 查找可用的SSH密钥（优先用 secret 里专门为 preview 配置的 key）
+        key_path=""
+        if [ -n "${PREVIEW_SSH_KEY:-}" ]; then
+          key_path="$HOME/.ssh/id_ed25519"
+          printf '%s\n' "$PREVIEW_SSH_KEY" > "$key_path"
+          chmod 600 "$key_path"
+          echo "Using key from PREVIEW_SSH_KEY secret"
+        elif [ -f /root/.ssh/xiaoxia_runtime_builder ]; then
+          key_path="/root/.ssh/xiaoxia_runtime_builder"
+          echo "Using key: $key_path (builder key)"
+        elif [ -f "$HOME/.ssh/xiaoxia_runtime_builder" ]; then
+          key_path="$HOME/.ssh/xiaoxia_runtime_builder"
+          echo "Using key: $key_path (home key)"
+        else
+          echo "ERROR: No SSH key available"
+          ls -la ~/.ssh/ 2>/dev/null || true
+          ls -la /root/.ssh/ 2>/dev/null || true
+          exit 1
+        fi
+
+        # SSH密钥完整性自检
+        if ! ssh-keygen -y -f "$key_path" > /dev/null 2>&1; then
+          echo "ERROR: SSH密钥损坏（private key contents do not match public）"
+          echo "请检查 PREVIEW_SSH_KEY secret 中的私钥是否完整正确"
+          echo "私钥文件大小: $(wc -c < "$key_path") 字节"
+          head -2 "$key_path"
+          exit 1
+        fi
+        echo "SSH key integrity check passed"
+
+        ssh-keyscan -p "$preview_port" -H "$preview_host" >> ~/.ssh/known_hosts 2>/dev/null
+        echo "SSH keyscan done"
+
+        # 测试SSH连接
+        ssh -p "$preview_port" -i "$key_path" -o StrictHostKeyChecking=no "${preview_user}@${preview_host}" "echo SSH_CONNECTION_OK && hostname"
+        echo "SSH connection verified"
+
+        # 创建预览目录并上传文件
+        ssh -p "$preview_port" -i "$key_path" -o StrictHostKeyChecking=no "${preview_user}@${preview_host}" \
+          "mkdir -p ${preview_dir} && echo 'Preview directory created: ${preview_dir}'"
+
+        # 使用rsync上传dist目录内容
+        rsync -avz --delete -e "ssh -p ${preview_port} -i ${key_path} -o StrictHostKeyChecking=no" \
+          apps/web/dist/ \
+          "${preview_user}@${preview_host}:${preview_dir}/"
+
+        echo "Preview deployed to: ${preview_dir}"
+        echo "Preview URL: https://pr-${PR_NUMBER}.preview.xiaoxiajianji.com"
+
+    - name: Comment preview link on PR
+      shell: sh
+      env:
+        GITHUB_TOKEN: ${{ github.token }}
+      run: |
+        set -eu
+        PR_NUMBER=$(echo "$GITHUB_REF" | sed 's|refs/pull/||; s|/.*||')
+        PREVIEW_URL="https://pr-${PR_NUMBER}.preview.xiaoxiajianji.com"
+        export PR_NUMBER PREVIEW_URL
+
+        COMMENT_BODY=$(python3 scripts/ci/preview_comment.py deploy)
+
+        API_URL="${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/issues/${PR_NUMBER}/comments"
+
+        EXISTING_COMMENT_ID=$(curl -s -H "Authorization: token ${GITHUB_TOKEN}" "$API_URL" | python3 -c "
+        import sys, json
+        try:
+            for c in json.load(sys.stdin):
+                if '预览环境已部署' in c.get('body', ''):
+                    print(c['id'])
+                    break
+        except Exception:
+            pass
+        ")
+
+        if [ -n "$EXISTING_COMMENT_ID" ]; then
+          curl -s -X PATCH \
+            -H "Authorization: token ${GITHUB_TOKEN}" \
+            -H "Content-Type: application/json" \
+            -d "$COMMENT_BODY" \
+            "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/issues/comments/${EXISTING_COMMENT_ID}" \
+            > /dev/null
+          echo "Comment updated"
+        else
+          curl -s -X POST \
+            -H "Authorization: token ${GITHUB_TOKEN}" \
+            -H "Content-Type: application/json" \
+            -d "$COMMENT_BODY" \
+            "$API_URL" \
+            > /dev/null
+          echo "Comment posted"
+        fi
+
+    - name: Job duration summary
+      if: always()
+      shell: sh
+      run: |
+        set +eu
+        if [ -n "$JOB_START_TIME" ]; then
+          END_TIME=$(date +%s)
+          DURATION=$((END_TIME - JOB_START_TIME))
+          MINS=$((DURATION / 60))
+          SECS=$((DURATION % 60))
+          echo "JOB_DURATION_SECONDS=$DURATION" >> $GITHUB_ENV
+          echo "=== Job Duration: ${MINS}m${SECS}s ==="
+        else
+          echo "JOB_DURATION_SECONDS=0" >> $GITHUB_ENV
+          echo "=== Job Duration: unknown ==="
+        fi
+
+    - name: Notify on failure
+      continue-on-error: true
+      if: failure()
+      shell: sh
+      env:
+        CI_NOTIFY_WEBHOOK: ${{ secrets.CI_NOTIFY_WEBHOOK }}
+      run: |
+        set +e
+        NOTIFY_MODE=failure JOB_NAME="Deploy Preview Environment" python3 scripts/ci_notify.py
+    - name: Report CI trace
+      if: always()
+      shell: sh
+      env:
+        AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
+      run: |
+        STATUS="ok"
+        [ ${{ job.status }} = "success" ] || STATUS="error"
+        START_TIME=""
+        [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
+        python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true
+

.gitea/workflows/tests.yml

@@ -1,107 +0,0 @@
-name: Tests
-
-on:
-  pull_request:
-    branches: [ main ]
-
-jobs:
-  test:
-    runs-on: runtime-builder
-
-    steps:
-    - name: Checkout code
-      shell: sh
-      run: |
-        set -eu
-        python - <<'PY'
-        import io
-        import os
-        import tarfile
-        import urllib.request
-
-        url = f"{os.environ['GITHUB_API_URL']}/repos/{os.environ['GITHUB_REPOSITORY']}/archive/{os.environ['GITHUB_SHA']}.tar.gz"
-        request = urllib.request.Request(url, headers={"Authorization": f"token {os.environ['GITHUB_TOKEN']}"})
-        with urllib.request.urlopen(request, timeout=120) as response:
-            archive = response.read()
-
-        with tarfile.open(fileobj=io.BytesIO(archive), mode='r:gz') as tar:
-            root_prefix = tar.getmembers()[0].name.split('/', 1)[0] + '/'
-            for member in tar.getmembers():
-                name = member.name
-                if name == root_prefix[:-1]:
-                    continue
-                if name.startswith(root_prefix):
-                    member.name = name[len(root_prefix):]
-                if member.name:
-                    tar.extract(member, '.')
-        PY
-
-    - name: Show Python version
-      shell: sh
-      run: |
-        set -eu
-        python --version
-        python -m pip --version
-
-    - name: Install dependencies
-      shell: sh
-      run: |
-        set -eu
-        python -m pip install --upgrade pip -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com
-        python -m pip install -r requirements.txt -r requirements-dev.txt -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com
-
-    - name: Run tests
-      shell: sh
-      run: |
-        set -eu
-        PYTHONPATH="$PWD/apps/api:$PWD" python -m pytest tests/unit -q
-
-  lint:
-    runs-on: runtime-builder
-
-    steps:
-    - name: Checkout code
-      shell: sh
-      run: |
-        set -eu
-        python - <<'PY'
-        import io
-        import os
-        import tarfile
-        import urllib.request
-
-        url = f"{os.environ['GITHUB_API_URL']}/repos/{os.environ['GITHUB_REPOSITORY']}/archive/{os.environ['GITHUB_SHA']}.tar.gz"
-        request = urllib.request.Request(url, headers={"Authorization": f"token {os.environ['GITHUB_TOKEN']}"})
-        with urllib.request.urlopen(request, timeout=120) as response:
-            archive = response.read()
-
-        with tarfile.open(fileobj=io.BytesIO(archive), mode='r:gz') as tar:
-            root_prefix = tar.getmembers()[0].name.split('/', 1)[0] + '/'
-            for member in tar.getmembers():
-                name = member.name
-                if name == root_prefix[:-1]:
-                    continue
-                if name.startswith(root_prefix):
-                    member.name = name[len(root_prefix):]
-                if member.name:
-                    tar.extract(member, '.')
-        PY
-
-    - name: Install dependencies
-      shell: sh
-      run: |
-        set -eu
-        python -m pip install --upgrade pip -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com
-        python -m pip install -r requirements.txt -r requirements-dev.txt -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com
-
-    - name: Run Black (check only)
-      shell: sh
-      run: |
-        set -eu
-        python -m black --check alembic apps packages tests scripts
-
-    - name: Run Flake8
-      shell: sh
-      run: |
-        set -eu
-        python -m flake8 apps packages tests --count --statistics

.gitea/workflows/worker-base-image.yml

@@ -0,0 +1,103 @@
+name: Worker Base Image Build
+
+on:
+  push:
+    branches:
+      - develop
+      - main
+    paths:
+      - 'requirements-base.txt'
+      - 'requirements-worker.txt'
+      - 'infra/docker/worker-base-builder.Dockerfile'
+      - 'infra/docker/worker-base-runtime.Dockerfile'
+  workflow_dispatch:  # 支持手动触发
+
+jobs:
+  build-worker-base:
+    name: Build Worker Base Images
+    runs-on: runtime-builder
+    timeout-minutes: 30
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+        - name: builder
+          dockerfile: infra/docker/worker-base-builder.Dockerfile
+          image_name: worker-base-builder
+          cache_name: worker-base-builder-cache
+        - name: runtime
+          dockerfile: infra/docker/worker-base-runtime.Dockerfile
+          image_name: worker-base-runtime
+          cache_name: worker-base-runtime-cache
+    steps:
+    - name: Checkout code
+      shell: sh
+      env:
+        GITHUB_TOKEN: ${{ github.token }}
+      run: |
+        curl -sH "Authorization: token $GITHUB_TOKEN"           "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/step_checkout.sh?ref=${GITHUB_SHA}"           | bash
+
+    - name: Docker login to Registry
+      shell: sh
+      env:
+        ACR_USERNAME: ${{ secrets.ACR_USERNAME }}
+        ACR_PASSWORD: ${{ secrets.ACR_PASSWORD }}
+        GITEA_REGISTRY_USER: xiaoxia
+        GITEA_REGISTRY_TOKEN: ${{ secrets.REGISTRY_TOKEN }}
+      run: |
+        set -eu
+        for i in 1 2 3; do
+          echo "=== Docker login 尝试 $i/3 ==="
+          if printf '%s' "${ACR_PASSWORD}" | docker login xiaoxia-registry.cn-hangzhou.cr.aliyuncs.com -u "${ACR_USERNAME}" --password-stdin              && docker login git.xiaoxiajianji.com -u "${GITEA_REGISTRY_USER}" -p "${GITEA_REGISTRY_TOKEN}"; then
+            echo "✅ Docker login successful"
+            break
+          fi
+          echo "❌ Docker login 失败（尝试 $i/3），5s 后重试..."
+          sleep 5
+        done
+
+    - name: Setup buildx builder
+      shell: sh
+      run: |
+        set -eu
+        BUILDER_NAME="ci-builder-${GITHUB_RUN_ID}-${{ matrix.name }}"
+        if ! docker buildx inspect "$BUILDER_NAME" > /dev/null 2>&1; then
+          docker buildx create --use --name "$BUILDER_NAME" --driver docker-container
+          echo "Created $BUILDER_NAME"
+        else
+          docker buildx use "$BUILDER_NAME"
+          echo "Using existing $BUILDER_NAME"
+        fi
+        docker buildx inspect --bootstrap
+
+    - name: Build and push base image
+      shell: sh
+      run: |
+        set -eu
+        REGISTRY="xiaoxia-registry.cn-hangzhou.cr.aliyuncs.com/xiaoxiakeji"
+        IMAGE_TAG="${REGISTRY}/${{ matrix.image_name }}:latest"
+        SAFE_REF_NAME=$(echo "${GITHUB_REF_NAME}" | tr '/' '-')
+        CACHE_REF="${REGISTRY}/${{ matrix.cache_name }}:${SAFE_REF_NAME}"
+        
+        echo "=== Building ${{ matrix.name }} base image ==="
+        echo "Image: ${IMAGE_TAG}"
+        echo "Cache: ${CACHE_REF}"
+        
+        # 用通用构建脚本
+        bash scripts/ci/docker_build_push.sh           ${{ matrix.dockerfile }}           "${IMAGE_TAG}"           "${CACHE_REF}"
+        
+        # 同时推送到 Gitea Packages 作为备份（可选）
+        GITEA_IMAGE="git.xiaoxiajianji.com/xiaoxia-saas/${{ matrix.image_name }}:latest"
+        docker tag "${IMAGE_TAG}" "${GITEA_IMAGE}"
+        docker push "${GITEA_IMAGE}" || echo "Gitea Packages push failed (non-fatal)"
+        
+        echo ""
+        echo "✅ ${{ matrix.name }} base image built and pushed"
+
+    - name: Cleanup buildx builder
+      if: always()
+      shell: sh
+      run: |
+        docker buildx rm "ci-builder-${GITHUB_RUN_ID}-${{ matrix.name }}" 2>/dev/null || true
+        docker buildx prune -f 2>/dev/null || true
+        echo "Builder cleanup done"

.gitignore

@@ -47,5 +47,4 @@ build/
 # Tracker temp files
 tracker_tasks.json
 
-# Schema metadata snapshot
-schema-metadata-snapshot.json
+frontend-v21-ui-prototype-final.html

CHANGELOG.md

@@ -1,19 +1,276 @@
-# Changelog
+## [v0.1.110] - 2026-07-03
 
-All notable changes to this project will be documented in this file.
+### 🔒 安全修复
 
-The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
-and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+- 注册登录接口添加 RateLimitMiddleware 防止暴力破解
+- JWT logout 黑名单机制，防止令牌重放攻击
+- 生产环境禁用 Swagger 文档防止信息泄露
+- `/metrics` 端点添加 Bearer Token 认证
+- 禁用 SVG 上传防止 XSS 风险
+- 删除 `decode_token_unsafe()` 方法，消除不安全的 JWT 解码
+- 移除遗留 `tasks.py` 消除 Celery 任务名冲突
+- 清理全局 `except:pass`（22处）改为 `logger.warning` 记录异常
+
+### ✨ 功能
+
+- 添加剪辑计划时间线场景 API (`GET /edit-plans/{id}/timeline`)
+- 前端对接真实 API 替换 mock 数据
+
+### 🐛 Bug 修复
+
+- **[P1]** 修复登录故障 — `password_hasher` 导入错误
+- 订阅续费事务修复 — 支付回调在数据库事务中更新订阅状态
+- 账单返回空数组修复 — 从数据库查询账单记录
+- 修复 `Image.open()` 资源泄漏
+- 清理已移除 workspace 概念的残留引用
+- 修复 AssetLibrary/TemplateLibrary 类型错误
+- 修复前端 workspace 残留导致项目创建失败
+- 永久修复 nginx `proxy_pass` 配置
+- 添加 Docker DNS resolver 防止 API 容器重启后 502
+- 修复 worker healthcheck YAML 语法
+- 修复 204 响应体断言崩溃
+- 修复 Alembic 元数据漂移检测
+- 修复 migration 009 DEFAULT 表达式 PostgreSQL 兼容性
+
+### 🔄 重构与清理
+
+- 后端代码清理 — 移除死代码和无用文件
+- 前端代码清理 — 移除无用代码和遗留 demo
+- 代码精简优化 — 移除无用代码和重复定义
+- 后端代码 black/isort 格式化
+
+### 🧪 测试
+
+- 完善 E2E 错误场景测试，Playwright 接入 CI
+- API 集成测试补充（145 项通过）
+- 添加核心流程 E2E 测试
+
+### 🚀 CI/CD & 基础设施
+
+- Validate 阶段添加 PostgreSQL 服务支持
+- 所有 workflow checkout 添加 5 次指数退避重试
+- 启用 BuildKit 分布式缓存 + Gitea Registry 优化构建速度
+- Deploy 阶段全面修复（E2E 服务器/Worker venv/Registry 登录）
+- Docker 网络隔离 staging/production 环境
+- 修复 CI 代码质量检查（black/flake8/bandit）
+
+---
+
+## [v0.1.88] - 2026-06-29
+
+### Phase 2 前端优化 - 完成 ✅
+
+**前端交互全面优化:**
+
+- 素材上传添加 project_id 参数
+- Drager 组件显示上传列表
+- 按钮防重复提交
+- 前端交互状态反馈补充（P0 第一批）
+
+---
+
+## [v0.1.87] - 2026-06-29
+
+### Bug 修复
+
+- Docker compose 修复 mem_limit 冲突
+
+---
+
+## [v0.1.86] - 2026-06-29
+
+### CI/CD 优化
+
+- CI 优化
+
+---
+
+## [v0.1.85] - 2026-06-29
+
+### CI/CD 优化
+
+- CI 优化
+
+---
+
+## [v0.1.84] - 2026-06-29
+
+### CI/CD 优化
+
+- CI runner label 匹配修复
+
+---
+
+## [v0.1.83] - 2026-06-29
+
+### CI/CD 优化
+
+- CI SSH debug 修正
+
+---
+
+## [v0.1.82] - 2026-06-29
+
+### CI/CD 优化
+
+- CI SSH debug 修正
+
+---
+
+## [v0.1.81] - 2026-06-29
+
+### CI/CD 优化
+
+- CI runner label 匹配修复
+
+---
+
+## [v0.1.80] - 2026-06-28
+
+### Bug 修复
+
+- 修复 redirect_slashes + 标题字段匹配
+
+---
+
+## [v0.1.79] - 2026-06-28
+
+### Deployment
+
+- Re-trigger deployment
+
+---
+
+## [v0.1.78] - 2026-06-28
+
+### Bug 修复
+
+- 修复 500 错误
+- CORS 配置修复
+- redirect_slashes 禁用
+
+---
+
+## [v0.1.77] - 2026-06-28
+
+### Bug 修复
+
+- 修复标题库新建/编辑 — 前后端字段名不匹配导致 422
+
+---
+
+### Phase 2 功能合并（v0.1.77 ~ v0.1.88）
+
+**新增功能 PR:**
+
+- PR#74: Phase 1 核心重构 — 标题库 API、配音库 API、去 Project 层清理
+- PR#75: Phase 2 查重功能前端页面
+- PR#76: Phase 2 查重功能后端 API（5 个端点）
+- PR#77: Phase 2 订阅管理前端页面
+- PR#78: Phase 2 订阅管理后端 API（5 个端点）
+- PR#79: 修复一键生成页面废弃 API 调用
+- PR#80: 回退域对象 extra_meta → metadata
+- PR#81: 删除查重 API 错误的 204 返回
+- PR#82: 查重上传接口错误信息不再泄露内部异常（安全审计）
+- PR#83: 订阅 + 查重单元测试（63 用例）
+- PR#84: 订阅管理前端对接真实 API
+- PR#85: 禁用 redirect_slashes 修复 307 重定向
+- PR#90: 标题库字段名修复
+- PR#91: 标题/配音创建 500 修复 + CORS
+- PR#94: 素材库新建自动获取默认 project_id
+- PR#97: 前端交互状态反馈全面补充
+
+---
+
+
+- Docker compose 修复 mem_limit 冲突
+
+---
+
+## [v0.1.86] - 2026-06-29
+
+### CI/CD 优化
+
+- CI 优化
+
+---
+
+## [v0.1.85] - 2026-06-29
+
+### CI/CD 优化
+
+- CI 优化
+
+---
+
+## [v0.1.84] - 2026-06-29
+
+### CI/CD 优化
+
+- CI runner label 匹配修复
+
+---
+
+## [v0.1.83] - 2026-06-29
+
+### CI/CD 优化
+
+- CI SSH debug 修正
+
+---
+
+## [v0.1.82] - 2026-06-29
+
+### CI/CD 优化
+
+- CI SSH debug 修正
+
+---
+
+## [v0.1.81] - 2026-06-29
+
+### CI/CD 优化
+
+- CI runner label 匹配修复
+
+---
+
+## [v0.1.80] - 2026-06-28
+
+### Bug 修复
+
+- 修复 redirect_slashes + 标题字段匹配
+
+---
+
+## [v0.1.79] - 2026-06-28
+
+### Deployment
+
+- Re-trigger deployment
+
+---
+
+## [v0.1.78] - 2026-06-28
+
+### Bug 修复
+
+- 修复 500 错误
+- CORS 配置修复
+- redirect_slashes 禁用
+
+---
+
+## [v0.1.77] - 2026-06-28
+
+### Bug 修复
+
+- 修复标题库新建/编辑 — 前后端字段名不匹配导致 422
+
+---
 
 ## [Unreleased]
 
-## [1.2.0] - 2026-06-19
-
-### Phase 7: 核心视频剪辑业务 - 完成 ✅
-
-**完成进度:** 100%  
-**状态:** 已完成并验证
-
 #### Added
 
 **素材管理:**

alembic/versions/007_add_editing_mode.py

@@ -4,25 +4,26 @@ Revision ID: 007
 Revises: 006
 Create Date: 2026-06-26
 """
-from alembic import op
+
 import sqlalchemy as sa
 
+from alembic import op
+
 # revision identifiers
-revision = '007'
-down_revision = '006'
+revision = "007"
+down_revision = "006"
 branch_labels = None
 depends_on = None
 
 
 def upgrade() -> None:
     op.add_column(
-        'generation_tasks',
-        sa.Column('editing_mode', sa.String(20), nullable=False, server_default='one_take')
+        "generation_tasks", sa.Column("editing_mode", sa.String(20), nullable=False, server_default="one_take")
     )
     # 添加索引以支持查询
-    op.create_index('ix_generation_tasks_editing_mode', 'generation_tasks', ['editing_mode'])
+    op.create_index("ix_generation_tasks_editing_mode", "generation_tasks", ["editing_mode"])
 
 
 def downgrade() -> None:
-    op.drop_index('ix_generation_tasks_editing_mode', table_name='generation_tasks')
-    op.drop_column('generation_tasks', 'editing_mode')
+    op.drop_index("ix_generation_tasks_editing_mode", table_name="generation_tasks")
+    op.drop_column("generation_tasks", "editing_mode")

alembic/versions/008_add_video_dedup_fields.py

@@ -4,6 +4,7 @@ Revision ID: 008
 Revises: 007
 Create Date: 2024-06-26
 """
+
 import sqlalchemy as sa
 
 from alembic import op
@@ -16,20 +17,11 @@ depends_on = None
 
 def upgrade() -> None:
     # Add video_fingerprint column as JSON text
-    op.add_column(
-        "generated_videos",
-        sa.Column("video_fingerprint", sa.Text(), nullable=True)
-    )
+    op.add_column("generated_videos", sa.Column("video_fingerprint", sa.Text(), nullable=True))
     # Add is_duplicate column
-    op.add_column(
-        "generated_videos",
-        sa.Column("is_duplicate", sa.Boolean(), nullable=False, server_default="false")
-    )
+    op.add_column("generated_videos", sa.Column("is_duplicate", sa.Boolean(), nullable=False, server_default="false"))
     # Add duplicate_of column for tracking original video
-    op.add_column(
-        "generated_videos",
-        sa.Column("duplicate_of", sa.String(32), nullable=True)
-    )
+    op.add_column("generated_videos", sa.Column("duplicate_of", sa.String(32), nullable=True))
 
 
 def downgrade() -> None:

alembic/versions/009_remove_workspace_concept.py

@@ -11,10 +11,12 @@ This migration:
 4. Removes workspace_id from all tables that had it
 5. Drops workspace-related tables: workspaces, workspace_members, workspace_invitations
 """
-from alembic import op
+
 import sqlalchemy as sa
 from sqlalchemy import text
 
+from alembic import op
+
 # revision identifiers
 revision = "009"
 down_revision = "008"
@@ -24,15 +26,15 @@ depends_on = None
 
 def upgrade() -> None:
     conn = op.get_bind()
-    
+
     # Step 1: Add subscription/quota fields to users table
     conn.execute(text("""
         ALTER TABLE users
-        ADD COLUMN IF NOT EXISTS subscription_plan VARCHAR(20) NOT NULL DEFAULT free
+        ADD COLUMN IF NOT EXISTS subscription_plan VARCHAR(20) NOT NULL DEFAULT 'free'
     """))
     conn.execute(text("""
         ALTER TABLE users
-        ADD COLUMN IF NOT EXISTS subscription_status VARCHAR(20) NOT NULL DEFAULT active
+        ADD COLUMN IF NOT EXISTS subscription_status VARCHAR(20) NOT NULL DEFAULT 'active'
     """))
     conn.execute(text("""
         ALTER TABLE users
@@ -50,7 +52,7 @@ def upgrade() -> None:
         ALTER TABLE users
         ADD COLUMN IF NOT EXISTS used_storage_gb FLOAT NOT NULL DEFAULT 0
     """))
-    
+
     # Step 2: Copy subscription data from workspaces to users
     conn.execute(text("""
         UPDATE users SET
@@ -63,7 +65,7 @@ def upgrade() -> None:
         FROM workspaces w
         WHERE w.owner_user_id = users.id
     """))
-    
+
     # Step 3: Add owner_user_id and shared_users to projects table
     conn.execute(text("""
         ALTER TABLE projects
@@ -73,7 +75,7 @@ def upgrade() -> None:
         ALTER TABLE projects
         ADD COLUMN IF NOT EXISTS shared_users JSON
     """))
-    
+
     # Step 4: Migrate workspace_id to owner_user_id (from workspace_members where role=owner)
     conn.execute(text("""
         UPDATE projects SET
@@ -82,13 +84,13 @@ def upgrade() -> None:
         WHERE wm.workspace_id = projects.workspace_id
         AND wm.role = 'owner'
     """))
-    
+
     # Set shared_users to empty array for all projects
     conn.execute(text("""
         UPDATE projects SET shared_users = '[]'::json
         WHERE shared_users IS NULL
     """))
-    
+
     # Step 5: Remove workspace_id from all tables
     tables_with_workspace_id = [
         "asset_libraries",
@@ -104,12 +106,12 @@ def upgrade() -> None:
         "tasks",
         "task_issues",
     ]
-    
+
     for table in tables_with_workspace_id:
         conn.execute(text(f"""
             ALTER TABLE {table} DROP COLUMN IF EXISTS workspace_id
         """))
-    
+
     # Step 6: Drop workspace-related tables
     conn.execute(text("""
         DROP TABLE IF EXISTS workspace_invitations
@@ -120,7 +122,7 @@ def upgrade() -> None:
     conn.execute(text("""
         DROP TABLE IF EXISTS workspaces
     """))
-    
+
     # Step 7: Drop workspace_id from projects table
     conn.execute(text("""
         ALTER TABLE projects DROP COLUMN IF EXISTS workspace_id
@@ -129,15 +131,15 @@ def upgrade() -> None:
 
 def downgrade() -> None:
     conn = op.get_bind()
-    
+
     # Add back workspace tables (simplified - in real scenario would need full recreation)
     conn.execute(text("""
         CREATE TABLE IF NOT EXISTS workspaces (
             id VARCHAR(36) PRIMARY KEY,
             name VARCHAR(100) NOT NULL,
             owner_user_id VARCHAR(36) NOT NULL,
-            subscription_plan VARCHAR(20) NOT NULL DEFAULT free,
-            subscription_status VARCHAR(20) NOT NULL DEFAULT active,
+            subscription_plan VARCHAR(20) NOT NULL DEFAULT 'free',
+            subscription_status VARCHAR(20) NOT NULL DEFAULT 'active',
             subscription_expires_at TIMESTAMP,
             max_projects FLOAT NOT NULL DEFAULT 3,
             max_storage_gb FLOAT NOT NULL DEFAULT 10,
@@ -145,7 +147,7 @@ def downgrade() -> None:
             created_at TIMESTAMP NOT NULL DEFAULT NOW()
         )
     """))
-    
+
     conn.execute(text("""
         CREATE TABLE IF NOT EXISTS workspace_members (
             id VARCHAR(36) PRIMARY KEY,
@@ -157,7 +159,7 @@ def downgrade() -> None:
             UNIQUE(workspace_id, user_id)
         )
     """))
-    
+
     conn.execute(text("""
         CREATE TABLE IF NOT EXISTS workspace_invitations (
             id VARCHAR(36) PRIMARY KEY,
@@ -166,18 +168,18 @@ def downgrade() -> None:
             invitee_email VARCHAR(255) NOT NULL,
             role VARCHAR(20) NOT NULL,
             invitation_token VARCHAR(255) NOT NULL UNIQUE,
-            status VARCHAR(20) NOT NULL DEFAULT pending,
+            status VARCHAR(20) NOT NULL DEFAULT 'pending',
             expires_at TIMESTAMP,
             accepted_at TIMESTAMP,
             created_at TIMESTAMP NOT NULL DEFAULT NOW()
         )
     """))
-    
+
     # Add back workspace_id column to projects
     conn.execute(text("""
         ALTER TABLE projects ADD COLUMN workspace_id VARCHAR(32)
     """))
-    
+
     # Add back workspace_id columns to other tables
     tables_with_workspace_id = [
         "asset_libraries",
@@ -193,11 +195,11 @@ def downgrade() -> None:
         "tasks",
         "task_issues",
     ]
-    
+
     for table in tables_with_workspace_id:
         conn.execute(text(f"""
             ALTER TABLE {table} ADD COLUMN workspace_id VARCHAR(36)
         """))
-    
+
     # Note: This downgrade is incomplete - projects.owner_user_id data would need to be
     # converted back to workspace_ids, which requires reconstructing workspace records.

alembic/versions/010_add_foreign_key_constraints.py

@@ -1,190 +0,0 @@
-"""add foreign key constraints and missing indexes
-
-Revision ID: 010
-Revises: 009
-Create Date: 2026-06-28
-
-Summary:
-- Add ForeignKey constraints to all 35 logical FK columns across 16 tables
-- ON DELETE CASCADE for parent-child relationships (project → children)
-- ON DELETE SET NULL for optional user references (created_by, assignee, etc.)
-- Add missing indexes on 10 FK columns
-- Alter columns from NOT NULL to nullable where SET NULL is needed
-"""
-
-from alembic import op
-import sqlalchemy as sa
-
-revision = "010"
-down_revision = "009"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # ── projects ──────────────────────────────────────────────
-    op.create_foreign_key("fk_projects_owner_user_id", "projects", "users", ["owner_user_id"], ["id"], ondelete="CASCADE")
-
-    # ── asset_libraries ───────────────────────────────────────
-    op.create_foreign_key("fk_asset_libraries_project_id", "asset_libraries", "projects", ["project_id"], ["id"], ondelete="CASCADE")
-
-    # ── assets ────────────────────────────────────────────────
-    # uploaded_by_user_id: nullable=False → nullable=True for SET NULL
-    op.alter_column("assets", "uploaded_by_user_id", existing_type=sa.String(36), nullable=True)
-    op.create_foreign_key("fk_assets_project_id", "assets", "projects", ["project_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_assets_asset_library_id", "assets", "asset_libraries", ["asset_library_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_assets_uploaded_by_user_id", "assets", "users", ["uploaded_by_user_id"], ["id"], ondelete="SET NULL")
-    # Add missing index
-    op.create_index("ix_assets_uploaded_by_user_id", "assets", ["uploaded_by_user_id"])
-
-    # ── project_titles ────────────────────────────────────────
-    # created_by_user_id: nullable=False → nullable=True for SET NULL
-    op.alter_column("project_titles", "created_by_user_id", existing_type=sa.String(36), nullable=True)
-    op.create_foreign_key("fk_project_titles_project_id", "project_titles", "projects", ["project_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_project_titles_created_by_user_id", "project_titles", "users", ["created_by_user_id"], ["id"], ondelete="SET NULL")
-    # Add missing index
-    op.create_index("ix_project_titles_created_by_user_id", "project_titles", ["created_by_user_id"])
-
-    # ── edit_templates ────────────────────────────────────────
-    # created_by_user_id: nullable=False → nullable=True for SET NULL
-    op.alter_column("edit_templates", "created_by_user_id", existing_type=sa.String(32), nullable=True)
-    op.create_foreign_key("fk_edit_templates_project_id", "edit_templates", "projects", ["project_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_edit_templates_created_by_user_id", "edit_templates", "users", ["created_by_user_id"], ["id"], ondelete="SET NULL")
-
-    # ── edit_plans ────────────────────────────────────────────
-    # title_id: nullable=False → nullable=True for SET NULL
-    # created_by_user_id: nullable=False → nullable=True for SET NULL
-    op.alter_column("edit_plans", "title_id", existing_type=sa.String(32), nullable=True)
-    op.alter_column("edit_plans", "created_by_user_id", existing_type=sa.String(32), nullable=True)
-    op.create_foreign_key("fk_edit_plans_project_id", "edit_plans", "projects", ["project_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_edit_plans_template_id", "edit_plans", "edit_templates", ["template_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_edit_plans_asset_library_id", "edit_plans", "asset_libraries", ["asset_library_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_edit_plans_title_id", "edit_plans", "project_titles", ["title_id"], ["id"], ondelete="SET NULL")
-    op.create_foreign_key("fk_edit_plans_created_by_user_id", "edit_plans", "users", ["created_by_user_id"], ["id"], ondelete="SET NULL")
-
-    # ── edit_plan_clips ───────────────────────────────────────
-    op.create_foreign_key("fk_edit_plan_clips_edit_plan_id", "edit_plan_clips", "edit_plans", ["edit_plan_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_edit_plan_clips_asset_id", "edit_plan_clips", "assets", ["asset_id"], ["id"], ondelete="CASCADE")
-
-    # ── ingest_jobs ───────────────────────────────────────────
-    # result_asset_id: nullable=False → nullable=True for SET NULL
-    op.alter_column("ingest_jobs", "result_asset_id", existing_type=sa.String(32), nullable=True)
-    op.create_foreign_key("fk_ingest_jobs_project_id", "ingest_jobs", "projects", ["project_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_ingest_jobs_library_id", "ingest_jobs", "asset_libraries", ["library_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_ingest_jobs_result_asset_id", "ingest_jobs", "assets", ["result_asset_id"], ["id"], ondelete="SET NULL")
-
-    # ── classification_jobs ───────────────────────────────────
-    op.create_foreign_key("fk_classification_jobs_project_id", "classification_jobs", "projects", ["project_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_classification_jobs_asset_id", "classification_jobs", "assets", ["asset_id"], ["id"], ondelete="CASCADE")
-
-    # ── generation_tasks ──────────────────────────────────────
-    # edit_plan_id: nullable=False → nullable=True for SET NULL
-    # created_by_user_id: nullable=False → nullable=True for SET NULL
-    op.alter_column("generation_tasks", "edit_plan_id", existing_type=sa.String(32), nullable=True)
-    op.alter_column("generation_tasks", "created_by_user_id", existing_type=sa.String(32), nullable=True)
-    op.create_foreign_key("fk_generation_tasks_project_id", "generation_tasks", "projects", ["project_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_generation_tasks_asset_library_id", "generation_tasks", "asset_libraries", ["asset_library_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_generation_tasks_edit_plan_id", "generation_tasks", "edit_plans", ["edit_plan_id"], ["id"], ondelete="SET NULL")
-    op.create_foreign_key("fk_generation_tasks_created_by_user_id", "generation_tasks", "users", ["created_by_user_id"], ["id"], ondelete="SET NULL")
-
-    # ── generated_videos ──────────────────────────────────────
-    op.create_foreign_key("fk_generated_videos_project_id", "generated_videos", "projects", ["project_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_generated_videos_generation_task_id", "generated_videos", "generation_tasks", ["generation_task_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_generated_videos_duplicate_of", "generated_videos", "generated_videos", ["duplicate_of"], ["id"], ondelete="SET NULL")
-
-    # ── tasks ─────────────────────────────────────────────────
-    # parent_task_id: nullable=False → nullable=True for SET NULL
-    # assignee_user_id: nullable=False → nullable=True for SET NULL
-    op.alter_column("tasks", "parent_task_id", existing_type=sa.String(32), nullable=True)
-    op.alter_column("tasks", "assignee_user_id", existing_type=sa.String(32), nullable=True)
-    op.create_foreign_key("fk_tasks_project_id", "tasks", "projects", ["project_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_tasks_parent_task_id", "tasks", "tasks", ["parent_task_id"], ["id"], ondelete="SET NULL")
-    op.create_foreign_key("fk_tasks_assignee_user_id", "tasks", "users", ["assignee_user_id"], ["id"], ondelete="SET NULL")
-
-    # ── milestones ────────────────────────────────────────────
-    op.create_foreign_key("fk_milestones_project_id", "milestones", "projects", ["project_id"], ["id"], ondelete="CASCADE")
-
-    # ── task_issues ───────────────────────────────────────────
-    # created_by_user_id: nullable=False → nullable=True for SET NULL
-    op.alter_column("task_issues", "created_by_user_id", existing_type=sa.String(32), nullable=True)
-    op.create_foreign_key("fk_task_issues_task_id", "task_issues", "tasks", ["task_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_task_issues_project_id", "task_issues", "projects", ["project_id"], ["id"], ondelete="CASCADE")
-    op.create_foreign_key("fk_task_issues_created_by_user_id", "task_issues", "users", ["created_by_user_id"], ["id"], ondelete="SET NULL")
-
-
-def downgrade() -> None:
-    # ── task_issues ───────────────────────────────────────────
-    op.drop_constraint("fk_task_issues_created_by_user_id", "task_issues", type_="foreignkey")
-    op.drop_constraint("fk_task_issues_project_id", "task_issues", type_="foreignkey")
-    op.drop_constraint("fk_task_issues_task_id", "task_issues", type_="foreignkey")
-    op.alter_column("task_issues", "created_by_user_id", existing_type=sa.String(32), nullable=False)
-
-    # ── milestones ────────────────────────────────────────────
-    op.drop_constraint("fk_milestones_project_id", "milestones", type_="foreignkey")
-
-    # ── tasks ─────────────────────────────────────────────────
-    op.drop_constraint("fk_tasks_assignee_user_id", "tasks", type_="foreignkey")
-    op.drop_constraint("fk_tasks_parent_task_id", "tasks", type_="foreignkey")
-    op.drop_constraint("fk_tasks_project_id", "tasks", type_="foreignkey")
-    op.alter_column("tasks", "assignee_user_id", existing_type=sa.String(32), nullable=False)
-    op.alter_column("tasks", "parent_task_id", existing_type=sa.String(32), nullable=False)
-
-    # ── generated_videos ──────────────────────────────────────
-    op.drop_constraint("fk_generated_videos_duplicate_of", "generated_videos", type_="foreignkey")
-    op.drop_constraint("fk_generated_videos_generation_task_id", "generated_videos", type_="foreignkey")
-    op.drop_constraint("fk_generated_videos_project_id", "generated_videos", type_="foreignkey")
-
-    # ── generation_tasks ──────────────────────────────────────
-    op.drop_constraint("fk_generation_tasks_created_by_user_id", "generation_tasks", type_="foreignkey")
-    op.drop_constraint("fk_generation_tasks_edit_plan_id", "generation_tasks", type_="foreignkey")
-    op.drop_constraint("fk_generation_tasks_asset_library_id", "generation_tasks", type_="foreignkey")
-    op.drop_constraint("fk_generation_tasks_project_id", "generation_tasks", type_="foreignkey")
-    op.alter_column("generation_tasks", "created_by_user_id", existing_type=sa.String(32), nullable=False)
-    op.alter_column("generation_tasks", "edit_plan_id", existing_type=sa.String(32), nullable=False)
-
-    # ── classification_jobs ───────────────────────────────────
-    op.drop_constraint("fk_classification_jobs_asset_id", "classification_jobs", type_="foreignkey")
-    op.drop_constraint("fk_classification_jobs_project_id", "classification_jobs", type_="foreignkey")
-
-    # ── ingest_jobs ───────────────────────────────────────────
-    op.drop_constraint("fk_ingest_jobs_result_asset_id", "ingest_jobs", type_="foreignkey")
-    op.drop_constraint("fk_ingest_jobs_library_id", "ingest_jobs", type_="foreignkey")
-    op.drop_constraint("fk_ingest_jobs_project_id", "ingest_jobs", type_="foreignkey")
-    op.alter_column("ingest_jobs", "result_asset_id", existing_type=sa.String(32), nullable=False)
-
-    # ── edit_plan_clips ───────────────────────────────────────
-    op.drop_constraint("fk_edit_plan_clips_asset_id", "edit_plan_clips", type_="foreignkey")
-    op.drop_constraint("fk_edit_plan_clips_edit_plan_id", "edit_plan_clips", type_="foreignkey")
-
-    # ── edit_plans ────────────────────────────────────────────
-    op.drop_constraint("fk_edit_plans_created_by_user_id", "edit_plans", type_="foreignkey")
-    op.drop_constraint("fk_edit_plans_title_id", "edit_plans", type_="foreignkey")
-    op.drop_constraint("fk_edit_plans_asset_library_id", "edit_plans", type_="foreignkey")
-    op.drop_constraint("fk_edit_plans_template_id", "edit_plans", type_="foreignkey")
-    op.drop_constraint("fk_edit_plans_project_id", "edit_plans", type_="foreignkey")
-    op.alter_column("edit_plans", "created_by_user_id", existing_type=sa.String(32), nullable=False)
-    op.alter_column("edit_plans", "title_id", existing_type=sa.String(32), nullable=False)
-
-    # ── edit_templates ────────────────────────────────────────
-    op.drop_constraint("fk_edit_templates_created_by_user_id", "edit_templates", type_="foreignkey")
-    op.drop_constraint("fk_edit_templates_project_id", "edit_templates", type_="foreignkey")
-    op.alter_column("edit_templates", "created_by_user_id", existing_type=sa.String(32), nullable=False)
-
-    # ── project_titles ────────────────────────────────────────
-    op.drop_index("ix_project_titles_created_by_user_id", table_name="project_titles")
-    op.drop_constraint("fk_project_titles_created_by_user_id", "project_titles", type_="foreignkey")
-    op.drop_constraint("fk_project_titles_project_id", "project_titles", type_="foreignkey")
-    op.alter_column("project_titles", "created_by_user_id", existing_type=sa.String(36), nullable=False)
-
-    # ── assets ────────────────────────────────────────────────
-    op.drop_index("ix_assets_uploaded_by_user_id", table_name="assets")
-    op.drop_constraint("fk_assets_uploaded_by_user_id", "assets", type_="foreignkey")
-    op.drop_constraint("fk_assets_asset_library_id", "assets", type_="foreignkey")
-    op.drop_constraint("fk_assets_project_id", "assets", type_="foreignkey")
-    op.alter_column("assets", "uploaded_by_user_id", existing_type=sa.String(36), nullable=False)
-
-    # ── asset_libraries ───────────────────────────────────────
-    op.drop_constraint("fk_asset_libraries_project_id", "asset_libraries", type_="foreignkey")
-
-    # ── projects ──────────────────────────────────────────────
-    op.drop_constraint("fk_projects_owner_user_id", "projects", type_="foreignkey")

alembic/versions/010_phase0_extensibility.py

@@ -10,9 +10,11 @@ This migration:
 2. Creates title_libraries table (独立标题库，支持跨项目复用)
 3. Creates voice_libraries table (配音库，支持 AI 配音管理)
 """
-from alembic import op
+
 import sqlalchemy as sa
 
+from alembic import op
+
 # revision identifiers
 revision = "010"
 down_revision = "009"
@@ -25,21 +27,11 @@ def upgrade() -> None:
 
     # ── 1. Add metadata JSONB to existing tables ──
 
-    conn.execute(sa.text(
-        "ALTER TABLE projects ADD COLUMN IF NOT EXISTS metadata JSONB NOT NULL DEFAULT '{}'"
-    ))
-    conn.execute(sa.text(
-        "ALTER TABLE asset_libraries ADD COLUMN IF NOT EXISTS metadata JSONB NOT NULL DEFAULT '{}'"
-    ))
-    conn.execute(sa.text(
-        "ALTER TABLE assets ADD COLUMN IF NOT EXISTS metadata JSONB NOT NULL DEFAULT '{}'"
-    ))
-    conn.execute(sa.text(
-        "ALTER TABLE edit_templates ADD COLUMN IF NOT EXISTS metadata JSONB NOT NULL DEFAULT '{}'"
-    ))
-    conn.execute(sa.text(
-        "ALTER TABLE generation_tasks ADD COLUMN IF NOT EXISTS metadata JSONB NOT NULL DEFAULT '{}'"
-    ))
+    conn.execute(sa.text("ALTER TABLE projects ADD COLUMN IF NOT EXISTS metadata JSONB NOT NULL DEFAULT '{}'"))
+    conn.execute(sa.text("ALTER TABLE asset_libraries ADD COLUMN IF NOT EXISTS metadata JSONB NOT NULL DEFAULT '{}'"))
+    conn.execute(sa.text("ALTER TABLE assets ADD COLUMN IF NOT EXISTS metadata JSONB NOT NULL DEFAULT '{}'"))
+    conn.execute(sa.text("ALTER TABLE edit_templates ADD COLUMN IF NOT EXISTS metadata JSONB NOT NULL DEFAULT '{}'"))
+    conn.execute(sa.text("ALTER TABLE generation_tasks ADD COLUMN IF NOT EXISTS metadata JSONB NOT NULL DEFAULT '{}'"))
 
     # ── 2. Create title_libraries table ──
 
@@ -59,15 +51,9 @@ def upgrade() -> None:
             updated_at TIMESTAMP NOT NULL DEFAULT NOW()
         )
     """))
-    conn.execute(sa.text(
-        "CREATE INDEX IF NOT EXISTS ix_title_libraries_user_id ON title_libraries(user_id)"
-    ))
-    conn.execute(sa.text(
-        "CREATE INDEX IF NOT EXISTS ix_title_libraries_category ON title_libraries(category)"
-    ))
-    conn.execute(sa.text(
-        "CREATE INDEX IF NOT EXISTS ix_title_libraries_is_active ON title_libraries(is_active)"
-    ))
+    conn.execute(sa.text("CREATE INDEX IF NOT EXISTS ix_title_libraries_user_id ON title_libraries(user_id)"))
+    conn.execute(sa.text("CREATE INDEX IF NOT EXISTS ix_title_libraries_category ON title_libraries(category)"))
+    conn.execute(sa.text("CREATE INDEX IF NOT EXISTS ix_title_libraries_is_active ON title_libraries(is_active)"))
 
     # ── 3. Create voice_libraries table ──
 
@@ -91,15 +77,9 @@ def upgrade() -> None:
             updated_at TIMESTAMP NOT NULL DEFAULT NOW()
         )
     """))
-    conn.execute(sa.text(
-        "CREATE INDEX IF NOT EXISTS ix_voice_libraries_user_id ON voice_libraries(user_id)"
-    ))
-    conn.execute(sa.text(
-        "CREATE INDEX IF NOT EXISTS ix_voice_libraries_project_id ON voice_libraries(project_id)"
-    ))
-    conn.execute(sa.text(
-        "CREATE INDEX IF NOT EXISTS ix_voice_libraries_status ON voice_libraries(status)"
-    ))
+    conn.execute(sa.text("CREATE INDEX IF NOT EXISTS ix_voice_libraries_user_id ON voice_libraries(user_id)"))
+    conn.execute(sa.text("CREATE INDEX IF NOT EXISTS ix_voice_libraries_project_id ON voice_libraries(project_id)"))
+    conn.execute(sa.text("CREATE INDEX IF NOT EXISTS ix_voice_libraries_status ON voice_libraries(status)"))
 
 
 def downgrade() -> None:

alembic/versions/011_phase1_core_refactor.py

@@ -14,9 +14,11 @@ This migration:
    - edit_plan_clips (编辑计划片段)
 2. Removes edit_plan_id column from generation_tasks table
 """
-from alembic import op
+
 import sqlalchemy as sa
 
+from alembic import op
+
 # revision identifiers
 revision = "011"
 down_revision = "010"
@@ -39,9 +41,7 @@ def upgrade() -> None:
 
     # ── 2. Remove edit_plan_id from generation_tasks ──
 
-    conn.execute(sa.text(
-        "ALTER TABLE generation_tasks DROP COLUMN IF EXISTS edit_plan_id"
-    ))
+    conn.execute(sa.text("ALTER TABLE generation_tasks DROP COLUMN IF EXISTS edit_plan_id"))
 
 
 def downgrade() -> None:
@@ -49,9 +49,7 @@ def downgrade() -> None:
 
     # ── 1. Re-add edit_plan_id to generation_tasks ──
 
-    conn.execute(sa.text(
-        "ALTER TABLE generation_tasks ADD COLUMN IF NOT EXISTS edit_plan_id VARCHAR(32)"
-    ))
+    conn.execute(sa.text("ALTER TABLE generation_tasks ADD COLUMN IF NOT EXISTS edit_plan_id VARCHAR(32)"))
 
     # ── 2. Recreate deprecated tables (basic structure) ──
 

alembic/versions/012_add_duplication_tables.py

@@ -8,9 +8,11 @@ This migration creates two new tables:
 1. duplication_records — 查重记录主表
 2. duplication_segments — 重复片段详情表
 """
-from alembic import op
+
 import sqlalchemy as sa
 
+from alembic import op
+
 # revision identifiers
 revision = "012"
 down_revision = "011"
@@ -40,12 +42,8 @@ def upgrade() -> None:
             updated_at TIMESTAMP NOT NULL DEFAULT NOW()
         )
     """))
-    conn.execute(sa.text(
-        "CREATE INDEX IF NOT EXISTS ix_duplication_records_user_id ON duplication_records(user_id)"
-    ))
-    conn.execute(sa.text(
-        "CREATE INDEX IF NOT EXISTS ix_duplication_records_status ON duplication_records(status)"
-    ))
+    conn.execute(sa.text("CREATE INDEX IF NOT EXISTS ix_duplication_records_user_id ON duplication_records(user_id)"))
+    conn.execute(sa.text("CREATE INDEX IF NOT EXISTS ix_duplication_records_status ON duplication_records(status)"))
 
     # ── 2. Create duplication_segments table ──
 
@@ -62,9 +60,9 @@ def upgrade() -> None:
             similarity FLOAT NOT NULL
         )
     """))
-    conn.execute(sa.text(
-        "CREATE INDEX IF NOT EXISTS ix_duplication_segments_record_id ON duplication_segments(record_id)"
-    ))
+    conn.execute(
+        sa.text("CREATE INDEX IF NOT EXISTS ix_duplication_segments_record_id ON duplication_segments(record_id)")
+    )
 
 
 def downgrade() -> None:

alembic/versions/013_add_recipe_tables.py

@@ -0,0 +1,62 @@
+"""Phase 2 - 配方复用：recipes + recipe_items
+
+Revision ID: 013
+Revises: 012
+Create Date: 2026-06-29
+
+This migration creates two new tables:
+1. recipes — 配方主表
+2. recipe_items — 配方素材项表
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+# revision identifiers
+revision = "013"
+down_revision = "012"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    conn = op.get_bind()
+
+    # ── 1. Create recipes table ──
+
+    conn.execute(sa.text("""
+        CREATE TABLE IF NOT EXISTS recipes (
+            id VARCHAR(36) PRIMARY KEY,
+            user_id VARCHAR(36) NOT NULL,
+            name VARCHAR(200) NOT NULL,
+            description TEXT NOT NULL DEFAULT '',
+            template_id VARCHAR(36) NOT NULL DEFAULT '',
+            generation_params JSONB NOT NULL DEFAULT '{}',
+            is_active BOOLEAN NOT NULL DEFAULT TRUE,
+            metadata JSONB NOT NULL DEFAULT '{}',
+            created_at TIMESTAMP NOT NULL DEFAULT NOW(),
+            updated_at TIMESTAMP NOT NULL DEFAULT NOW()
+        )
+    """))
+    conn.execute(sa.text("CREATE INDEX IF NOT EXISTS ix_recipes_user_id ON recipes(user_id)"))
+
+    # ── 2. Create recipe_items table ──
+
+    conn.execute(sa.text("""
+        CREATE TABLE IF NOT EXISTS recipe_items (
+            id VARCHAR(36) PRIMARY KEY,
+            recipe_id VARCHAR(36) NOT NULL,
+            item_type VARCHAR(20) NOT NULL,
+            item_id VARCHAR(36) NOT NULL,
+            position INTEGER NOT NULL DEFAULT 0,
+            metadata JSONB NOT NULL DEFAULT '{}'
+        )
+    """))
+    conn.execute(sa.text("CREATE INDEX IF NOT EXISTS ix_recipe_items_recipe_id ON recipe_items(recipe_id)"))
+
+
+def downgrade() -> None:
+    conn = op.get_bind()
+    conn.execute(sa.text("DROP TABLE IF EXISTS recipe_items"))
+    conn.execute(sa.text("DROP TABLE IF EXISTS recipes"))

alembic/versions/014_add_template_tables.py

@@ -0,0 +1,83 @@
+"""Phase 3 - 剪辑计划模板：templates + template_segments + template_categories
+
+Revision ID: 014
+Revises: 013
+Create Date: 2026-06-29
+
+This migration creates three new tables:
+1. templates — 剪辑计划模板主表
+2. template_segments — 模板片段表
+3. template_categories — 模板分类表
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+# revision identifiers
+revision = "014"
+down_revision = "013"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    conn = op.get_bind()
+
+    # ── 1. Create templates table ──
+    conn.execute(sa.text("""
+        CREATE TABLE IF NOT EXISTS templates (
+            id VARCHAR(36) PRIMARY KEY,
+            user_id VARCHAR(36) NOT NULL,
+            name VARCHAR(200) NOT NULL,
+            mode VARCHAR(30) NOT NULL,
+            category VARCHAR(100) NOT NULL DEFAULT '',
+            tags JSONB NOT NULL DEFAULT '[]',
+            title_config JSONB NOT NULL DEFAULT '{}',
+            subtitle_config JSONB NOT NULL DEFAULT '{}',
+            bgm_config JSONB NOT NULL DEFAULT '{}',
+            estimated_duration FLOAT NOT NULL DEFAULT 0.0,
+            is_active BOOLEAN NOT NULL DEFAULT TRUE,
+            created_at TIMESTAMP NOT NULL DEFAULT NOW(),
+            updated_at TIMESTAMP NOT NULL DEFAULT NOW()
+        )
+    """))
+    conn.execute(sa.text("CREATE INDEX IF NOT EXISTS ix_templates_user_id ON templates(user_id)"))
+    conn.execute(sa.text("CREATE INDEX IF NOT EXISTS ix_templates_mode ON templates(mode)"))
+
+    # ── 2. Create template_segments table ──
+    conn.execute(sa.text("""
+        CREATE TABLE IF NOT EXISTS template_segments (
+            id VARCHAR(36) PRIMARY KEY,
+            template_id VARCHAR(36) NOT NULL,
+            segment_order INTEGER NOT NULL,
+            duration_min FLOAT NOT NULL,
+            duration_max FLOAT NOT NULL,
+            material_type VARCHAR(20),
+            created_at TIMESTAMP NOT NULL DEFAULT NOW(),
+            updated_at TIMESTAMP NOT NULL DEFAULT NOW()
+        )
+    """))
+    conn.execute(
+        sa.text("CREATE INDEX IF NOT EXISTS ix_template_segments_template_id " "ON template_segments(template_id)")
+    )
+
+    # ── 3. Create template_categories table ──
+    conn.execute(sa.text("""
+        CREATE TABLE IF NOT EXISTS template_categories (
+            id VARCHAR(36) PRIMARY KEY,
+            user_id VARCHAR(36) NOT NULL,
+            name VARCHAR(100) NOT NULL,
+            created_at TIMESTAMP NOT NULL DEFAULT NOW()
+        )
+    """))
+    conn.execute(
+        sa.text("CREATE INDEX IF NOT EXISTS ix_template_categories_user_id " "ON template_categories(user_id)")
+    )
+
+
+def downgrade() -> None:
+    conn = op.get_bind()
+    conn.execute(sa.text("DROP TABLE IF EXISTS template_categories"))
+    conn.execute(sa.text("DROP TABLE IF EXISTS template_segments"))
+    conn.execute(sa.text("DROP TABLE IF EXISTS templates"))

alembic/versions/015_add_generation_task_extensions.py

@@ -0,0 +1,34 @@
+"""add generation task extensions
+
+Revision ID: 015
+Revises: 014
+Create Date: 2026-06-29
+"""
+
+import sqlalchemy as sa
+from sqlalchemy.dialects import mysql
+
+from alembic import op
+
+# revision identifiers
+revision = "015"
+down_revision = "014"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column("generation_tasks", sa.Column("template_id", sa.String(36), nullable=False, server_default=""))
+    op.add_column("generation_tasks", sa.Column("asset_ids", mysql.JSON(), nullable=False, server_default="[]"))
+    op.add_column("generation_tasks", sa.Column("title_ids", mysql.JSON(), nullable=False, server_default="[]"))
+    op.add_column("generation_tasks", sa.Column("voice_ids", mysql.JSON(), nullable=False, server_default="[]"))
+
+    op.create_index(op.f("ix_generation_tasks_template_id"), "generation_tasks", ["template_id"])
+
+
+def downgrade() -> None:
+    op.drop_index(op.f("ix_generation_tasks_template_id"), table_name="generation_tasks")
+    op.drop_column("generation_tasks", "voice_ids")
+    op.drop_column("generation_tasks", "title_ids")
+    op.drop_column("generation_tasks", "asset_ids")
+    op.drop_column("generation_tasks", "template_id")

alembic/versions/016_phase8_edit_template_plan.py

@@ -0,0 +1,116 @@
+"""phase8 edit template plan
+
+Revision ID: 016
+Revises: 015
+Create Date: 2026-07-01
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "016"
+down_revision = "015"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # --- edit_templates: 替换为 Phase 8 新 schema ---
+    # 删除旧列
+    op.drop_column("edit_templates", "project_id")
+    op.drop_column("edit_templates", "target_duration")
+    op.drop_column("edit_templates", "clip_count")
+    op.drop_column("edit_templates", "is_active")
+    op.drop_column("edit_templates", "created_by_user_id")
+    op.drop_column("edit_templates", "metadata")
+
+    # 添加新列
+    op.add_column(
+        "edit_templates",
+        sa.Column("template_type", sa.String(50), nullable=False, server_default="default"),
+    )
+    op.add_column(
+        "edit_templates",
+        sa.Column("config", sa.JSON(), nullable=False, server_default="{}"),
+    )
+    op.add_column(
+        "edit_templates",
+        sa.Column("preview_url", sa.String(1000), nullable=False, server_default=""),
+    )
+    op.add_column(
+        "edit_templates",
+        sa.Column("sort_weight", sa.Integer(), nullable=False, server_default="0"),
+    )
+    op.add_column(
+        "edit_templates",
+        sa.Column("status", sa.String(20), nullable=False, server_default="active"),
+    )
+
+    # 添加索引
+    op.create_index("ix_edit_templates_template_type", "edit_templates", ["template_type"])
+    op.create_index("ix_edit_templates_sort_weight", "edit_templates", ["sort_weight"])
+    op.create_index("ix_edit_templates_status", "edit_templates", ["status"])
+
+    # --- edit_plans: 重建表（在 011 中被删除） ---
+    op.create_table(
+        "edit_plans",
+        sa.Column("id", sa.String(32), primary_key=True),
+        sa.Column("template_id", sa.String(32), nullable=False, index=True),
+        sa.Column("name", sa.String(200), nullable=False),
+        sa.Column("status", sa.String(20), nullable=False, server_default="draft", index=True),
+        sa.Column("total_duration", sa.Float(), nullable=False, server_default="0"),
+        sa.Column("config", sa.JSON(), nullable=False, server_default="{}"),
+        sa.Column(
+            "created_at",
+            sa.DateTime(),
+            nullable=False,
+            server_default=sa.func.now(),
+        ),
+        sa.Column(
+            "updated_at",
+            sa.DateTime(),
+            nullable=False,
+            server_default=sa.func.now(),
+        ),
+    )
+
+
+def downgrade() -> None:
+    op.drop_table("edit_plans")
+
+    op.drop_index("ix_edit_templates_status", "edit_templates")
+    op.drop_index("ix_edit_templates_sort_weight", "edit_templates")
+    op.drop_index("ix_edit_templates_template_type", "edit_templates")
+
+    op.drop_column("edit_templates", "status")
+    op.drop_column("edit_templates", "sort_weight")
+    op.drop_column("edit_templates", "preview_url")
+    op.drop_column("edit_templates", "config")
+    op.drop_column("edit_templates", "template_type")
+
+    # 恢复旧列
+    op.add_column(
+        "edit_templates",
+        sa.Column("project_id", sa.String(32), nullable=False, server_default=""),
+    )
+    op.add_column(
+        "edit_templates",
+        sa.Column("target_duration", sa.Float(), nullable=False, server_default="30"),
+    )
+    op.add_column(
+        "edit_templates",
+        sa.Column("clip_count", sa.Integer(), nullable=False, server_default="3"),
+    )
+    op.add_column(
+        "edit_templates",
+        sa.Column("is_active", sa.Boolean(), nullable=False, server_default=sa.true()),
+    )
+    op.add_column(
+        "edit_templates",
+        sa.Column("created_by_user_id", sa.String(32), nullable=False, server_default=""),
+    )
+    op.add_column(
+        "edit_templates",
+        sa.Column("metadata", sa.JSON(), nullable=False, server_default="{}"),
+    )

alembic/versions/017_phase8_clip_config_plan_clip.py

@@ -0,0 +1,82 @@
+"""Phase 8: Create template_clip_configs and edit_plan_clips tables
+
+Revision ID: 017
+Revises: 016
+Create Date: 2026-07-01
+
+新增两张表：
+- template_clip_configs: 模板片段配置（定义模板中每个片段的规则）
+- edit_plan_clips: 剪辑计划片段（剪辑计划中的具体片段实例）
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "017"
+down_revision = "016"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # template_clip_configs: 模板片段配置表
+    op.create_table(
+        "template_clip_configs",
+        sa.Column("id", sa.String(32), primary_key=True),
+        sa.Column("template_id", sa.String(32), nullable=False, index=True),
+        sa.Column("clip_type", sa.String(20), nullable=False, index=True),
+        sa.Column("order", sa.Integer, nullable=False),
+        sa.Column("min_duration", sa.Float, nullable=False, server_default="0.0"),
+        sa.Column("max_duration", sa.Float, nullable=False, server_default="0.0"),
+        sa.Column("text_template", sa.Text, nullable=False, server_default=""),
+        sa.Column("material_requirements", sa.JSON, nullable=False, server_default="{}"),
+        sa.Column("transition_effect", sa.String(20), nullable=False, server_default="cut"),
+        sa.Column("config", sa.JSON, nullable=False, server_default="{}"),
+        sa.Column(
+            "created_at",
+            sa.DateTime,
+            nullable=False,
+            server_default=sa.func.now(),
+        ),
+        sa.Column(
+            "updated_at",
+            sa.DateTime,
+            nullable=False,
+            server_default=sa.func.now(),
+        ),
+    )
+
+    # edit_plan_clips: 剪辑计划片段表
+    op.create_table(
+        "edit_plan_clips",
+        sa.Column("id", sa.String(32), primary_key=True),
+        sa.Column("plan_id", sa.String(32), nullable=False, index=True),
+        sa.Column("clip_type", sa.String(20), nullable=False, index=True),
+        sa.Column("order", sa.Integer, nullable=False),
+        sa.Column("template_clip_config_id", sa.String(32), nullable=False, server_default="", index=True),
+        sa.Column("asset_id", sa.String(32), nullable=False, server_default="", index=True),
+        sa.Column("text_content", sa.Text, nullable=False, server_default=""),
+        sa.Column("start_time", sa.Float, nullable=False, server_default="0.0"),
+        sa.Column("duration", sa.Float, nullable=False, server_default="0.0"),
+        sa.Column("transition_effect", sa.String(20), nullable=False, server_default="cut"),
+        sa.Column("status", sa.String(20), nullable=False, server_default="pending", index=True),
+        sa.Column("config", sa.JSON, nullable=False, server_default="{}"),
+        sa.Column(
+            "created_at",
+            sa.DateTime,
+            nullable=False,
+            server_default=sa.func.now(),
+        ),
+        sa.Column(
+            "updated_at",
+            sa.DateTime,
+            nullable=False,
+            server_default=sa.func.now(),
+        ),
+    )
+
+
+def downgrade() -> None:
+    op.drop_table("edit_plan_clips")
+    op.drop_table("template_clip_configs")

alembic/versions/018_add_jobs_table.py

@@ -0,0 +1,55 @@
+"""Phase 8 任务 2.10: Create jobs table for unified async task management
+
+Revision ID: 018
+Revises: 017
+Create Date: 2026-07-01
+
+新增 jobs 表，用于统一管理异步任务（视频合成、渲染等）的生命周期。
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "018"
+down_revision = "017"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.create_table(
+        "jobs",
+        sa.Column("id", sa.String(32), primary_key=True),
+        sa.Column("project_id", sa.String(32), nullable=False, index=True),
+        sa.Column("job_type", sa.String(30), nullable=False, index=True),
+        sa.Column("status", sa.String(20), nullable=False, server_default="pending", index=True),
+        sa.Column("progress", sa.Float, nullable=False, server_default="0.0"),
+        sa.Column("current_stage", sa.String(200), nullable=False, server_default=""),
+        sa.Column("payload", sa.JSON, nullable=False, server_default="{}"),
+        sa.Column("result", sa.JSON, nullable=False, server_default="{}"),
+        sa.Column("error_message", sa.Text, nullable=False, server_default=""),
+        sa.Column("retry_count", sa.Integer, nullable=False, server_default="0"),
+        sa.Column("max_retries", sa.Integer, nullable=False, server_default="3"),
+        sa.Column("celery_task_id", sa.String(100), nullable=False, server_default=""),
+        sa.Column("source_id", sa.String(32), nullable=False, server_default="", index=True),
+        sa.Column("created_by_user_id", sa.String(32), nullable=False, server_default="", index=True),
+        sa.Column("started_at", sa.DateTime, nullable=True),
+        sa.Column("completed_at", sa.DateTime, nullable=True),
+        sa.Column(
+            "created_at",
+            sa.DateTime,
+            nullable=False,
+            server_default=sa.func.now(),
+        ),
+        sa.Column(
+            "updated_at",
+            sa.DateTime,
+            nullable=False,
+            server_default=sa.func.now(),
+        ),
+    )
+
+
+def downgrade() -> None:
+    op.drop_table("jobs")

alembic/versions/019_add_voice_clone_profiles_table.py

@@ -0,0 +1,53 @@
+"""Task 3.05: Create voice_clone_profiles table
+
+Revision ID: 019
+Revises: 018
+Create Date: 2026-07-02
+
+新增 voice_clone_profiles 表，用于存储音色克隆档案。
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "019"
+down_revision = "018"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.create_table(
+        "voice_clone_profiles",
+        sa.Column("id", sa.String(36), primary_key=True),
+        sa.Column("user_id", sa.String(36), nullable=False, index=True),
+        sa.Column("name", sa.String(100), nullable=False),
+        sa.Column("description", sa.Text(), nullable=False, server_default=""),
+        sa.Column("source_audio_url", sa.Text(), nullable=False, server_default=""),
+        sa.Column("voice_id", sa.String(100), nullable=False, server_default=""),
+        sa.Column("voice_model", sa.String(100), nullable=False, server_default=""),
+        sa.Column("language", sa.String(20), nullable=False, server_default="zh-CN"),
+        sa.Column("gender", sa.String(20), nullable=False, server_default="unknown"),
+        sa.Column("status", sa.String(20), nullable=False, server_default="pending", index=True),
+        sa.Column("error_message", sa.Text(), nullable=False, server_default=""),
+        sa.Column("retry_count", sa.Integer(), nullable=False, server_default="0"),
+        sa.Column("max_retries", sa.Integer(), nullable=False, server_default="3"),
+        sa.Column("metadata", sa.JSON(), nullable=False, server_default="{}"),
+        sa.Column(
+            "created_at",
+            sa.DateTime(),
+            nullable=False,
+            server_default=sa.func.now(),
+        ),
+        sa.Column(
+            "updated_at",
+            sa.DateTime(),
+            nullable=False,
+            server_default=sa.func.now(),
+        ),
+    )
+
+
+def downgrade() -> None:
+    op.drop_table("voice_clone_profiles")

alembic/versions/020_add_tts_jobs_table.py

@@ -0,0 +1,59 @@
+"""Task 3.06: Create tts_jobs table
+
+Revision ID: 020
+Revises: 019
+Create Date: 2026-07-02
+
+新增 tts_jobs 表，用于存储 TTS 合成任务。
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "020"
+down_revision = "019"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.create_table(
+        "tts_jobs",
+        sa.Column("id", sa.String(36), primary_key=True),
+        sa.Column("user_id", sa.String(36), nullable=False, index=True),
+        sa.Column("input_text", sa.Text(), nullable=False),
+        sa.Column("voice_id", sa.String(100), nullable=False, server_default=""),
+        sa.Column("voice_model", sa.String(100), nullable=False, server_default=""),
+        sa.Column("project_id", sa.String(36), nullable=False, server_default=""),
+        sa.Column("voice_clone_profile_id", sa.String(36), nullable=False, server_default=""),
+        sa.Column("status", sa.String(20), nullable=False, server_default="pending", index=True),
+        sa.Column("output_audio_url", sa.Text(), nullable=False, server_default=""),
+        sa.Column("output_audio_key", sa.String(500), nullable=False, server_default=""),
+        sa.Column("duration", sa.Float(), nullable=False, server_default="0"),
+        sa.Column("file_size", sa.Integer(), nullable=False, server_default="0"),
+        sa.Column("sample_rate", sa.Integer(), nullable=False, server_default="22050"),
+        sa.Column("format", sa.String(20), nullable=False, server_default="mp3"),
+        sa.Column("error_message", sa.Text(), nullable=False, server_default=""),
+        sa.Column("retry_count", sa.Integer(), nullable=False, server_default="0"),
+        sa.Column("max_retries", sa.Integer(), nullable=False, server_default="3"),
+        sa.Column("metadata", sa.JSON(), nullable=False, server_default="{}"),
+        sa.Column("started_at", sa.DateTime(), nullable=True),
+        sa.Column("completed_at", sa.DateTime(), nullable=True),
+        sa.Column(
+            "created_at",
+            sa.DateTime(),
+            nullable=False,
+            server_default=sa.func.now(),
+        ),
+        sa.Column(
+            "updated_at",
+            sa.DateTime(),
+            nullable=False,
+            server_default=sa.func.now(),
+        ),
+    )
+
+
+def downgrade() -> None:
+    op.drop_table("tts_jobs")

alembic/versions/021_add_billing_records_table.py

@@ -0,0 +1,43 @@
+"""Task 3.09: Create billing_records table
+
+Revision ID: 021
+Revises: 020
+Create Date: 2026-07-03
+
+新增 billing_records 表，用于存储账单记录。
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "021"
+down_revision = "020"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.create_table(
+        "billing_records",
+        sa.Column("id", sa.String(36), primary_key=True),
+        sa.Column("user_id", sa.String(36), nullable=False, index=True),
+        sa.Column("plan_name", sa.String(50), nullable=False),
+        sa.Column("amount", sa.Float, nullable=False),
+        sa.Column("billing_cycle", sa.String(20), nullable=False),
+        sa.Column("status", sa.String(20), nullable=False, server_default="pending"),
+        sa.Column("payment_method", sa.String(50), nullable=True),
+        sa.Column("payment_id", sa.String(100), nullable=True),
+        sa.Column("invoice_url", sa.String(500), nullable=True),
+        sa.Column(
+            "created_at",
+            sa.DateTime(),
+            nullable=False,
+            server_default=sa.func.now(),
+        ),
+        sa.Column("paid_at", sa.DateTime(), nullable=True),
+    )
+
+
+def downgrade() -> None:
+    op.drop_table("billing_records")

alembic/versions/022_add_source_edit_plan_id.py

@@ -0,0 +1,56 @@
+"""Task: Add source_edit_plan_id to edit_plans and generation_tasks
+
+Revision ID: 022
+Revises: 021
+Create Date: 2026-07-04
+
+新增 source_edit_plan_id 字段到 edit_plans 和 generation_tasks 表，
+用于关联生成记录到其来源的剪辑计划。
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "022"
+down_revision = "021"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "edit_plans",
+        sa.Column("source_edit_plan_id", sa.String(32), nullable=True),
+    )
+    op.create_index(
+        op.f("ix_edit_plans_source_edit_plan_id"),
+        "edit_plans",
+        ["source_edit_plan_id"],
+        unique=False,
+    )
+
+    op.add_column(
+        "generation_tasks",
+        sa.Column("source_edit_plan_id", sa.String(32), nullable=True),
+    )
+    op.create_index(
+        op.f("ix_generation_tasks_source_edit_plan_id"),
+        "generation_tasks",
+        ["source_edit_plan_id"],
+        unique=False,
+    )
+
+
+def downgrade() -> None:
+    op.drop_index(
+        op.f("ix_generation_tasks_source_edit_plan_id"),
+        table_name="generation_tasks",
+    )
+    op.drop_column("generation_tasks", "source_edit_plan_id")
+
+    op.drop_index(
+        op.f("ix_edit_plans_source_edit_plan_id"),
+        table_name="edit_plans",
+    )
+    op.drop_column("edit_plans", "source_edit_plan_id")

alembic/versions/023_add_edit_plan_project_and_user.py

@@ -0,0 +1,56 @@
+"""Task: Add project_id and created_by_user_id to edit_plans
+
+Revision ID: 023
+Revises: 022
+Create Date: 2026-07-05
+
+新增 project_id 和 created_by_user_id 字段到 edit_plans 表，
+用于项目归属鉴权和用户归属追踪，修复审计发现的 P1 越权漏洞。
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "023"
+down_revision = "022"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "edit_plans",
+        sa.Column("project_id", sa.String(32), nullable=False, server_default=""),
+    )
+    op.create_index(
+        op.f("ix_edit_plans_project_id"),
+        "edit_plans",
+        ["project_id"],
+        unique=False,
+    )
+
+    op.add_column(
+        "edit_plans",
+        sa.Column("created_by_user_id", sa.String(32), nullable=False, server_default=""),
+    )
+    op.create_index(
+        op.f("ix_edit_plans_created_by_user_id"),
+        "edit_plans",
+        ["created_by_user_id"],
+        unique=False,
+    )
+
+
+def downgrade() -> None:
+    op.drop_index(
+        op.f("ix_edit_plans_created_by_user_id"),
+        table_name="edit_plans",
+    )
+    op.drop_column("edit_plans", "created_by_user_id")
+
+    op.drop_index(
+        op.f("ix_edit_plans_project_id"),
+        table_name="edit_plans",
+    )
+    op.drop_column("edit_plans", "project_id")

alembic/versions/024_add_user_is_admin.py

@@ -0,0 +1,28 @@
+"""Task: Add is_admin to users
+
+Revision ID: 024
+Revises: 023
+Create Date: 2026-07-05
+
+新增 is_admin 字段到 users 表，用于模板管理等管理员权限校验。
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "024"
+down_revision = "023"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "users",
+        sa.Column("is_admin", sa.Boolean(), nullable=False, server_default=sa.text("false")),
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("users", "is_admin")

alembic/versions/025_add_user_wechat_fields.py

@@ -0,0 +1,79 @@
+"""Task: Add wechat_openid / wechat_unionid to users
+
+Revision ID: 025
+Revises: 024
+Create Date: 2026-07-05
+
+补录微信小程序登录所需的 wechat 字段。
+生产数据库已手动添加过这些字段和索引，因此 upgrade 做幂等检查，
+避免在已有字段的库上执行报错。
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "025"
+down_revision = "024"
+branch_labels = None
+depends_on = None
+
+
+def _column_exists(table: str, column: str) -> bool:
+    """检查列是否已存在。离线模式下返回 False。"""
+    conn = op.get_bind()
+    try:
+        result = conn.execute(
+            sa.text("SELECT 1 FROM information_schema.columns " "WHERE table_name = :table AND column_name = :column"),
+            {"table": table, "column": column},
+        )
+        if result is None:
+            return False
+        return result.scalar() is not None
+    except Exception:
+        return False
+
+
+def _index_exists(index: str) -> bool:
+    """检查索引是否已存在。离线模式下返回 False。"""
+    conn = op.get_bind()
+    try:
+        result = conn.execute(
+            sa.text("SELECT 1 FROM pg_indexes WHERE indexname = :index"),
+            {"index": index},
+        )
+        if result is None:
+            return False
+        return result.scalar() is not None
+    except Exception:
+        return False
+
+
+def upgrade() -> None:
+    # wechat_openid
+    if not _column_exists("users", "wechat_openid"):
+        op.add_column(
+            "users",
+            sa.Column("wechat_openid", sa.String(length=128), nullable=True),
+        )
+
+    # wechat_unionid
+    if not _column_exists("users", "wechat_unionid"):
+        op.add_column(
+            "users",
+            sa.Column("wechat_unionid", sa.String(length=128), nullable=True),
+        )
+
+    # 唯一索引
+    if not _index_exists("ix_users_wechat_openid"):
+        op.create_index("ix_users_wechat_openid", "users", ["wechat_openid"], unique=True)
+
+    if not _index_exists("ix_users_wechat_unionid"):
+        op.create_index("ix_users_wechat_unionid", "users", ["wechat_unionid"], unique=True)
+
+
+def downgrade() -> None:
+    op.drop_index("ix_users_wechat_unionid", table_name="users")
+    op.drop_index("ix_users_wechat_openid", table_name="users")
+    op.drop_column("users", "wechat_unionid")
+    op.drop_column("users", "wechat_openid")

alembic/versions/026_add_user_profile_fields.py

@@ -0,0 +1,56 @@
+"""Add user profile fields (name, avatar, updated_at)
+
+Revision ID: 026
+Revises: 025
+Create Date: 2026-07-05
+
+补录用户资料字段。生产数据库已手动添加过这些字段，
+因此 upgrade 做幂等检查，避免在已有字段的库上执行报错。
+"""
+
+import sqlalchemy as sa
+
+from alembic import context, op
+
+revision = "026"
+down_revision = "025"
+branch_labels = None
+depends_on = None
+
+
+def _column_exists(table: str, column: str) -> bool:
+    if context.is_offline_mode():
+        return False
+    conn = op.get_bind()
+    result = conn.execute(
+        sa.text(
+            "SELECT COUNT(*) FROM information_schema.columns " "WHERE table_name = :table AND column_name = :column"
+        ),
+        {"table": table, "column": column},
+    )
+    return result.scalar() > 0
+
+
+def upgrade() -> None:
+    if not _column_exists("users", "name"):
+        op.add_column("users", sa.Column("name", sa.String(100), nullable=True))
+
+    if not _column_exists("users", "avatar"):
+        op.add_column("users", sa.Column("avatar", sa.String(500), nullable=True))
+
+    if not _column_exists("users", "updated_at"):
+        op.add_column(
+            "users",
+            sa.Column(
+                "updated_at",
+                sa.DateTime(),
+                nullable=True,
+                server_default=sa.func.now(),
+            ),
+        )
+
+
+def downgrade() -> None:
+    op.drop_column("users", "updated_at")
+    op.drop_column("users", "avatar")
+    op.drop_column("users", "name")

alembic/versions/027_add_user_ban_fields.py

@@ -0,0 +1,44 @@
+"""Add user ban fields (ban_reason, ban_at)
+
+Revision ID: 027
+Revises: 026
+Create Date: 2026-07-05
+
+补录用户封禁字段。生产数据库已手动添加过这些字段，
+因此 upgrade 做幂等检查。
+"""
+
+import sqlalchemy as sa
+
+from alembic import context, op
+
+revision = "027"
+down_revision = "026"
+branch_labels = None
+depends_on = None
+
+
+def _column_exists(table: str, column: str) -> bool:
+    if context.is_offline_mode():
+        return False
+    conn = op.get_bind()
+    result = conn.execute(
+        sa.text(
+            "SELECT COUNT(*) FROM information_schema.columns " "WHERE table_name = :table AND column_name = :column"
+        ),
+        {"table": table, "column": column},
+    )
+    return result.scalar() > 0
+
+
+def upgrade() -> None:
+    if not _column_exists("users", "ban_reason"):
+        op.add_column("users", sa.Column("ban_reason", sa.Text(), nullable=True))
+
+    if not _column_exists("users", "ban_at"):
+        op.add_column("users", sa.Column("ban_at", sa.DateTime(), nullable=True))
+
+
+def downgrade() -> None:
+    op.drop_column("users", "ban_at")
+    op.drop_column("users", "ban_reason")

alembic/versions/028_add_user_admin_fields.py

@@ -0,0 +1,44 @@
+"""Add user admin fields (admin_status, admin_remarks)
+
+Revision ID: 028
+Revises: 027
+Create Date: 2026-07-05
+
+补录管理员备注字段。生产数据库已手动添加过这些字段，
+因此 upgrade 做幂等检查。
+"""
+
+import sqlalchemy as sa
+
+from alembic import context, op
+
+revision = "028"
+down_revision = "027"
+branch_labels = None
+depends_on = None
+
+
+def _column_exists(table: str, column: str) -> bool:
+    if context.is_offline_mode():
+        return False
+    conn = op.get_bind()
+    result = conn.execute(
+        sa.text(
+            "SELECT COUNT(*) FROM information_schema.columns " "WHERE table_name = :table AND column_name = :column"
+        ),
+        {"table": table, "column": column},
+    )
+    return result.scalar() > 0
+
+
+def upgrade() -> None:
+    if not _column_exists("users", "admin_status"):
+        op.add_column("users", sa.Column("admin_status", sa.String(50), nullable=True))
+
+    if not _column_exists("users", "admin_remarks"):
+        op.add_column("users", sa.Column("admin_remarks", sa.Text(), nullable=True))
+
+
+def downgrade() -> None:
+    op.drop_column("users", "admin_remarks")
+    op.drop_column("users", "admin_status")

alembic/versions/029_add_user_phone.py

@@ -0,0 +1,40 @@
+"""Add user phone field
+
+Revision ID: 029
+Revises: 028
+Create Date: 2026-07-05
+
+补录用户手机号字段。生产数据库已手动添加过该字段，
+因此 upgrade 做幂等检查。
+"""
+
+import sqlalchemy as sa
+
+from alembic import context, op
+
+revision = "029"
+down_revision = "028"
+branch_labels = None
+depends_on = None
+
+
+def _column_exists(table: str, column: str) -> bool:
+    if context.is_offline_mode():
+        return False
+    conn = op.get_bind()
+    result = conn.execute(
+        sa.text(
+            "SELECT COUNT(*) FROM information_schema.columns " "WHERE table_name = :table AND column_name = :column"
+        ),
+        {"table": table, "column": column},
+    )
+    return result.scalar() > 0
+
+
+def upgrade() -> None:
+    if not _column_exists("users", "phone"):
+        op.add_column("users", sa.Column("phone", sa.String(20), nullable=True))
+
+
+def downgrade() -> None:
+    op.drop_column("users", "phone")

alembic/versions/030_add_tags_and_asset_tags.py

@@ -0,0 +1,68 @@
+"""Add tags and asset_tags tables
+
+Revision ID: 030
+Revises: 029
+Create Date: 2026-07-07
+
+新增标签表和素材-标签关联表，支持规范化多对多标签管理。
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "030"
+down_revision = "029"
+branch_labels = None
+depends_on = None
+
+
+def _table_exists(table: str) -> bool:
+    ctx = op.get_context()
+    if ctx.as_sql:
+        return False
+    conn = op.get_bind()
+    result = conn.execute(
+        sa.text("SELECT COUNT(*) FROM information_schema.tables WHERE table_name = :table"),
+        {"table": table},
+    )
+    return (result.scalar() or 0) > 0
+
+
+def upgrade() -> None:
+    if not _table_exists("tags"):
+        op.create_table(
+            "tags",
+            sa.Column("id", sa.String(36), primary_key=True),
+            sa.Column("user_id", sa.String(36), nullable=False),
+            sa.Column("name", sa.String(100), nullable=False),
+            sa.Column(
+                "created_at",
+                sa.DateTime(),
+                nullable=False,
+                server_default=sa.func.now(),
+            ),
+            sa.UniqueConstraint("user_id", "name", name="uq_tags_user_name"),
+        )
+        op.create_index("ix_tags_user_id", "tags", ["user_id"])
+
+    if not _table_exists("asset_tags"):
+        op.create_table(
+            "asset_tags",
+            sa.Column("asset_id", sa.String(36), primary_key=True),
+            sa.Column("tag_id", sa.String(36), primary_key=True),
+            sa.Column(
+                "created_at",
+                sa.DateTime(),
+                nullable=False,
+                server_default=sa.func.now(),
+            ),
+        )
+        op.create_index("ix_asset_tags_tag_id", "asset_tags", ["tag_id"])
+
+
+def downgrade() -> None:
+    op.drop_index("ix_asset_tags_tag_id", table_name="asset_tags")
+    op.drop_table("asset_tags")
+    op.drop_index("ix_tags_user_id", table_name="tags")
+    op.drop_table("tags")

alembic/versions/031_add_file_hash_to_assets_and_ing_jobs.py

@@ -0,0 +1,33 @@
+"""Add file_hash to assets and ingest_jobs
+
+Revision ID: 031
+Revises: 030
+Create Date: 2026-07-07
+
+为素材去重检测功能添加 file_hash 字段。
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "031"
+down_revision = "030"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column("assets", sa.Column("file_hash", sa.String(64), nullable=True))
+    op.create_index(op.f("ix_assets_file_hash"), "assets", ["file_hash"])
+
+    op.add_column("ingest_jobs", sa.Column("file_hash", sa.String(64), nullable=True))
+    op.create_index(op.f("ix_ingest_jobs_file_hash"), "ingest_jobs", ["file_hash"])
+
+
+def downgrade() -> None:
+    op.drop_index(op.f("ix_ingest_jobs_file_hash"), table_name="ingest_jobs")
+    op.drop_column("ingest_jobs", "file_hash")
+
+    op.drop_index(op.f("ix_assets_file_hash"), table_name="assets")
+    op.drop_column("assets", "file_hash")

alembic/versions/032_add_asset_select_mode_to_generation_tasks.py

@@ -0,0 +1,28 @@
+"""Add asset_select_mode to generation_tasks
+
+Revision ID: 032
+Revises: 031
+Create Date: 2026-07-07
+
+素材库自动匹配功能：为 generation_tasks 表添加 asset_select_mode 字段。
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "032"
+down_revision = "031"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "generation_tasks",
+        sa.Column("asset_select_mode", sa.String(20), nullable=False, server_default=""),
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("generation_tasks", "asset_select_mode")

alembic/versions/033_add_batch_id_to_generation_tasks.py

@@ -0,0 +1,31 @@
+"""Add batch_id to generation_tasks
+
+Revision ID: 033
+Revises: 032
+Create Date: 2026-07-07
+
+视频查重功能：为 generation_tasks 表添加 batch_id 字段，
+用于关联同一次批量生成请求中的多个任务。
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "033"
+down_revision = "032"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "generation_tasks",
+        sa.Column("batch_id", sa.String(32), nullable=False, server_default=""),
+    )
+    op.create_index(op.f("ix_generation_tasks_batch_id"), "generation_tasks", ["batch_id"])
+
+
+def downgrade() -> None:
+    op.drop_index(op.f("ix_generation_tasks_batch_id"), table_name="generation_tasks")
+    op.drop_column("generation_tasks", "batch_id")

alembic/versions/034_add_confirm_generation_fields.py

@@ -0,0 +1,60 @@
+"""Add confirm generation fields
+
+Revision ID: 034
+Revises: 033
+Create Date: 2026-07-08
+
+确认生成 API 改造：为 generation_tasks 表添加 is_preview、source_task_id、
+output_width、output_height、cover_url、custom_title 字段。
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "034"
+down_revision = "033"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "generation_tasks",
+        sa.Column("is_preview", sa.Boolean(), nullable=False, server_default=sa.true()),
+    )
+    op.add_column(
+        "generation_tasks",
+        sa.Column("source_task_id", sa.String(32), nullable=False, server_default=""),
+    )
+    op.add_column(
+        "generation_tasks",
+        sa.Column("output_width", sa.Integer(), nullable=False, server_default=sa.text("1280")),
+    )
+    op.add_column(
+        "generation_tasks",
+        sa.Column("output_height", sa.Integer(), nullable=False, server_default=sa.text("720")),
+    )
+    op.add_column(
+        "generation_tasks",
+        sa.Column("cover_url", sa.String(1000), nullable=False, server_default=""),
+    )
+    op.add_column(
+        "generation_tasks",
+        sa.Column("custom_title", sa.String(500), nullable=False, server_default=""),
+    )
+    op.create_index(
+        op.f("ix_generation_tasks_source_task_id"),
+        "generation_tasks",
+        ["source_task_id"],
+    )
+
+
+def downgrade() -> None:
+    op.drop_index(op.f("ix_generation_tasks_source_task_id"), table_name="generation_tasks")
+    op.drop_column("generation_tasks", "custom_title")
+    op.drop_column("generation_tasks", "cover_url")
+    op.drop_column("generation_tasks", "output_height")
+    op.drop_column("generation_tasks", "output_width")
+    op.drop_column("generation_tasks", "source_task_id")
+    op.drop_column("generation_tasks", "is_preview")

apps/api/app/api/router.py

@@ -4,16 +4,26 @@ from app.api.routes.assets import router as assets_router
 from app.api.routes.auth import router as auth_router
 from app.api.routes.chunked_upload import router as chunked_upload_router
 from app.api.routes.classification_jobs import router as classification_jobs_router
+from app.api.routes.dashboard import router as dashboard_router
 from app.api.routes.duplication import router as duplication_router
+from app.api.routes.edit_plans import router as edit_plans_router
+from app.api.routes.edit_templates import router as edit_templates_router
 from app.api.routes.generated_videos import router as generated_videos_router
-from app.api.routes.titles import router as titles_router
-from app.api.routes.voices import router as voices_router
 from app.api.routes.generation_tasks import router as generation_tasks_router
 from app.api.routes.health import router as health_check_router
 from app.api.routes.ingest_jobs import router as ingest_jobs_router
+from app.api.routes.jobs import router as jobs_router
 from app.api.routes.projects import router as projects_router
+from app.api.routes.recipes import router as recipes_router
+from app.api.routes.subscription import router as subscription_router
+from app.api.routes.tags import router as tags_router
 from app.api.routes.task_center import router as task_center_router
+from app.api.routes.templates import router as templates_router
+from app.api.routes.titles import router as titles_router
+from app.api.routes.tts import router as tts_router
 from app.api.routes.upload import router as upload_router
+from app.api.routes.voice_clones import router as voice_clones_router
+from app.api.routes.voices import router as voices_router
 from fastapi import APIRouter
 
 api_router = APIRouter(prefix="/api/v1")
@@ -29,6 +39,11 @@ api_router.include_router(
     prefix="/projects",
     tags=["Project"],
 )
+api_router.include_router(
+    tags_router,
+    prefix="/tags",
+    tags=["Tag"],
+)
 api_router.include_router(
     task_center_router,
     tags=["TaskCenter"],
@@ -72,6 +87,10 @@ api_router.include_router(
     prefix="/generation",
     tags=["Generation"],
 )
+api_router.include_router(
+    jobs_router,
+    tags=["Job"],
+)
 api_router.include_router(
     generated_videos_router,
     prefix="/generated-videos",
@@ -87,8 +106,48 @@ api_router.include_router(
     prefix="/voices",
     tags=["VoiceLibrary"],
 )
+api_router.include_router(
+    voice_clones_router,
+    prefix="/voice-clones",
+    tags=["VoiceClone"],
+)
 api_router.include_router(
     duplication_router,
     prefix="/duplication",
     tags=["Duplication"],
 )
+api_router.include_router(
+    subscription_router,
+    prefix="/subscription",
+    tags=["Subscription"],
+)
+api_router.include_router(
+    recipes_router,
+    prefix="/recipes",
+    tags=["Recipe"],
+)
+api_router.include_router(
+    templates_router,
+    prefix="/templates",
+    tags=["Template"],
+)
+api_router.include_router(
+    dashboard_router,
+    prefix="/dashboard",
+    tags=["Dashboard"],
+)
+api_router.include_router(
+    edit_templates_router,
+    prefix="/edit-templates",
+    tags=["EditTemplate"],
+)
+api_router.include_router(
+    edit_plans_router,
+    prefix="/edit-plans",
+    tags=["EditPlan"],
+)
+api_router.include_router(
+    tts_router,
+    prefix="/tts",
+    tags=["TTS"],
+)

apps/api/app/api/routes/asset_diagnosis.py

@@ -1,4 +1,5 @@
-from typing import Any
+import logging
+from typing import Any, Optional
 
 from app.auth import AuthenticatedUser, get_current_user
 from app.dependencies import (
@@ -7,10 +8,12 @@ from app.dependencies import (
     get_project_repository,
 )
 from app.schemas.asset_diagnosis import AssetGapItem, AssetSmartViewItem, ProjectAssetDiagnosisResponse
-from fastapi import APIRouter, Depends, HTTPException
+from fastapi import APIRouter, Depends, HTTPException, Query
 
 from packages.domain import Asset, AssetLibraryKind, AssetStatus
 
+logger = logging.getLogger(__name__)
+
 router = APIRouter()
 
 
@@ -190,21 +193,152 @@ def _build_diagnosis(project_id: str, assets: list[Asset]) -> ProjectAssetDiagno
     )
 
 
+def _build_single_asset_diagnosis(project_id: str, asset: Asset) -> ProjectAssetDiagnosisResponse:
+    """为单个素材构建诊断结果"""
+    kind = _asset_kind(asset)
+    is_ready = asset.status == AssetStatus.READY
+    is_problem = asset.status in {AssetStatus.ERROR, AssetStatus.UPLOADING, AssetStatus.PROCESSING}
+    is_risky = is_ready and (
+        (asset.quality_score is not None and asset.quality_score < 60)
+        or asset.metadata.get("review_status") == "rejected"
+        or asset.status == AssetStatus.ERROR
+    )
+    is_unclassified = is_ready and asset.classification_status.value in {"pending", "failed"}
+
+    # 单素材评分
+    score = 0
+    if is_ready:
+        score = 60
+        if kind == "video":
+            score += 20
+        if asset.duration and asset.duration >= 5:
+            score += 10
+        if asset.quality_score and asset.quality_score >= 60:
+            score += 10
+    if is_problem:
+        score = max(score - 30, 0)
+    if is_risky:
+        score = max(score - 20, 0)
+    score = max(0, min(100, score))
+
+    gaps: list[AssetGapItem] = []
+    if not is_ready:
+        gaps.append(
+            AssetGapItem(
+                key="asset_not_ready",
+                severity="critical",
+                message=f"素材状态为 {asset.status.value}，尚未就绪",
+                recommendation="等待素材导入完成后再使用。",
+            )
+        )
+    if is_risky:
+        gaps.append(
+            AssetGapItem(
+                key="asset_low_quality",
+                severity="warning",
+                message="素材质量分偏低或已被拒绝",
+                recommendation="建议使用更清晰、稳定的素材替代。",
+            )
+        )
+    if is_unclassified:
+        gaps.append(
+            AssetGapItem(
+                key="asset_unclassified",
+                severity="info",
+                message="素材尚未完成分类",
+                recommendation="等待分类完成或手动检查素材类型。",
+            )
+        )
+    if kind == "video" and (asset.duration is None or asset.duration < 5):
+        gaps.append(
+            AssetGapItem(
+                key="short_video",
+                severity="warning",
+                message="视频时长偏短",
+                recommendation="建议使用时长 5 秒以上的视频素材。",
+            )
+        )
+
+    used_count = int(asset.metadata.get("generation_use_count") or 0)
+    smart_views = [
+        AssetSmartViewItem(
+            key="asset_info",
+            label="素材信息",
+            count=1,
+            description=f"类型: {kind}，状态: {asset.status.value}",
+        ),
+        AssetSmartViewItem(
+            key="asset_quality",
+            label="质量评分",
+            count=int(asset.quality_score or 0),
+            description=f"质量分: {asset.quality_score or '未评分'}",
+        ),
+        AssetSmartViewItem(
+            key="asset_usage",
+            label="使用次数",
+            count=used_count,
+            description=f"参与生成 {used_count} 次",
+        ),
+    ]
+
+    video_count = 1 if kind == "video" and is_ready else 0
+    image_count = 1 if kind == "image" and is_ready else 0
+    voice_count = 1 if kind == "voice" and is_ready else 0
+    total_duration = round(float(asset.duration or 0), 2) if kind == "video" else 0.0
+
+    return ProjectAssetDiagnosisResponse(
+        project_id=project_id,
+        readiness_score=score,
+        readiness_label=_readiness_label(score),
+        total_assets=1,
+        ready_assets=1 if is_ready else 0,
+        video_assets=video_count,
+        image_assets=image_count,
+        voice_assets=voice_count,
+        total_duration_seconds=total_duration,
+        estimated_video_count=1 if video_count and total_duration >= 5 else 0,
+        used_assets=1 if used_count > 0 else 0,
+        unused_assets=1 if used_count == 0 and is_ready else 0,
+        pending_review_assets=1 if asset.metadata.get("review_status") == "pending_review" else 0,
+        smart_views=smart_views,
+        gaps=gaps,
+    )
+
+
 @router.get("/projects/{project_id}/asset-diagnosis", response_model=ProjectAssetDiagnosisResponse)
 def get_project_asset_diagnosis(
     project_id: str,
+    asset_id: Optional[str] = Query(None),
     authenticated_user: AuthenticatedUser = Depends(get_current_user),
     project_repository: Any = Depends(get_project_repository),
     asset_library_repository: Any = Depends(get_asset_library_repository),
     asset_repository: Any = Depends(get_asset_repository),
 ) -> ProjectAssetDiagnosisResponse:
-    project = project_repository.find_by_id(project_id)
-    if project is None:
-        raise HTTPException(status_code=404, detail=f"Project {project_id} not found")
+    try:
+        project = project_repository.find_by_id(project_id)
+        if project is None:
+            raise HTTPException(status_code=404, detail=f"Project {project_id} not found")
+        if not project.can_access(authenticated_user.user.id):
+            raise HTTPException(status_code=403, detail="Access denied to project")
 
-    libraries = asset_library_repository.list_by_project(project_id)
-    assets: list[Asset] = []
-    for library in libraries:
-        assets.extend(asset_repository.list_by_library(library.id))
+        # 单素材诊断模式
+        if asset_id:
+            asset = asset_repository.get(asset_id)
+            if asset is None:
+                raise HTTPException(status_code=404, detail=f"Asset {asset_id} not found")
+            if asset.project_id != project_id:
+                raise HTTPException(status_code=403, detail="Asset does not belong to this project")
+            return _build_single_asset_diagnosis(project_id, asset)
+
+        libraries = asset_library_repository.find_by_project(project_id)
+        assets: list[Asset] = []
+        for library in libraries:
+            assets.extend(asset_repository.list_by_library(library.id))
+    except HTTPException:
+        raise
+    except Exception:
+        logger.exception("素材诊断查询失败: project_id=%s", project_id)
+        # 返回空诊断结果，避免 500
+        return _build_diagnosis(project_id, [])
 
     return _build_diagnosis(project_id, assets)

apps/api/app/api/routes/asset_libraries.py

@@ -5,6 +5,7 @@ from app.dependencies import get_asset_library_repository, get_project_repositor
 from app.schemas.asset_library import (
     AssetLibraryResponse,
     CreateAssetLibraryRequest,
+    EnsureDefaultLibraryRequest,
     ListAssetLibrariesResponse,
 )
 from fastapi import APIRouter, Depends, HTTPException, Query, status
@@ -15,7 +16,7 @@ from packages.application import (
     GetProjectUseCase,
     ListAssetLibrariesUseCase,
 )
-from packages.domain import AssetLibraryKind
+from packages.domain import AssetLibrary, AssetLibraryKind
 
 router = APIRouter()
 
@@ -43,13 +44,14 @@ def _to_asset_library_response(item) -> AssetLibraryResponse:
 @router.get("", response_model=ListAssetLibrariesResponse)
 def list_asset_libraries(
     project_id: str | None = Query(None),
+    kind: str | None = Query(None, pattern="^(video|voice|image)$"),
     authenticated_user: AuthenticatedUser = Depends(get_current_user),
     asset_library_repository: Any = Depends(get_asset_library_repository),
     project_repository: Any = Depends(get_project_repository),
 ) -> ListAssetLibrariesResponse:
     user_id = authenticated_user.user.id
     use_case = ListAssetLibrariesUseCase(asset_library_repository)
-    
+
     if project_id:
         # If project_id provided, check access and filter by project
         project = GetProjectUseCase(project_repository).execute(project_id)
@@ -65,7 +67,12 @@ def list_asset_libraries(
         for proj in accessible_projects:
             all_items.extend(use_case.execute(proj.id))
         items = all_items
-    
+
+    # 按 kind 过滤（可选）
+    if kind:
+        kind_enum = AssetLibraryKind(kind)
+        items = [item for item in items if item.kind == kind_enum]
+
     return ListAssetLibrariesResponse(items=[_to_asset_library_response(item) for item in items])
 
 
@@ -90,3 +97,53 @@ def create_asset_library(
         )
     )
     return _to_asset_library_response(item)
+
+
+# 默认素材库名称映射
+_DEFAULT_LIBRARY_NAMES = {
+    "video": "视频素材库",
+    "voice": "配音素材库",
+    "image": "图片素材库",
+}
+
+
+@router.post("/ensure-default", response_model=AssetLibraryResponse)
+def ensure_default_library(
+    request: EnsureDefaultLibraryRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    asset_library_repository: Any = Depends(get_asset_library_repository),
+    project_repository: Any = Depends(get_project_repository),
+) -> AssetLibraryResponse:
+    """确保项目下指定 kind 的默认素材库存在，已存在则直接返回，不存在则自动创建。"""
+    project = project_repository.find_by_id(request.project_id)
+    if project is None:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Project not found")
+    if not project.can_access(authenticated_user.user.id):
+        raise HTTPException(status_code=status.HTTP_403_FORBIDDEN, detail="Access denied to project")
+
+    kind = AssetLibraryKind(request.kind)
+
+    # 查找该项目下同 kind 的素材库，返回第一个
+    existing = asset_library_repository.find_by_project(request.project_id)
+    for lib in existing:
+        if lib.kind == kind:
+            return _to_asset_library_response(lib)
+
+    # 不存在 → 自动创建
+    import uuid
+    from datetime import datetime, timezone
+
+    now = datetime.now(timezone.utc)
+    default_name = _DEFAULT_LIBRARY_NAMES.get(request.kind, f"{request.kind}素材库")
+    library = AssetLibrary(
+        id=str(uuid.uuid4()),
+        project_id=request.project_id,
+        name=default_name,
+        kind=kind,
+        asset_count=0,
+        total_size=0,
+        created_at=now,
+        updated_at=now,
+    )
+    created = asset_library_repository.create(library)
+    return _to_asset_library_response(created)

apps/api/app/api/routes/assets.py

@@ -1,25 +1,53 @@
-from typing import Any
+import logging
+from typing import Any, Optional
 
 from app.auth import AuthenticatedUser, get_current_user
+from app.core.storage import get_storage_service
 from app.dependencies import (
     get_asset_library_repository,
     get_asset_repository,
     get_project_repository,
+    get_tag_repository,
 )
-from app.schemas.asset import AssetResponse, CreateAssetRequest, ListAssetsResponse, UpdateAssetReviewRequest
-from fastapi import APIRouter, Depends, HTTPException
+from app.schemas.asset import (
+    AssetResponse,
+    BatchDeleteRequest,
+    BatchDeleteResponse,
+    CreateAssetRequest,
+    ListAssetsResponse,
+    UpdateAssetRequest,
+    UpdateAssetReviewRequest,
+)
+from app.schemas.tag import TagAssetsRequest
+from fastapi import APIRouter, Depends, HTTPException, Query
 
 from packages.application import (
     CreateAssetCommand,
     CreateAssetUseCase,
-    ListAssetsUseCase,
 )
 from packages.domain import AssetStatus, ClassificationStatus
 
+logger = logging.getLogger(__name__)
+
 router = APIRouter()
 
 
-def _to_asset_response(item) -> AssetResponse:
+def _to_asset_response(item, storage_service=None) -> AssetResponse:
+    # 生成签名文件 URL（用于视频播放 / 文件下载）
+    file_url = None
+    if item.storage_key:
+        try:
+            svc = storage_service or get_storage_service()
+            file_url = svc.get_download_url(item.storage_key)
+        except Exception:
+            logger.warning("生成签名URL失败: storage_key=%s", item.storage_key, exc_info=True)
+            file_url = None
+
+    # 缩略图：优先用已有 thumbnail_url，否则对视频素材复用文件签名 URL
+    thumbnail_url = item.thumbnail_url
+    if not thumbnail_url and item.mime_type and item.mime_type.startswith("video") and file_url:
+        thumbnail_url = file_url
+
     return AssetResponse(
         id=item.id,
         project_id=item.project_id,
@@ -29,7 +57,8 @@ def _to_asset_response(item) -> AssetResponse:
         mime_type=item.mime_type,
         metadata=item.metadata,
         file_size=item.file_size,
-        thumbnail_url=item.thumbnail_url,
+        file_url=file_url,
+        thumbnail_url=thumbnail_url,
         duration=item.duration,
         width=item.width,
         height=item.height,
@@ -39,6 +68,7 @@ def _to_asset_response(item) -> AssetResponse:
         classification_status=item.classification_status.value,
         quality_score=item.quality_score,
         uploaded_by_user_id=item.uploaded_by_user_id,
+        tag_ids=getattr(item, "tag_ids", []),
     )
 
 
@@ -53,19 +83,164 @@ def _check_project_access(project_id: str, user_id: str, project_repository) ->
 
 @router.get("", response_model=ListAssetsResponse)
 def list_assets(
-    library_id: str,
+    library_id: Optional[str] = Query(None),
+    project_id: Optional[str] = Query(None),
+    kind: Optional[str] = Query(None, pattern="^(video|voice|image)$"),
+    keyword: Optional[str] = Query(None, description="按名称模糊匹配"),
+    gender: Optional[str] = Query(None, description="按 metadata.gender 筛选"),
+    style: Optional[str] = Query(None, description="按 metadata.style 筛选"),
+    tag_ids: Optional[str] = Query(None, description="按标签 ID 筛选（逗号分隔，取交集）"),
+    skip: int = Query(0, ge=0),
+    limit: int = Query(100, ge=1, le=500),
     authenticated_user: AuthenticatedUser = Depends(get_current_user),
     asset_repository: Any = Depends(get_asset_repository),
     asset_library_repository: Any = Depends(get_asset_library_repository),
     project_repository: Any = Depends(get_project_repository),
 ) -> ListAssetsResponse:
-    library = asset_library_repository.get(library_id)
-    if library is None:
-        raise HTTPException(status_code=404, detail=f"AssetLibrary {library_id} not found")
-    _check_project_access(library.project_id, authenticated_user.user.id, project_repository)
-    use_case = ListAssetsUseCase(asset_repository)
-    items = use_case.execute(library_id)
-    return ListAssetsResponse(items=[_to_asset_response(item) for item in items])
+    user_id = authenticated_user.user.id
+
+    # kind → file_type 映射（voice 对应 audio）
+    kind_to_file_type = {"video": "video", "voice": "audio", "image": "image"}
+
+    # 解析 tag_ids 参数（逗号分隔）
+    filter_tag_ids: list[str] | None = None
+    if tag_ids:
+        filter_tag_ids = [t.strip() for t in tag_ids.split(",") if t.strip()]
+        if not filter_tag_ids:
+            filter_tag_ids = None
+
+    # 需要内存过滤的标志（keyword/gender/style/tag_ids 无法在 DB 层过滤）
+    needs_memory_filter = bool(keyword or gender or style or filter_tag_ids)
+
+    def _apply_memory_filters(items):
+        """应用 keyword / gender / style / tag_ids 内存过滤。"""
+        result = items
+        if keyword:
+            kw = keyword.lower()
+            result = [i for i in result if kw in (i.name or "").lower()]
+        if gender:
+            result = [i for i in result if (i.metadata or {}).get("gender") == gender]
+        if style:
+            result = [i for i in result if (i.metadata or {}).get("style") == style]
+        if filter_tag_ids:
+            tag_set = set(filter_tag_ids)
+            result = [i for i in result if tag_set.issubset(set(getattr(i, "tag_ids", [])))]
+        return result
+
+    # ── 优化路径：无内存过滤时，使用 DB 级分页 ──
+    if not needs_memory_filter:
+        ft = kind_to_file_type.get(kind) if kind else None
+
+        # 模式1：指定 library_id
+        if library_id:
+            library = asset_library_repository.get(library_id)
+            if library is None:
+                raise HTTPException(status_code=404, detail=f"AssetLibrary {library_id} not found")
+            _check_project_access(library.project_id, user_id, project_repository)
+            if ft:
+                items = asset_repository.find_by_library_and_file_type(library_id, ft, skip=skip, limit=limit)
+                total = asset_repository.count_by_project(library.project_id) if not kind else len(items)
+            else:
+                items = asset_repository.find_by_library(library_id, skip=skip, limit=limit)
+                total = asset_repository.count_by_project(library.project_id)
+            return ListAssetsResponse(
+                items=[_to_asset_response(item) for item in items],
+                total=total,
+                skip=skip,
+                limit=limit,
+            )
+
+        # 模式2：指定 project_id
+        if project_id:
+            _check_project_access(project_id, user_id, project_repository)
+            if ft:
+                # 无直接方法，加载后按 file_type 过滤（仍比全量加载好）
+                all_items = asset_repository.find_by_project(project_id)
+                items = [i for i in all_items if i.mime_type and i.mime_type.startswith(ft)]
+                total = len(items)
+                paged = items[skip : skip + limit]
+            else:
+                items = asset_repository.find_by_project(project_id, skip=skip, limit=limit)
+                total = asset_repository.count_by_project(project_id)
+                paged = items
+            return ListAssetsResponse(
+                items=[_to_asset_response(item) for item in paged],
+                total=total,
+                skip=skip,
+                limit=limit,
+            )
+
+        # 模式3：跨项目（无 library_id/project_id）
+        try:
+            projects = project_repository.find_accessible_projects(user_id)
+        except Exception:
+            logger.exception("查询用户可访问项目失败: user_id=%s", user_id)
+            return ListAssetsResponse(items=[], total=0, skip=skip, limit=limit)
+
+        project_ids = [p.id for p in projects]
+        if not project_ids:
+            return ListAssetsResponse(items=[], total=0, skip=skip, limit=limit)
+
+        total = asset_repository.count_by_project_ids(project_ids)
+        # 跨项目分页：逐项目累积直到凑够一页
+        paged_items: list = []
+        offset = skip
+        remaining = limit
+        for pid in project_ids:
+            proj_total = asset_repository.count_by_project(pid)
+            if offset >= proj_total:
+                offset -= proj_total
+                continue
+            proj_items = asset_repository.find_by_project(pid, skip=offset, limit=remaining)
+            paged_items.extend(proj_items)
+            remaining -= len(proj_items)
+            offset = 0
+            if remaining <= 0:
+                break
+
+        return ListAssetsResponse(
+            items=[_to_asset_response(item) for item in paged_items],
+            total=total,
+            skip=skip,
+            limit=limit,
+        )
+
+    # ── 内存过滤路径：有 keyword/gender/style 时，加载全量后内存过滤 ──
+    if library_id:
+        library = asset_library_repository.get(library_id)
+        if library is None:
+            raise HTTPException(status_code=404, detail=f"AssetLibrary {library_id} not found")
+        _check_project_access(library.project_id, user_id, project_repository)
+        if kind:
+            all_items = asset_repository.find_by_library_and_file_type(library_id, kind_to_file_type[kind])
+        else:
+            all_items = asset_repository.find_by_library(library_id)
+    elif project_id:
+        _check_project_access(project_id, user_id, project_repository)
+        all_items = asset_repository.find_by_project(project_id)
+    else:
+        try:
+            projects = project_repository.find_accessible_projects(user_id)
+        except Exception:
+            logger.exception("查询用户可访问项目失败: user_id=%s", user_id)
+            return ListAssetsResponse(items=[], total=0, skip=skip, limit=limit)
+        all_items = []
+        for proj in projects:
+            all_items.extend(asset_repository.find_by_project(proj.id))
+
+    # 应用 kind 过滤（如果有）+ keyword/gender/style
+    if kind:
+        ft = kind_to_file_type.get(kind)
+        all_items = [i for i in all_items if i.mime_type and i.mime_type.startswith(ft or "")]
+    filtered = _apply_memory_filters(all_items)
+    total = len(filtered)
+    paged = filtered[skip : skip + limit]
+    return ListAssetsResponse(
+        items=[_to_asset_response(item) for item in paged],
+        total=total,
+        skip=skip,
+        limit=limit,
+    )
 
 
 def _apply_asset_review_status(item, review_status: str):
@@ -93,6 +268,130 @@ def update_asset_review_status(
     return _to_asset_response(updated)
 
 
+@router.post("/batch-delete", response_model=BatchDeleteResponse)
+def batch_delete_assets(
+    request: BatchDeleteRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    asset_repository: Any = Depends(get_asset_repository),
+    project_repository: Any = Depends(get_project_repository),
+) -> BatchDeleteResponse:
+    """批量删除素材（配音素材等），需逐项校验项目权限。"""
+    user_id = authenticated_user.user.id
+    deleted_ids: list[str] = []
+    failed_ids: list[str] = []
+
+    for asset_id in request.ids:
+        item = asset_repository.find_by_id(asset_id)
+        if item is None:
+            failed_ids.append(asset_id)
+            continue
+        try:
+            _check_project_access(item.project_id, user_id, project_repository)
+            deleted_ids.append(asset_id)
+        except HTTPException:
+            failed_ids.append(asset_id)
+
+    if deleted_ids:
+        asset_repository.batch_delete(deleted_ids)
+
+    return BatchDeleteResponse(deleted_count=len(deleted_ids), failed_ids=failed_ids)
+
+
+@router.get("/{asset_id}", response_model=AssetResponse)
+def get_asset(
+    asset_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    asset_repository: Any = Depends(get_asset_repository),
+    project_repository: Any = Depends(get_project_repository),
+) -> AssetResponse:
+    item = asset_repository.find_by_id(asset_id)
+    if item is None:
+        raise HTTPException(status_code=404, detail=f"Asset {asset_id} not found")
+    _check_project_access(item.project_id, authenticated_user.user.id, project_repository)
+    return _to_asset_response(item)
+
+
+@router.put("/{asset_id}", response_model=AssetResponse)
+def update_asset(
+    asset_id: str,
+    request: UpdateAssetRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    asset_repository: Any = Depends(get_asset_repository),
+    project_repository: Any = Depends(get_project_repository),
+) -> AssetResponse:
+    item = asset_repository.find_by_id(asset_id)
+    if item is None:
+        raise HTTPException(status_code=404, detail=f"Asset {asset_id} not found")
+    _check_project_access(item.project_id, authenticated_user.user.id, project_repository)
+
+    # 合并可修改字段
+    if request.name is not None:
+        item.name = request.name
+    if request.metadata is not None:
+        item.metadata = {**item.metadata, **request.metadata}
+    if request.tags is not None:
+        item.metadata = {**item.metadata, "tags": request.tags}
+
+    updated = asset_repository.update(item)
+    return _to_asset_response(updated)
+
+
+@router.delete("/{asset_id}", status_code=204)
+def delete_asset(
+    asset_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    asset_repository: Any = Depends(get_asset_repository),
+    project_repository: Any = Depends(get_project_repository),
+) -> None:
+    item = asset_repository.find_by_id(asset_id)
+    if item is None:
+        raise HTTPException(status_code=404, detail=f"Asset {asset_id} not found")
+    _check_project_access(item.project_id, authenticated_user.user.id, project_repository)
+    asset_repository.delete(asset_id)
+
+
+@router.post("/{asset_id}/tags", response_model=AssetResponse)
+def tag_asset(
+    asset_id: str,
+    request: TagAssetsRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    asset_repository: Any = Depends(get_asset_repository),
+    project_repository: Any = Depends(get_project_repository),
+    tag_repository: Any = Depends(get_tag_repository),
+) -> AssetResponse:
+    """给素材打标签。"""
+    item = asset_repository.find_by_id(asset_id)
+    if item is None:
+        raise HTTPException(status_code=404, detail=f"Asset {asset_id} not found")
+    _check_project_access(item.project_id, authenticated_user.user.id, project_repository)
+    for tag_id in request.tag_ids:
+        tag = tag_repository.get(tag_id)
+        if tag is None:
+            raise HTTPException(status_code=404, detail=f"Tag {tag_id} not found")
+        if tag.user_id != authenticated_user.user.id:
+            raise HTTPException(status_code=403, detail=f"无权使用标签 {tag_id}")
+        item.add_tag(tag_id)
+    updated = asset_repository.update(item)
+    return _to_asset_response(updated)
+
+
+@router.delete("/{asset_id}/tags/{tag_id}", status_code=204)
+def untag_asset(
+    asset_id: str,
+    tag_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    asset_repository: Any = Depends(get_asset_repository),
+    project_repository: Any = Depends(get_project_repository),
+) -> None:
+    """取消素材的标签。"""
+    item = asset_repository.find_by_id(asset_id)
+    if item is None:
+        raise HTTPException(status_code=404, detail=f"Asset {asset_id} not found")
+    _check_project_access(item.project_id, authenticated_user.user.id, project_repository)
+    item.remove_tag(tag_id)
+    asset_repository.update(item)
+
+
 @router.post("", response_model=AssetResponse)
 def create_asset(
     request: CreateAssetRequest,
@@ -106,7 +405,7 @@ def create_asset(
         raise HTTPException(status_code=404, detail=f"Project {request.project_id} not found")
     if not project.can_access(authenticated_user.user.id):
         raise HTTPException(status_code=403, detail="Access denied to project")
-    
+
     library = asset_library_repository.get(request.library_id)
     if library is None or library.project_id != request.project_id:
         raise HTTPException(status_code=404, detail=f"AssetLibrary {request.library_id} not found")

apps/api/app/api/routes/auth.py

@@ -5,12 +5,16 @@ The route layer is intentionally thin: repository construction lives in
 app.dependencies and authentication behavior lives in application use cases.
 """
 
+import logging
+import os
 from typing import Optional
 
-from app.auth import AuthenticatedUser, get_current_user
+import jwt
+from app.auth import AuthenticatedUser, blacklist_token, get_current_user
 from app.config import settings
 from app.dependencies import get_auth_email_service, get_auth_session_store, get_user_repository
-from fastapi import APIRouter, Depends, HTTPException, status
+from fastapi import APIRouter, Depends, Header, HTTPException, status
+from fastapi.security import HTTPAuthorizationCredentials, HTTPBearer
 from pydantic import BaseModel, EmailStr
 
 from packages.adapters.redis import NoopSessionStore
@@ -29,6 +33,10 @@ from packages.application.auth.register_user_use_case import RegisterUserRequest
 from packages.application.auth.register_user_use_case import RegisterUserUseCase, VerifyEmailRequest, VerifyEmailUseCase
 from packages.ports.user_repository import UserRepository
 
+logger = logging.getLogger(__name__)
+
+bearer_scheme = HTTPBearer(auto_error=False)
+
 router = APIRouter(prefix="/auth", tags=["认证"])
 
 
@@ -55,6 +63,7 @@ class LoginRequest(BaseModel):
 class RefreshRequest(BaseModel):
     refresh_token: str
 
+
 class LoginResponse(BaseModel):
     access_token: str
     refresh_token: str
@@ -173,7 +182,6 @@ async def refresh(
     )
 
 
-
 def _verify_email_token(token: str, user_repository: UserRepository) -> MessageResponse:
     success, error = VerifyEmailUseCase(user_repository=user_repository).execute(VerifyEmailRequest(token=token))
     if not success:
@@ -229,6 +237,24 @@ async def reset_password(
     return MessageResponse(message="密码重置成功")
 
 
+@router.post("/logout")
+async def logout(
+    credentials: HTTPAuthorizationCredentials | None = Depends(bearer_scheme),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+):
+    """登出 - 将当前 token 加入黑名单"""
+    from fastapi.security import HTTPAuthorizationCredentials, HTTPBearer
+
+    if credentials:
+        try:
+            payload = jwt.decode(credentials.credentials, settings.JWT_SECRET_KEY, algorithms=["HS256"])
+            exp = payload.get("exp", 0)
+            blacklist_token(credentials.credentials, exp)
+        except Exception as e:
+            logger.warning(f"Operation failed in apps/api/app/api/routes/auth.py: {e}", exc_info=True)
+    return MessageResponse(message="已登出")
+
+
 @router.get("/me", response_model=CurrentUserResponse)
 async def get_current_user_info(
     authenticated_user: AuthenticatedUser = Depends(get_current_user),
@@ -259,3 +285,99 @@ def _translate_auth_error(error: str | None) -> str:
         "Display name is required": "显示名称不能为空",
     }
     return translations.get(error or "", error or "注册失败")
+
+
+class WechatSyncRequest(BaseModel):
+    openid: str
+    unionid: Optional[str] = None
+    nickname: Optional[str] = None
+    avatar_url: Optional[str] = None
+    source: str = "miniapp"
+
+
+class WechatSyncResponse(BaseModel):
+    access_token: str
+    token: str
+    refresh_token: str
+    user_id: str
+    user: dict
+    user_info: dict
+    is_new_user: bool
+    expires_in: int
+
+
+def _get_internal_api_keys() -> list[str]:
+    """获取内部 API Key 列表
+
+    优先级:
+    1. INTERNAL_API_KEYS 环境变量
+    2. /app/generated/internal_api_keys.txt 文件 (volume 持久化)
+    """
+    env_keys = os.environ.get("INTERNAL_API_KEYS", "")
+    if env_keys:
+        return [k.strip() for k in env_keys.split(",") if k.strip()]
+
+    # 从持久化文件读取
+    try:
+        with open("/app/generated/internal_api_keys.txt", "r") as f:
+            content = f.read().strip()
+            if content:
+                return [k.strip() for k in content.split(",") if k.strip()]
+    except Exception:
+        logger.debug("Failed to read internal API keys from file", exc_info=True)
+
+    return []
+
+
+def _verify_internal_api_key(x_api_key: str | None = Header(None)) -> bool:
+    """验证内部 API Key
+
+    - 已配置时：必须匹配 INTERNAL_API_KEYS 中的 key
+    - 未配置且非生产环境：放行（方便开发）
+    - 未配置且生产环境：拒绝
+    """
+    env = os.environ.get("APP_ENV", os.environ.get("ENV", "development")).lower()
+    key_list = _get_internal_api_keys()
+
+    if not key_list:
+        if env in ("production", "prod"):
+            raise HTTPException(status_code=401, detail="内部接口未配置 API Key")
+        return True
+
+    if x_api_key and x_api_key.strip() in key_list:
+        return True
+
+    raise HTTPException(status_code=401, detail="无效的 API Key")
+
+
+@router.post("/wechat-sync", response_model=WechatSyncResponse, include_in_schema=False)
+async def wechat_sync(
+    request: WechatSyncRequest,
+    user_repository: UserRepository = Depends(get_user_repository),
+    _: bool = Depends(_verify_internal_api_key),
+):
+    """
+    微信同步登录/注册（系统级内部接口）
+
+    由 BFF 层通过 API Key 调用，不直接面向终端用户。
+    根据 openid 查找或创建用户，返回 SaaS token。
+    """
+    from packages.application.auth.wechat_sync_use_case import WechatSyncRequest as UseCaseRequest
+    from packages.application.auth.wechat_sync_use_case import (
+        WechatSyncUseCase,
+    )
+
+    use_case = WechatSyncUseCase(user_repository=user_repository)
+    use_case_request = UseCaseRequest(
+        openid=request.openid,
+        unionid=request.unionid,
+        nickname=request.nickname,
+        avatar_url=request.avatar_url,
+        source=request.source,
+    )
+
+    response, error = use_case.execute(use_case_request)
+    if error:
+        raise HTTPException(status_code=400, detail=error)
+
+    return WechatSyncResponse(**response.to_dict())

apps/api/app/api/routes/chunked_upload.py

@@ -6,8 +6,8 @@ Supports chunked upload, resume, and automatic cleanup of expired uploads.
 import fcntl
 import json
 import logging
-import os
 import shutil
+import tempfile
 from datetime import datetime, timedelta, timezone
 from pathlib import Path
 from typing import Any
@@ -19,6 +19,7 @@ from app.core.celery_app import celery_app
 from app.core.storage import OSSStorageService, get_storage_service
 from app.dependencies import (
     get_asset_library_repository,
+    get_asset_repository,
     get_ingest_job_repository,
     get_project_repository,
 )
@@ -42,15 +43,37 @@ DEFAULT_CHUNK_SIZE = 5 * 1024 * 1024  # 5MB
 MAX_FILE_SIZE = 2 * 1024 * 1024 * 1024  # 2GB
 CHUNK_EXPIRY_HOURS = 24
 
-# Allowed file types (consistent with existing upload.py)
+# Allowed file types — must stay in sync with upload.py ALLOWED_MIME_TYPES
 ALLOWED_MIME_TYPES = {
-    "image/jpeg", "image/png", "image/gif", "image/webp",
-    "video/mp4", "video/quicktime", "video/x-msvideo", "video/webm",
-    "audio/mpeg", "audio/wav", "audio/ogg", "audio/mp3",
+    # Images
+    "image/jpeg",
+    "image/png",
+    "image/gif",
+    "image/webp",
+    "image/bmp",
+    "image/tiff",
+    "image/svg+xml",
+    # Video
+    "video/mp4",
+    "video/quicktime",
+    "video/mpeg",
+    "video/x-msvideo",
+    "video/webm",
+    "video/x-matroska",
+    "video/3gpp",
+    # Audio
+    "audio/mpeg",
+    "audio/wav",
+    "audio/ogg",
+    "audio/mp3",
+    "audio/flac",
+    "audio/aac",
+    "audio/x-m4a",
+    "audio/webm",
 }
 
 # Chunk storage root directory
-CHUNK_STORAGE_ROOT = Path("/tmp/chunked_uploads")
+CHUNK_STORAGE_ROOT = Path(tempfile.gettempdir()) / "chunked_uploads"
 
 
 def _get_chunk_dir(upload_id: str) -> Path:
@@ -67,13 +90,13 @@ def _atomic_check_and_record(upload_id: str, chunk_index: int) -> bool:
     """
     Atomically check if chunk is uploaded and record if not.
     Uses file locking to prevent race conditions.
-    
+
     Returns:
         True if chunk was newly recorded, False if already exists
     """
     meta_path = _get_upload_meta_path(upload_id)
     CHUNK_STORAGE_ROOT.mkdir(parents=True, exist_ok=True)
-    
+
     with open(meta_path, "r+", encoding="utf-8") as f:
         fcntl.flock(f.fileno(), fcntl.LOCK_EX)
         try:
@@ -101,7 +124,7 @@ def _require_project_and_library(
     if project is None:
         raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Project not found")
 
-    libraries = asset_library_repository.list_by_project(project_id)
+    libraries = asset_library_repository.find_by_project(project_id)
     if not any(item.id == library_id for item in libraries):
         raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Asset library not found")
 
@@ -128,15 +151,17 @@ def _validate_file_type(content: bytes, filename: str) -> str:
     """Validate file type"""
     try:
         import magic
+
         detected_mime = magic.from_buffer(content, mime=True)
     except ImportError:
         import mimetypes
+
         detected_mime = mimetypes.guess_type(filename)[0] or "application/octet-stream"
 
     if detected_mime not in ALLOWED_MIME_TYPES:
         raise HTTPException(
             status_code=status.HTTP_400_BAD_REQUEST,
-            detail=f"Unsupported file type: {detected_mime}. Allowed types: {', '.join(sorted(ALLOWED_MIME_TYPES))}"
+            detail=f"Unsupported file type: {detected_mime}. Allowed types: {', '.join(sorted(ALLOWED_MIME_TYPES))}",
         )
     return detected_mime
 
@@ -349,6 +374,7 @@ async def complete_chunked_upload(
     authenticated_user: AuthenticatedUser = Depends(get_current_user),
     project_repository: Any = Depends(get_project_repository),
     asset_library_repository: Any = Depends(get_asset_library_repository),
+    asset_repository: Any = Depends(get_asset_repository),
     ingest_job_repository: Any = Depends(get_ingest_job_repository),
     storage_service: OSSStorageService = Depends(get_storage_service),
 ) -> ChunkedUploadCompleteResponse:
@@ -412,6 +438,29 @@ async def complete_chunked_upload(
             content_type=meta["content_type"],
         )
 
+        # ── 素材去重检测：同素材库 + 同 file_hash 视为重复 ──
+        if request.file_hash:
+            existing = asset_repository.find_by_library_and_file_hash(
+                library_id=request.library_id,
+                file_hash=request.file_hash,
+            )
+            if existing is not None:
+                logger.info(
+                    "素材去重命中(chunked): library=%s hash=%s existing_asset=%s",
+                    request.library_id,
+                    request.file_hash,
+                    existing.id,
+                )
+                meta["status"] = "completed"
+                _save_upload_meta(upload_id, meta)
+                return ChunkedUploadCompleteResponse(
+                    storage_key=storage_key,
+                    ingest_job_id="",
+                    url=file_url,
+                    duplicated=True,
+                    asset_id=existing.id,
+                )
+
         # Create ingest job
         use_case = SubmitIngestJobUseCase(ingest_job_repository)
         job = use_case.execute(
@@ -419,6 +468,7 @@ async def complete_chunked_upload(
                 project_id=meta["project_id"],
                 library_id=meta["library_id"],
                 storage_key=storage_key,
+                file_hash=request.file_hash,
             )
         )
         celery_app.send_task("worker.ingest_asset", args=[job.id])

apps/api/app/api/routes/classification_jobs.py

@@ -1,4 +1,3 @@
-from datetime import datetime, timezone
 from typing import Any
 
 from app.core.celery_app import celery_app

apps/api/app/api/routes/dashboard.py

@@ -0,0 +1,92 @@
+from typing import Any
+
+from app.auth import AuthenticatedUser, get_current_user
+from app.dependencies import (
+    get_asset_repository,
+    get_generation_task_repository,
+    get_project_repository,
+    get_title_library_repository,
+    get_voice_library_repository,
+)
+from app.schemas.dashboard import DashboardOverviewResponse, RecentTaskItem, SubscriptionInfo
+from fastapi import APIRouter, Depends
+
+router = APIRouter()
+
+
+def _status_value(status) -> str:
+    return status.value if hasattr(status, "value") else str(status)
+
+
+def _generation_step(status: str) -> str:
+    if status == "pending":
+        return "等待 Worker 执行"
+    if status == "running":
+        return "正在生成成片"
+    if status == "completed":
+        return "生成完成"
+    if status == "failed":
+        return "生成失败"
+    return status
+
+
+@router.get("/overview", response_model=DashboardOverviewResponse)
+def get_dashboard_overview(
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    project_repository: Any = Depends(get_project_repository),
+    asset_repository: Any = Depends(get_asset_repository),
+    generation_task_repository: Any = Depends(get_generation_task_repository),
+    title_library_repository: Any = Depends(get_title_library_repository),
+    voice_library_repository: Any = Depends(get_voice_library_repository),
+) -> DashboardOverviewResponse:
+    """Dashboard 概览：用户级汇总数据。"""
+    user_id = authenticated_user.user.id
+
+    # 获取用户可访问的所有 project
+    projects = project_repository.find_accessible_projects(user_id)
+    project_ids = [p.id for p in projects]
+
+    # 素材统计
+    total_assets = asset_repository.count_by_project_ids(project_ids)
+    used_storage_bytes = asset_repository.sum_storage_by_project_ids(project_ids)
+
+    # 标题库 / 配音库统计
+    total_titles = title_library_repository.count_by_user(user_id)
+    total_voices = voice_library_repository.count_by_user(user_id)
+
+    # 生成任务统计
+    total_tasks = generation_task_repository.count_by_user(user_id)
+
+    # 最近任务（SQL 层 LIMIT 5）
+    recent = generation_task_repository.list_recent_by_user(user_id, limit=5)
+    recent_tasks = []
+    for task in recent:
+        s = _status_value(task.status)
+        recent_tasks.append(
+            RecentTaskItem(
+                id=task.id,
+                task_type="generation",
+                status=s,
+                current_step=_generation_step(s),
+                error_message=task.error_message or "",
+                updated_at=task.completed_at or task.started_at or task.created_at,
+            )
+        )
+
+    # 订阅信息
+    user = authenticated_user.user
+    subscription = SubscriptionInfo(
+        plan=getattr(user, "subscription_plan", "free") or "free",
+        is_active=getattr(user, "subscription_status", "") == "active",
+    )
+
+    return DashboardOverviewResponse(
+        total_assets=total_assets,
+        used_storage_bytes=used_storage_bytes,
+        total_titles=total_titles,
+        total_voices=total_voices,
+        total_tasks=total_tasks,
+        total_products=len(projects),
+        subscription=subscription,
+        recent_tasks=recent_tasks,
+    )

apps/api/app/api/routes/duplication.py

@@ -1,4 +1,5 @@
 """查重 API 路由。"""
+
 from __future__ import annotations
 
 import logging
@@ -9,12 +10,12 @@ from app.auth import AuthenticatedUser, get_current_user
 from app.core.storage import OSSStorageService, get_storage_service
 from app.dependencies import get_duplication_repository
 from app.schemas.duplication import (
+    DuplicateSegmentResponse,
     DuplicationDetailResponse,
     DuplicationRecordResponse,
     DuplicationUploadResponse,
-    DuplicateSegmentResponse,
 )
-from fastapi import APIRouter, Depends, File, HTTPException, Response, UploadFile, status
+from fastapi import APIRouter, Depends, File, HTTPException, Query, Response, UploadFile, status
 
 from packages.application import (
     DeleteDuplicationRecordUseCase,
@@ -28,13 +29,22 @@ from packages.domain.duplication import DuplicationRecord
 
 logger = logging.getLogger(__name__)
 
-router = APIRouter()
+router = APIRouter(
+    tags=["查重"],
+)
 
 # 查重功能只接受视频文件
-ALLOWED_VIDEO_MIME_TYPES = frozenset({
-    "video/mp4", "video/mpeg", "video/quicktime", "video/x-msvideo",
-    "video/webm", "video/x-matroska", "video/3gpp",
-})
+ALLOWED_VIDEO_MIME_TYPES = frozenset(
+    {
+        "video/mp4",
+        "video/mpeg",
+        "video/quicktime",
+        "video/x-msvideo",
+        "video/webm",
+        "video/x-matroska",
+        "video/3gpp",
+    }
+)
 
 
 def _validate_video_mime_type(content_type: str | None) -> str:
@@ -44,16 +54,16 @@ def _validate_video_mime_type(content_type: str | None) -> str:
             status_code=status.HTTP_400_BAD_REQUEST,
             detail="Content-Type header is required",
         )
-    
+
     # 处理带参数的类型，如 "video/mp4; charset=utf-8"
     base_type = content_type.split(";")[0].strip().lower()
-    
+
     if base_type not in ALLOWED_VIDEO_MIME_TYPES:
         raise HTTPException(
             status_code=status.HTTP_415_UNSUPPORTED_MEDIA_TYPE,
-            detail=f"只支持视频文件。支持的类型: mp4, mpeg, mov, avi, webm, mkv, 3gp",
+            detail="只支持视频文件。支持的类型: mp4, mpeg, mov, avi, webm, mkv, 3gp",
         )
-    
+
     return base_type
 
 
@@ -117,9 +127,10 @@ async def upload_for_duplication(
 
     # P0-2: 验证文件大小（参考 OSS_DIRECT_UPLOAD_MAX_MB）
     from app.config import get_settings
+
     settings = get_settings()
     max_size_bytes = settings.OSS_DIRECT_UPLOAD_MAX_MB * 1024 * 1024
-    
+
     # 先检查 Content-Length header（如果可用）
     if file.size is not None and file.size > max_size_bytes:
         raise HTTPException(
@@ -135,7 +146,7 @@ async def upload_for_duplication(
     try:
         content = await file.read()
         file_size = len(content)
-        
+
         # 再次检查实际文件大小
         if file_size > max_size_bytes:
             raise HTTPException(
@@ -145,9 +156,10 @@ async def upload_for_duplication(
     except HTTPException:
         raise
     except Exception as exc:
+        logger.error("读取查重文件失败: %s", exc, exc_info=True)
         raise HTTPException(
             status_code=status.HTTP_400_BAD_REQUEST,
-            detail=f"读取文件失败: {exc}",
+            detail="文件读取失败，请稍后重试",
         ) from exc
 
     try:
@@ -157,9 +169,10 @@ async def upload_for_duplication(
             content_type=validated_content_type,
         )
     except Exception as exc:
+        logger.error("查重文件上传 OSS 失败: %s", exc, exc_info=True)
         raise HTTPException(
             status_code=status.HTTP_503_SERVICE_UNAVAILABLE,
-            detail=f"文件上传失败: {exc}",
+            detail="文件上传失败，请稍后重试",
         ) from exc
 
     use_case = UploadForDuplicationUseCase(duplication_repository)
@@ -188,12 +201,19 @@ async def upload_for_duplication(
 
 @router.get("/records", response_model=list[DuplicationRecordResponse])
 def list_duplication_records(
+    offset: int = Query(0, ge=0, description="分页偏移量"),
+    limit: int = Query(50, ge=1, le=200, description="每页数量，最大 200"),
     authenticated_user: AuthenticatedUser = Depends(get_current_user),
     duplication_repository: Any = Depends(get_duplication_repository),
 ) -> list[DuplicationRecordResponse]:
-    """获取当前用户的查重记录列表。"""
+    """
+    获取当前用户的查重记录列表。
+
+    支持分页：通过 offset 和 limit 参数控制。
+    返回按创建时间倒序排列的记录。
+    """
     use_case = ListDuplicationRecordsUseCase(duplication_repository)
-    records = use_case.execute(authenticated_user.user.id)
+    records = use_case.execute(user_id=authenticated_user.user.id, offset=offset, limit=limit)
     return [_to_record_response(r) for r in records]
 
 
@@ -219,7 +239,7 @@ def get_duplication_detail(
     return _to_detail_response(record)
 
 
-@router.delete("/records/{record_id}", status_code=status.HTTP_204_NO_CONTENT, response_class=Response)
+@router.delete("/records/{record_id}", status_code=status.HTTP_204_NO_CONTENT, response_model=None)
 def delete_duplication_record(
     record_id: str,
     authenticated_user: AuthenticatedUser = Depends(get_current_user),
@@ -246,7 +266,11 @@ def retry_duplication(
     authenticated_user: AuthenticatedUser = Depends(get_current_user),
     duplication_repository: Any = Depends(get_duplication_repository),
 ) -> DuplicationUploadResponse:
-    """重新提交查重。"""
+    """
+    重新提交查重。
+
+    仅 failed 状态的记录允许重试，其他状态返回 400。
+    """
     # 检查记录存在且属于当前用户
     detail_uc = GetDuplicationDetailUseCase(duplication_repository)
     record = detail_uc.execute(record_id)
@@ -257,7 +281,13 @@ def retry_duplication(
         )
 
     use_case = RetryDuplicationUseCase(duplication_repository)
-    updated = use_case.execute(record_id)
+    try:
+        updated = use_case.execute(record_id)
+    except ValueError as e:
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=str(e),
+        )
     if updated is None:
         raise HTTPException(
             status_code=status.HTTP_404_NOT_FOUND,

apps/api/app/api/routes/edit_plans.py

@@ -0,0 +1,935 @@
+"""剪辑计划管理 API — Phase 8 模板编排引擎.
+
+RESTful CRUD for EditPlan:
+- GET    /api/v1/edit-plans            列表（分页 + 状态/模板筛选）
+- GET    /api/v1/edit-plans/{id}       详情
+- POST   /api/v1/edit-plans            创建
+- PUT    /api/v1/edit-plans/{id}       更新（含状态机流转）
+- DELETE /api/v1/edit-plans/{id}       删除
+- POST   /api/v1/edit-plans/{id}/generate          触发剪辑渲染生成（任务 2.05）
+- GET    /api/v1/edit-plans/{id}/generation-status  查询生成进度（任务 2.05）
+- POST   /api/v1/edit-plans/{id}/ai-recommend       AI 推荐片段方案（任务 3.09）
+- POST   /api/v1/edit-plans/{id}/generate-cover     AI 生成封面（任务 3.09）
+
+业务逻辑委托给 EditPlanService 服务层。
+"""
+
+from __future__ import annotations
+
+import logging
+from datetime import datetime
+from typing import Any, List, Optional
+
+from app.auth import AuthenticatedUser, get_current_user
+from app.core.celery_app import celery_app
+from app.dependencies import get_db_session, get_project_repository
+from app.schemas.generation_task import GenerationTaskResponse
+from app.services import EditPlanService
+from fastapi import APIRouter, Depends, HTTPException, Query, status
+from pydantic import BaseModel, Field
+from sqlalchemy.orm import Session
+
+from packages.adapters.sqlalchemy_impl.generation_task_repository import (
+    SQLAlchemyGenerationTaskRepository,
+)
+from packages.application.generation_tasks import (
+    CreateGenerationTaskCommand,
+    CreateGenerationTaskUseCase,
+)
+from packages.domain.config_schemas import normalize_plan_config
+from packages.domain.edit_plan import EditPlan, EditPlanStatus
+
+logger = logging.getLogger(__name__)
+
+router = APIRouter()
+
+
+# ── Pydantic Schemas ─────────────────────────────────────────────────────────
+
+
+class EditPlanCreateRequest(BaseModel):
+    """创建剪辑计划请求体"""
+
+    template_id: str = Field(..., min_length=1, max_length=32, description="关联模板 ID")
+    name: str = Field(..., min_length=1, max_length=200, description="计划名称")
+    config: dict[str, Any] = Field(default_factory=dict, description="计划配置 (JSON)")
+    total_duration: float = Field(default=0.0, ge=0.0, description="总时长 (秒)")
+    project_id: str = Field(default="", description="所属项目 ID")
+
+
+class EditPlanUpdateRequest(BaseModel):
+    """更新剪辑计划请求体"""
+
+    name: Optional[str] = Field(default=None, min_length=1, max_length=200, description="计划名称")
+    config: Optional[dict[str, Any]] = Field(default=None, description="计划配置 (JSON)")
+    total_duration: Optional[float] = Field(default=None, ge=0.0, description="总时长 (秒)")
+    status: Optional[str] = Field(
+        default=None,
+        description="目标状态 (通过状态机流转): editing / rendering / completed / failed / draft",
+    )
+
+
+class EditPlanResponse(BaseModel):
+    """剪辑计划响应体"""
+
+    id: str
+    template_id: str
+    name: str
+    status: str
+    total_duration: float
+    project_id: str = ""
+    created_by_user_id: str = ""
+    config: dict[str, Any]
+    created_at: datetime
+    updated_at: datetime
+
+    model_config = {"from_attributes": True}
+
+
+class EditPlanListResponse(BaseModel):
+    """剪辑计划列表响应体"""
+
+    items: List[EditPlanResponse]
+    total: int
+    page: int
+    page_size: int
+
+
+class ClipStatusItem(BaseModel):
+    """片段生成状态"""
+
+    clip_id: str
+    clip_type: str
+    order: int
+    status: str
+    asset_id: str
+    text_content: str
+    duration: float
+
+
+class EditPlanGenerationStatusResponse(BaseModel):
+    """剪辑计划生成进度响应体"""
+
+    plan_id: str
+    plan_status: str
+    generation_task_id: Optional[str] = None
+    clips: List[ClipStatusItem]
+
+
+class EditPlanGenerateResponse(BaseModel):
+    """剪辑计划触发生成响应体"""
+
+    plan_id: str
+    plan_status: str
+    generation_task_id: str
+    clip_count: int
+
+
+class EditPlanGenerationsResponse(BaseModel):
+    """剪辑计划关联的生成记录列表响应体"""
+
+    items: List[GenerationTaskResponse]
+    total: int
+
+
+# ── AI 推荐片段方案 Schemas（任务 3.09） ──────────────────────────────────────
+
+
+class AIRecommendRequest(BaseModel):
+    """AI 推荐片段方案请求体"""
+
+    asset_ids: List[str] = Field(default_factory=list, description="素材 ID 列表")
+    editing_mode: str = Field(default="one_take", description="剪辑模式: one_take / pip / voice_over / voice_pip")
+    target_duration: float = Field(default=30.0, ge=1.0, le=600.0, description="目标时长（秒）")
+
+
+class AIRecommendClipItem(BaseModel):
+    """AI 推荐的单个片段"""
+
+    clip_type: str = Field(..., description="片段类型: intro / showcase / title / subtitle / cta / outro")
+    order: int = Field(..., ge=0, description="片段顺序")
+    text_content: str = Field(default="", description="文字内容")
+    duration: float = Field(..., ge=0.0, description="片段时长（秒）")
+    transition_effect: str = Field(default="cut", description="转场效果")
+    asset_id: str = Field(default="", description="关联素材 ID")
+    start_time: float = Field(default=0.0, ge=0.0, description="素材截取起始时间（秒）")
+    config: dict[str, Any] = Field(default_factory=dict, description="片段额外配置")
+
+
+class AIRecommendResponse(BaseModel):
+    """AI 推荐片段方案响应体"""
+
+    plan_id: str = Field(..., description="剪辑计划 ID")
+    clips: List[AIRecommendClipItem] = Field(..., description="推荐的片段列表")
+    config: dict[str, Any] = Field(..., description="推荐的 plan config（cover/title/subtitle/bgm）")
+    total_duration: float = Field(..., ge=0.0, description="推荐方案总时长（秒）")
+    confidence: float = Field(..., ge=0.0, le=1.0, description="AI 推荐置信度 (0~1)")
+
+
+# ── AI 封面生成 Schemas（任务 3.09） ─────────────────────────────────────────
+
+
+class GenerateCoverRequest(BaseModel):
+    """AI 封面生成请求体"""
+
+    asset_ids: List[str] = Field(default_factory=list, description="素材 ID 列表（确定视频来源）")
+    cover_type: str = Field(
+        default="ai_frame",
+        description="封面类型: ai_frame / manual / upload / ai_regenerate",
+    )
+    frame_time: Optional[float] = Field(
+        default=None,
+        ge=0.0,
+        description="手动选帧时间点（秒），仅 cover_type=manual 时有效",
+    )
+
+
+class GenerateCoverResponse(BaseModel):
+    """AI 封面生成响应体"""
+
+    plan_id: str = Field(..., description="剪辑计划 ID")
+    cover: dict[str, Any] = Field(..., description="封面数据（type / image_url / frame_time 等）")
+
+
+# ── Helpers ───────────────────────────────────────────────────────────────────
+
+
+def _check_project_access(project_id: str, user_id: str, project_repository: Any) -> None:
+    """校验用户对项目的访问权限（参照 assets.py 的 can_access 模式）"""
+    if not project_id or not project_id.strip():
+        return
+    project = project_repository.find_by_id(project_id)
+    if project is None:
+        raise HTTPException(status_code=404, detail=f"Project {project_id} not found")
+    if not project.can_access(user_id):
+        raise HTTPException(status_code=403, detail="Access denied to project")
+
+
+def _to_response(p: EditPlan) -> EditPlanResponse:
+    return EditPlanResponse(
+        id=p.id,
+        template_id=p.template_id,
+        name=p.name,
+        status=p.status.value if hasattr(p.status, "value") else p.status,
+        total_duration=p.total_duration,
+        project_id=p.project_id or "",
+        created_by_user_id=p.created_by_user_id or "",
+        config=p.config,
+        created_at=p.created_at,
+        updated_at=p.updated_at,
+    )
+
+
+# ── Routes ────────────────────────────────────────────────────────────────────
+
+
+@router.get("", response_model=EditPlanListResponse)
+def list_plans(
+    page: int = Query(default=1, ge=1, description="页码"),
+    page_size: int = Query(default=20, ge=1, le=100, description="每页数量"),
+    template_id: Optional[str] = Query(default=None, description="按模板 ID 筛选"),
+    project_id: Optional[str] = Query(default=None, description="按项目 ID 筛选"),
+    status_filter: Optional[str] = Query(
+        default=None,
+        alias="status",
+        description="按状态筛选: draft / editing / rendering / completed / failed",
+    ),
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+    project_repository: Any = Depends(get_project_repository),
+) -> EditPlanListResponse:
+    """获取剪辑计划列表（支持分页、按模板/状态/项目筛选）"""
+    svc = EditPlanService(db)
+
+    # 空串 project_id 视为未传（避免 DB 中匹配到空串记录）
+    if project_id is not None and not project_id.strip():
+        project_id = None
+
+    # 解析状态筛选
+    status_enum: Optional[EditPlanStatus] = None
+    if status_filter:
+        try:
+            status_enum = EditPlanStatus(status_filter)
+        except ValueError:
+            raise HTTPException(
+                status_code=status.HTTP_400_BAD_REQUEST,
+                detail=(f"无效的状态值: {status_filter}，" f"可选值: draft, editing, rendering, completed, failed"),
+            )
+
+    # 项目鉴权：如果指定了 project_id，校验用户是否有权访问
+    if project_id:
+        _check_project_access(project_id, current_user.user.id, project_repository)
+
+    skip = (page - 1) * page_size
+    plans = svc.list_plans(
+        template_id=template_id,
+        project_id=project_id,
+        status=status_enum,
+        skip=skip,
+        limit=page_size,
+    )
+    total = svc.count_plans(
+        template_id=template_id,
+        project_id=project_id,
+        status=status_enum,
+    )
+
+    return EditPlanListResponse(
+        items=[_to_response(p) for p in plans],
+        total=total,
+        page=page,
+        page_size=page_size,
+    )
+
+
+@router.get("/{plan_id}", response_model=EditPlanResponse)
+def get_plan(
+    plan_id: str,
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+    project_repository: Any = Depends(get_project_repository),
+) -> EditPlanResponse:
+    """获取单个剪辑计划详情"""
+    svc = EditPlanService(db)
+    try:
+        plan = svc.get_plan_or_raise(plan_id)
+    except ValueError as exc:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND,
+            detail=str(exc),
+        )
+    # 项目鉴权
+    if plan.project_id:
+        _check_project_access(plan.project_id, current_user.user.id, project_repository)
+    return _to_response(plan)
+
+
+@router.post("", response_model=EditPlanResponse, status_code=status.HTTP_201_CREATED)
+def create_plan(
+    body: EditPlanCreateRequest,
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+    project_repository: Any = Depends(get_project_repository),
+) -> EditPlanResponse:
+    """创建剪辑计划"""
+    # 空串 project_id 统一为 ""
+    project_id = (body.project_id or "").strip()
+    # 项目鉴权
+    if project_id:
+        _check_project_access(project_id, current_user.user.id, project_repository)
+    svc = EditPlanService(db)
+    # 标准化 config，填充 cover/title/subtitle/bgm 默认值
+    normalized_config = normalize_plan_config(body.config)
+    try:
+        created = svc.create_plan(
+            template_id=body.template_id,
+            name=body.name,
+            config=normalized_config,
+            total_duration=body.total_duration,
+            project_id=project_id,
+            created_by_user_id=current_user.user.id,
+        )
+    except ValueError as exc:
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=str(exc),
+        )
+    logger.info(
+        "创建剪辑计划: id=%s name=%s by user=%s",
+        created.id,
+        created.name,
+        current_user.user.id,
+    )
+    return _to_response(created)
+
+
+@router.put("/{plan_id}", response_model=EditPlanResponse)
+def update_plan(
+    plan_id: str,
+    body: EditPlanUpdateRequest,
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+    project_repository: Any = Depends(get_project_repository),
+) -> EditPlanResponse:
+    """更新剪辑计划（支持状态机流转）"""
+    svc = EditPlanService(db)
+    # 项目鉴权
+    existing = svc.get_plan(plan_id)
+    if existing is None:
+        raise HTTPException(status_code=404, detail=f"剪辑计划不存在: {plan_id}")
+    if existing.project_id:
+        _check_project_access(existing.project_id, current_user.user.id, project_repository)
+
+    # 基础字段更新
+    try:
+        if body.name is not None or body.config is not None or body.total_duration is not None:
+            # 标准化 config（如果提供了）
+            config_to_update = normalize_plan_config(body.config) if body.config is not None else None
+            svc.update_plan(
+                plan_id,
+                name=body.name,
+                config=config_to_update,
+                total_duration=body.total_duration,
+            )
+
+        # 状态机流转
+        if body.status is not None:
+            try:
+                target_status = EditPlanStatus(body.status)
+            except ValueError:
+                raise HTTPException(
+                    status_code=status.HTTP_400_BAD_REQUEST,
+                    detail=(f"无效的状态值: {body.status}，" f"可选值: draft, editing, rendering, completed, failed"),
+                )
+            svc.transition_status(plan_id, target_status)
+    except ValueError as exc:
+        err_msg = str(exc)
+        if "不存在" in err_msg:
+            raise HTTPException(
+                status_code=status.HTTP_404_NOT_FOUND,
+                detail=err_msg,
+            )
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=err_msg,
+        )
+
+    # 返回最新状态
+    result = svc.get_plan_or_raise(plan_id)
+    logger.info("更新剪辑计划: id=%s by user=%s", plan_id, current_user.user.id)
+    return _to_response(result)
+
+
+@router.delete("/{plan_id}", status_code=status.HTTP_204_NO_CONTENT, response_model=None)
+def delete_plan(
+    plan_id: str,
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+    project_repository: Any = Depends(get_project_repository),
+) -> None:
+    """删除剪辑计划"""
+    svc = EditPlanService(db)
+    # 项目鉴权
+    existing = svc.get_plan(plan_id)
+    if existing and existing.project_id:
+        _check_project_access(existing.project_id, current_user.user.id, project_repository)
+    deleted = svc.delete_plan(plan_id)
+    if not deleted:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND,
+            detail=f"剪辑计划不存在: {plan_id}",
+        )
+    logger.info(
+        "删除剪辑计划: id=%s by user=%s",
+        plan_id,
+        current_user.user.id,
+    )
+
+
+# ── 生成相关端点（任务 2.05） ─────────────────────────────────────────────────
+
+
+@router.post("/{plan_id}/generate", response_model=EditPlanGenerateResponse)
+def generate_plan(
+    plan_id: str,
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+    project_repository: Any = Depends(get_project_repository),
+) -> EditPlanGenerateResponse:
+    """触发剪辑计划渲染生成
+
+    前置条件：计划状态必须为 editing，且至少有一个片段。
+    流程：
+    1. 验证计划状态为 editing
+    2. 将 pending 片段标记为 ready
+    3. 创建 GenerationTask
+    4. 调度 Celery 任务 worker.render_edit_plan
+    5. 将计划状态流转为 rendering
+    """
+    svc = EditPlanService(db)
+    # 项目鉴权
+    plan_check = svc.get_plan(plan_id)
+    if plan_check is None:
+        raise HTTPException(status_code=404, detail=f"剪辑计划不存在: {plan_id}")
+    if plan_check.project_id:
+        _check_project_access(plan_check.project_id, current_user.user.id, project_repository)
+
+    # 检查是否可生成
+    try:
+        can_gen, reason = svc.can_generate(plan_id)
+    except ValueError as exc:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND,
+            detail=str(exc),
+        )
+    if not can_gen:
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=reason,
+        )
+
+    # 将 pending 片段标记为 ready
+    clip_count = svc.mark_clips_ready(plan_id)
+
+    # 创建 GenerationTask
+    gen_task_repo = SQLAlchemyGenerationTaskRepository(db)
+    gen_task_use_case = CreateGenerationTaskUseCase(gen_task_repo)
+    plan = svc.get_plan_or_raise(plan_id)
+    gen_task = gen_task_use_case.execute(
+        CreateGenerationTaskCommand(
+            project_id="",
+            template_id=plan.template_id,
+            created_by_user_id=current_user.user.id,
+            source_edit_plan_id=plan_id,
+        )
+    )
+
+    # 将 generation_task_id 存入 plan config
+    svc.update_plan_config(plan_id, {"generation_task_id": gen_task.id})
+
+    # 流转状态为 rendering
+    svc.transition_status(plan_id, EditPlanStatus.RENDERING)
+
+    # 调度 Celery 任务
+    celery_app.send_task("worker.render_edit_plan", args=[plan_id])
+
+    # 获取最新状态
+    updated_plan = svc.get_plan_or_raise(plan_id)
+
+    logger.info(
+        "触发剪辑计划生成: plan_id=%s gen_task_id=%s clips=%d by user=%s",
+        plan_id,
+        gen_task.id,
+        clip_count,
+        current_user.user.id,
+    )
+
+    return EditPlanGenerateResponse(
+        plan_id=plan_id,
+        plan_status=updated_plan.status.value if hasattr(updated_plan.status, "value") else updated_plan.status,
+        generation_task_id=gen_task.id,
+        clip_count=clip_count,
+    )
+
+
+@router.get(
+    "/{plan_id}/generation-status",
+    response_model=EditPlanGenerationStatusResponse,
+)
+def get_generation_status(
+    plan_id: str,
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+    project_repository: Any = Depends(get_project_repository),
+) -> EditPlanGenerationStatusResponse:
+    """查询剪辑计划生成进度
+
+    返回计划状态、关联的 GenerationTask ID、以及每个片段的状态。
+    """
+    svc = EditPlanService(db)
+    try:
+        gen_status = svc.get_generation_status(plan_id)
+    except ValueError as exc:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND,
+            detail=str(exc),
+        )
+
+    plan = gen_status["plan"]
+    # 项目鉴权
+    if plan.project_id:
+        _check_project_access(plan.project_id, current_user.user.id, project_repository)
+    clips = gen_status["clips"]
+
+    clip_items = [
+        ClipStatusItem(
+            clip_id=c.id,
+            clip_type=c.clip_type,
+            order=c.order,
+            status=c.status.value if hasattr(c.status, "value") else c.status,
+            asset_id=c.asset_id or "",
+            text_content=c.text_content or "",
+            duration=c.duration,
+        )
+        for c in clips
+    ]
+
+    return EditPlanGenerationStatusResponse(
+        plan_id=plan_id,
+        plan_status=plan.status.value if hasattr(plan.status, "value") else plan.status,
+        generation_task_id=gen_status["generation_task_id"],
+        clips=clip_items,
+    )
+
+
+@router.get(
+    "/{plan_id}/generations",
+    response_model=EditPlanGenerationsResponse,
+)
+def list_plan_generations(
+    plan_id: str,
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+    project_repository: Any = Depends(get_project_repository),
+) -> EditPlanGenerationsResponse:
+    """查询剪辑计划关联的所有生成记录
+
+    返回该剪辑计划触发的所有 GenerationTask，按创建时间倒序。
+    """
+    svc = EditPlanService(db)
+    # 验证计划存在 + 项目鉴权
+    plan = svc.get_plan_or_raise(plan_id)
+    if plan.project_id:
+        _check_project_access(plan.project_id, current_user.user.id, project_repository)
+
+    gen_task_repo = SQLAlchemyGenerationTaskRepository(db)
+    tasks = gen_task_repo.list_by_source_edit_plan(plan_id)
+    items = [
+        GenerationTaskResponse(
+            id=t.id,
+            project_id=t.project_id,
+            asset_library_id=t.asset_library_id,
+            strategy_id=t.strategy_id,
+            voice_library_id=t.voice_library_id,
+            template_id=t.template_id,
+            asset_ids=t.asset_ids,
+            title_ids=t.title_ids,
+            voice_ids=t.voice_ids,
+            source_edit_plan_id=t.source_edit_plan_id or "",
+            status=t.status.value if hasattr(t.status, "value") else t.status,
+            progress=t.progress,
+            result_count=t.result_count,
+            error_message=t.error_message,
+        )
+        for t in tasks
+    ]
+    return EditPlanGenerationsResponse(items=items, total=len(items))
+
+
+# ── AI 推荐 & 封面生成端点（任务 3.09） ────────────────────────────────────────
+
+
+@router.post(
+    "/{plan_id}/ai-recommend",
+    response_model=AIRecommendResponse,
+)
+def ai_recommend_clips(
+    plan_id: str,
+    body: AIRecommendRequest,
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+    project_repository: Any = Depends(get_project_repository),
+) -> AIRecommendResponse:
+    """AI 推荐片段方案
+
+    调用 AI 服务分析素材，自动生成片段编排方案并写入剪辑计划。
+
+    流程：
+    1. 验证计划存在且状态为 draft/editing
+    2. 调用 AI 推荐服务（当前为 stub，后续接入真实 AI）
+    3. 清除计划现有片段，按推荐方案重新创建
+    4. 更新计划 config（cover/title/subtitle/bgm）和 total_duration
+    5. 返回推荐方案详情
+
+    前端对接：
+    - 请求体只需传 asset_ids（必填），editing_mode 和 target_duration 可选
+    - 返回的 clips 可直接渲染到时间线
+    - 返回的 config 包含推荐的封面/标题/字幕/BGM 配置
+    """
+    svc = EditPlanService(db)
+
+    # 验证计划存在
+    try:
+        plan = svc.get_plan_or_raise(plan_id)
+    except ValueError as exc:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND,
+            detail=str(exc),
+        )
+
+    # 项目鉴权
+    if plan.project_id:
+        _check_project_access(plan.project_id, current_user.user.id, project_repository)
+
+    # 验证状态：只允许 draft 或 editing
+    plan_status = plan.status.value if hasattr(plan.status, "value") else plan.status
+    if plan_status not in ("draft", "editing"):
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=f"AI 推荐仅支持 draft/editing 状态的计划，当前状态: {plan_status}",
+        )
+
+    # 调用 AI 推荐服务（同步调用 stub，后续改为 Celery 异步）
+    from apps.worker.worker_app.tasks.ai_tasks import run_ai_recommend
+
+    result = run_ai_recommend(
+        plan_id=plan_id,
+        template_id=plan.template_id,
+        asset_ids=body.asset_ids,
+        editing_mode=body.editing_mode,
+        target_duration=body.target_duration,
+    )
+
+    # ── 事务保护：清除 → 重建 → 更新 必须在同一逻辑事务中 ──
+    # TODO: 当前各 repo 方法内部 commit()，无法真正回滚。
+    #       后续重构 repo 为 flush() 模式后，此处改为统一 commit。
+    try:
+        # 清除现有片段
+        svc.delete_all_clips(plan_id)
+
+        # 按推荐方案创建新片段
+        for clip_data in result["clips"]:
+            svc.create_clip(
+                plan_id=plan_id,
+                clip_type=clip_data["clip_type"],
+                order=clip_data["order"],
+                text_content=clip_data.get("text_content", ""),
+                duration=clip_data["duration"],
+                transition_effect=clip_data.get("transition_effect", "cut"),
+                asset_id=clip_data.get("asset_id", ""),
+                start_time=clip_data.get("start_time", 0.0),
+                config=clip_data.get("config", {}),
+            )
+
+        # 更新计划 config 和 total_duration
+        normalized_config = normalize_plan_config(result.get("config", {}))
+        svc.update_plan(
+            plan_id,
+            config=normalized_config,
+            total_duration=result["total_duration"],
+        )
+    except Exception as exc:
+        logger.exception("AI 推荐写入失败，plan_id=%s 数据可能不一致", plan_id)
+        # 尝试回滚未提交的变更
+        try:
+            db.rollback()
+        except Exception:
+            pass
+        raise HTTPException(
+            status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,
+            detail=f"AI 推荐结果写入失败: {exc}",
+        )
+
+    logger.info(
+        "AI 推荐片段方案: plan_id=%s clips=%d duration=%.1f by user=%s",
+        plan_id,
+        len(result["clips"]),
+        result["total_duration"],
+        current_user.user.id,
+    )
+
+    return AIRecommendResponse(
+        plan_id=plan_id,
+        clips=[
+            AIRecommendClipItem(
+                clip_type=c["clip_type"],
+                order=c["order"],
+                text_content=c.get("text_content", ""),
+                duration=c["duration"],
+                transition_effect=c.get("transition_effect", "cut"),
+                asset_id=c.get("asset_id", ""),
+                start_time=c.get("start_time", 0.0),
+                config=c.get("config", {}),
+            )
+            for c in result["clips"]
+        ],
+        config=normalized_config,
+        total_duration=result["total_duration"],
+        confidence=result["confidence"],
+    )
+
+
+@router.post(
+    "/{plan_id}/generate-cover",
+    response_model=GenerateCoverResponse,
+)
+def generate_cover(
+    plan_id: str,
+    body: GenerateCoverRequest,
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+    project_repository: Any = Depends(get_project_repository),
+) -> GenerateCoverResponse:
+    """AI 生成封面
+
+    调用 AI 服务从视频中选帧或生成封面图，并更新计划 config.cover。
+
+    流程：
+    1. 验证计划存在
+    2. 调用 AI 封面生成服务（当前为 stub，后续接入真实 AI）
+    3. 更新 plan.config["cover"] 为生成的封面数据
+    4. 返回封面数据
+
+    前端对接：
+    - cover_type=ai_frame: AI 智能选帧（默认）
+    - cover_type=manual: 手动选帧，需传 frame_time
+    - cover_type=upload: 用户上传，接口返回空 image_url，前端自行上传后更新
+    - cover_type=ai_regenerate: AI 重新生成
+    """
+    svc = EditPlanService(db)
+
+    # 验证计划存在
+    try:
+        plan = svc.get_plan_or_raise(plan_id)
+    except ValueError as exc:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND,
+            detail=str(exc),
+        )
+
+    # 项目鉴权
+    if plan.project_id:
+        _check_project_access(plan.project_id, current_user.user.id, project_repository)
+
+    # 调用 AI 封面生成服务
+    from apps.worker.worker_app.tasks.ai_tasks import run_generate_cover
+
+    cover_data = run_generate_cover(
+        plan_id=plan_id,
+        asset_ids=body.asset_ids,
+        cover_type=body.cover_type,
+        frame_time=body.frame_time,
+    )
+
+    # 更新 plan.config["cover"]
+    current_config = dict(plan.config)
+    current_config["cover"] = cover_data
+    normalized = normalize_plan_config(current_config)
+    svc.update_plan_config(plan_id, {"cover": normalized["cover"]})
+
+    logger.info(
+        "AI 封面生成: plan_id=%s type=%s by user=%s",
+        plan_id,
+        body.cover_type,
+        current_user.user.id,
+    )
+
+    return GenerateCoverResponse(
+        plan_id=plan_id,
+        cover=cover_data,
+    )
+
+
+# ── Timeline / Scene 端点（P2-6） ─────────────────────────────────────────────
+
+
+class TimelineSceneResponse(BaseModel):
+    """时间线场景"""
+
+    scene: str = Field(..., description="场景描述")
+    time: str = Field(..., description='时间范围，如 "0:00 - 0:05"')
+    duration: float = Field(..., ge=0, description="时长（秒）")
+    color: str = Field(..., description="展示颜色")
+    clip_id: str = Field(default="", description="关联的片段 ID")
+    clip_type: str = Field(default="", description="片段类型")
+
+
+class TimelineResponse(BaseModel):
+    """时间线响应"""
+
+    plan_id: str
+    total_duration: float
+    scenes: List[TimelineSceneResponse]
+
+
+# clip_type → 颜色映射
+_CLIP_TYPE_COLORS = {
+    "intro": "#6366f1",
+    "title": "#6366f1",
+    "product": "#818cf8",
+    "showcase": "#10b981",
+    "scene": "#10b981",
+    "subtitle": "#f59e0b",
+    "text": "#f59e0b",
+    "cta": "#ef4444",
+    "outro": "#ef4444",
+    "voiceover": "#8b5cf6",
+    "transition": "#64748b",
+}
+
+_DEFAULT_COLOR = "#6366f1"
+
+
+def _format_time(seconds: float) -> str:
+    """将秒数格式化为 M:SS"""
+    m = int(seconds) // 60
+    s = int(seconds) % 60
+    return f"{m}:{s:02d}"
+
+
+def _clip_type_to_scene_label(clip_type: str, text_content: str) -> str:
+    """根据 clip_type 和 text_content 生成场景描述"""
+    type_labels = {
+        "intro": "开场",
+        "title": "标题",
+        "product": "产品展示",
+        "showcase": "场景展示",
+        "scene": "场景",
+        "subtitle": "字幕",
+        "text": "文字",
+        "cta": "结尾 CTA",
+        "outro": "结尾",
+        "voiceover": "配音",
+        "transition": "转场",
+    }
+    label = type_labels.get(clip_type, clip_type or "片段")
+    if text_content:
+        # 截取前 20 个字符作为副标题
+        short = text_content[:20].strip()
+        if short:
+            return f"{label} - {short}"
+    return label
+
+
+@router.get(
+    "/{plan_id}/timeline",
+    response_model=TimelineResponse,
+)
+def get_plan_timeline(
+    plan_id: str,
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+    project_repository: Any = Depends(get_project_repository),
+) -> TimelineResponse:
+    """获取剪辑计划的时间线场景数据
+
+    返回按计划片段排序的时间线场景列表，供前端 GeneratePage 渲染使用。
+    """
+    svc = EditPlanService(db)
+    plan = svc.get_plan_or_raise(plan_id)
+    # 项目鉴权
+    if plan.project_id:
+        _check_project_access(plan.project_id, current_user.user.id, project_repository)
+
+    clips = svc.list_clips(plan_id=plan_id, skip=0, limit=200)
+    # 按 order 排序
+    clips.sort(key=lambda c: c.order)
+
+    scenes: List[TimelineSceneResponse] = []
+    current_time = 0.0
+
+    for clip in clips:
+        start = current_time
+        end = start + clip.duration
+        color = _CLIP_TYPE_COLORS.get(clip.clip_type, _DEFAULT_COLOR)
+        scene_label = _clip_type_to_scene_label(clip.clip_type, clip.text_content)
+
+        scenes.append(
+            TimelineSceneResponse(
+                scene=scene_label,
+                time=f"{_format_time(start)} - {_format_time(end)}",
+                duration=clip.duration,
+                color=color,
+                clip_id=clip.id,
+                clip_type=clip.clip_type,
+            )
+        )
+        current_time = end
+
+    total_duration = sum(s.duration for s in scenes) or plan.total_duration
+
+    return TimelineResponse(
+        plan_id=plan_id,
+        total_duration=total_duration,
+        scenes=scenes,
+    )

apps/api/app/api/routes/edit_templates.py

@@ -0,0 +1,279 @@
+"""模板管理 API — Phase 8 模板编排引擎.
+
+RESTful CRUD for EditTemplate:
+- GET  /api/v1/edit-templates      列表（分页 + 类型筛选）
+- GET  /api/v1/edit-templates/{id} 详情
+- POST /api/v1/edit-templates      创建（管理员）
+- PUT  /api/v1/edit-templates/{id} 更新
+- DELETE /api/v1/edit-templates/{id} 删除（软删除 → inactive）
+
+业务逻辑委托给 EditTemplateService 服务层。
+"""
+
+from __future__ import annotations
+
+import logging
+from datetime import datetime
+from typing import Any, List, Optional
+
+from app.auth import AuthenticatedUser, get_current_user
+from app.dependencies import get_db_session
+from app.services import EditTemplateService
+from fastapi import APIRouter, Depends, HTTPException, Query, status
+from fastapi.responses import Response
+from pydantic import BaseModel, Field
+from sqlalchemy.orm import Session
+
+from packages.domain.config_schemas import normalize_template_config
+from packages.domain.edit_template import EditTemplate, EditTemplateStatus
+
+logger = logging.getLogger(__name__)
+
+router = APIRouter()
+
+
+# ── Pydantic Schemas ─────────────────────────────────────────────────────────
+
+
+class EditTemplateCreateRequest(BaseModel):
+    """创建模板请求体"""
+
+    name: str = Field(..., min_length=1, max_length=200, description="模板名称")
+    description: str = Field(default="", max_length=2000, description="模板描述")
+    template_type: str = Field(default="default", max_length=50, description="模板类型")
+    config: dict[str, Any] = Field(default_factory=dict, description="模板配置 (JSON)")
+    preview_url: str = Field(default="", max_length=500, description="预览地址")
+    sort_weight: int = Field(default=0, ge=0, le=9999, description="排序权重")
+
+
+class EditTemplateUpdateRequest(BaseModel):
+    """更新模板请求体"""
+
+    name: Optional[str] = Field(default=None, min_length=1, max_length=200, description="模板名称")
+    description: Optional[str] = Field(default=None, max_length=2000, description="模板描述")
+    template_type: Optional[str] = Field(default=None, max_length=50, description="模板类型")
+    config: Optional[dict[str, Any]] = Field(default=None, description="模板配置 (JSON)")
+    preview_url: Optional[str] = Field(default=None, max_length=500, description="预览地址")
+    sort_weight: Optional[int] = Field(default=None, ge=0, le=9999, description="排序权重")
+    status: Optional[str] = Field(default=None, description="状态: active / inactive")
+
+
+class EditTemplateResponse(BaseModel):
+    """模板响应体"""
+
+    id: str
+    name: str
+    description: str
+    template_type: str
+    config: dict[str, Any]
+    preview_url: str
+    sort_weight: int
+    status: str
+    created_at: datetime
+    updated_at: datetime
+
+    model_config = {"from_attributes": True}
+
+
+class EditTemplateListResponse(BaseModel):
+    """模板列表响应体"""
+
+    items: List[EditTemplateResponse]
+    total: int
+    page: int
+    page_size: int
+
+
+# ── Helpers ───────────────────────────────────────────────────────────────────
+
+
+def _require_admin(current_user: AuthenticatedUser) -> None:
+    """校验当前用户是否为管理员，非管理员返回 403"""
+    if not getattr(current_user.user, "is_admin", False):
+        raise HTTPException(
+            status_code=status.HTTP_403_FORBIDDEN,
+            detail="仅管理员可执行此操作",
+        )
+
+
+def _to_response(t: EditTemplate) -> EditTemplateResponse:
+    return EditTemplateResponse(
+        id=t.id,
+        name=t.name,
+        description=t.description,
+        template_type=t.template_type,
+        config=t.config,
+        preview_url=t.preview_url,
+        sort_weight=t.sort_weight,
+        status=t.status.value if hasattr(t.status, "value") else t.status,
+        created_at=t.created_at,
+        updated_at=t.updated_at,
+    )
+
+
+# ── Routes ────────────────────────────────────────────────────────────────────
+
+
+@router.get("", response_model=EditTemplateListResponse)
+def list_templates(
+    page: int = Query(default=1, ge=1, description="页码"),
+    page_size: int = Query(default=20, ge=1, le=100, description="每页数量"),
+    template_type: Optional[str] = Query(default=None, description="按类型筛选"),
+    status_filter: Optional[str] = Query(
+        default=None,
+        alias="status",
+        description="按状态筛选: active / inactive",
+    ),
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+) -> EditTemplateListResponse:
+    """获取模板列表（支持分页、按类型/状态筛选）"""
+    svc = EditTemplateService(db)
+
+    # 解析状态筛选
+    status_enum: Optional[EditTemplateStatus] = None
+    if status_filter:
+        try:
+            status_enum = EditTemplateStatus(status_filter)
+        except ValueError:
+            raise HTTPException(
+                status_code=status.HTTP_400_BAD_REQUEST,
+                detail=f"无效的状态值: {status_filter}，可选值: active, inactive",
+            )
+
+    skip = (page - 1) * page_size
+    templates = svc.list_templates(
+        template_type=template_type,
+        status=status_enum,
+        skip=skip,
+        limit=page_size,
+    )
+    total = svc.count_templates(
+        template_type=template_type,
+        status=status_enum,
+    )
+
+    return EditTemplateListResponse(
+        items=[_to_response(t) for t in templates],
+        total=total,
+        page=page,
+        page_size=page_size,
+    )
+
+
+@router.get("/{template_id}", response_model=EditTemplateResponse)
+def get_template(
+    template_id: str,
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+) -> EditTemplateResponse:
+    """获取单个模板详情"""
+    svc = EditTemplateService(db)
+    try:
+        template = svc.get_template_or_raise(template_id)
+    except ValueError as exc:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND,
+            detail=str(exc),
+        )
+    return _to_response(template)
+
+
+@router.post("", response_model=EditTemplateResponse, status_code=status.HTTP_201_CREATED)
+def create_template(
+    body: EditTemplateCreateRequest,
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+) -> EditTemplateResponse:
+    """创建模板（管理员）"""
+    _require_admin(current_user)
+    svc = EditTemplateService(db)
+    # 标准化 config，填充 cover/title/subtitle/bgm 默认值
+    normalized_config = normalize_template_config(body.config)
+    try:
+        created = svc.create_template(
+            name=body.name,
+            description=body.description,
+            template_type=body.template_type,
+            config=normalized_config,
+            preview_url=body.preview_url,
+            sort_weight=body.sort_weight,
+        )
+    except ValueError as exc:
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=str(exc),
+        )
+    logger.info("创建模板: id=%s name=%s by user=%s", created.id, created.name, current_user.user.id)
+    return _to_response(created)
+
+
+@router.put("/{template_id}", response_model=EditTemplateResponse)
+def update_template(
+    template_id: str,
+    body: EditTemplateUpdateRequest,
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+) -> EditTemplateResponse:
+    """更新模板"""
+    _require_admin(current_user)
+    svc = EditTemplateService(db)
+
+    # 解析状态
+    status_enum: Optional[EditTemplateStatus] = None
+    if body.status is not None:
+        try:
+            status_enum = EditTemplateStatus(body.status)
+        except ValueError:
+            raise HTTPException(
+                status_code=status.HTTP_400_BAD_REQUEST,
+                detail=f"无效的状态值: {body.status}，可选值: active, inactive",
+            )
+
+    # 标准化 config（如果提供了）
+    config_to_update = normalize_template_config(body.config) if body.config is not None else None
+
+    try:
+        result = svc.update_template(
+            template_id,
+            name=body.name,
+            description=body.description,
+            template_type=body.template_type,
+            config=config_to_update,
+            preview_url=body.preview_url,
+            sort_weight=body.sort_weight,
+            status=status_enum,
+        )
+    except ValueError as exc:
+        err_msg = str(exc)
+        if "不存在" in err_msg:
+            raise HTTPException(
+                status_code=status.HTTP_404_NOT_FOUND,
+                detail=err_msg,
+            )
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=err_msg,
+        )
+    logger.info("更新模板: id=%s by user=%s", template_id, current_user.user.id)
+    return _to_response(result)
+
+
+@router.delete("/{template_id}", status_code=status.HTTP_204_NO_CONTENT, response_model=None)
+def delete_template(
+    template_id: str,
+    db: Session = Depends(get_db_session),
+    current_user: AuthenticatedUser = Depends(get_current_user),
+) -> Response:
+    """删除模板（软删除 → 设为 inactive）"""
+    _require_admin(current_user)
+    svc = EditTemplateService(db)
+    try:
+        svc.deactivate_template(template_id)
+    except ValueError as exc:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND,
+            detail=str(exc),
+        )
+    logger.info("删除模板(软删除): id=%s by user=%s", template_id, current_user.user.id)
+    return Response(status_code=204)

apps/api/app/api/routes/generated_videos.py

@@ -50,7 +50,7 @@ def list_generated_videos(
 ) -> ListGeneratedVideosResponse:
     user_id = authenticated_user.user.id
     use_case = ListGeneratedVideosUseCase(generated_video_repository)
-    
+
     if project_id:
         # If project_id provided, check access and filter by project
         project = project_repository.find_by_id(project_id)
@@ -64,7 +64,7 @@ def list_generated_videos(
         for proj in accessible_projects:
             all_items.extend(use_case.execute(proj.id))
         items = all_items
-    
+
     # Generate download URLs for each video
     responses = []
     for item in items:

apps/api/app/api/routes/generation_tasks.py

@@ -1,3 +1,5 @@
+import random
+import uuid
 from typing import Any
 
 from app.auth import AuthenticatedUser, get_current_user
@@ -14,8 +16,11 @@ from app.schemas.generated_video import (
     ListGeneratedVideosResponse,
 )
 from app.schemas.generation_task import (
+    BatchGenerationTaskResponse,
+    ConfirmGenerationRequest,
     CreateGenerationTaskRequest,
     GenerationTaskResponse,
+    ListGenerationTasksResponse,
 )
 from fastapi import APIRouter, Depends, HTTPException
 
@@ -45,6 +50,19 @@ def _to_generation_task_response(task) -> GenerationTaskResponse:
         asset_library_id=task.asset_library_id,
         strategy_id=task.strategy_id,
         voice_library_id=task.voice_library_id,
+        template_id=task.template_id,
+        asset_ids=task.asset_ids,
+        title_ids=task.title_ids,
+        voice_ids=task.voice_ids,
+        source_edit_plan_id=task.source_edit_plan_id or "",
+        asset_select_mode=getattr(task, "asset_select_mode", ""),
+        batch_id=getattr(task, "batch_id", ""),
+        is_preview=getattr(task, "is_preview", True),
+        source_task_id=getattr(task, "source_task_id", ""),
+        output_width=getattr(task, "output_width", 1280),
+        output_height=getattr(task, "output_height", 720),
+        cover_url=getattr(task, "cover_url", ""),
+        custom_title=getattr(task, "custom_title", ""),
         status=task.status,
         progress=task.progress,
         result_count=task.result_count,
@@ -79,7 +97,87 @@ def _ensure_library_has_ready_video_assets(assets) -> None:
         )
 
 
-@router.post("/tasks", response_model=GenerationTaskResponse)
+def _select_assets_from_library(
+    assets: list,
+    mode: str,
+    count: int,
+) -> list[str]:
+    """根据选取模式从素材库中选取 ready 状态的视频素材 ID。
+
+    Args:
+        assets: 素材库中所有素材（Asset 实体列表）
+        mode: 选取模式 — all=全部, random=随机, smart=按质量评分
+        count: 选取数量，0 表示全部（仅 random/smart 模式有效）
+
+    Returns:
+        选中的素材 ID 列表
+    """
+    ready_video_assets = [a for a in assets if a.status.value == "ready" and a.mime_type.startswith("video")]
+
+    if not ready_video_assets:
+        return []
+
+    if mode == "random":
+        selected = (
+            ready_video_assets if count <= 0 else random.sample(ready_video_assets, min(count, len(ready_video_assets)))
+        )
+        return [a.id for a in selected]
+
+    if mode == "smart":
+        # 按质量分降序排列（质量分高的优先），质量分相同时按时长降序
+        sorted_assets = sorted(
+            ready_video_assets,
+            key=lambda a: (
+                a.quality_score if a.quality_score is not None else 0.0,
+                a.duration if a.duration is not None else 0.0,
+            ),
+            reverse=True,
+        )
+        selected = sorted_assets if count <= 0 else sorted_assets[:count]
+        return [a.id for a in selected]
+
+    # 默认 all 模式：返回全部 ready 视频素材
+    return [a.id for a in ready_video_assets]
+
+
+def _resolve_project_and_library(
+    request: CreateGenerationTaskRequest,
+    project_repository: Any,
+    asset_library_repository: Any,
+    asset_repository: Any,
+    authenticated_user: AuthenticatedUser,
+) -> tuple[str, str]:
+    """解析 project_id 和 asset_library_id。
+
+    支持两种模式：
+    - 显式传入（向后兼容）
+    - 从 asset_ids 反查 asset_library（模板模式）
+    返回 (project_id, asset_library_id)。
+    """
+    project_id = request.project_id.strip()
+    asset_library_id = request.asset_library_id.strip()
+
+    # 模板模式：project_id 未提供时，从 asset_ids 反查所属 project
+    if not project_id and request.asset_ids:
+        first_asset_id = request.asset_ids[0]
+        asset = asset_repository.find_by_id(first_asset_id)
+        if asset is not None:
+            project_id = asset.project_id
+            if not asset_library_id:
+                asset_library_id = asset.library_id
+
+    # 向后兼容校验：project_id 已提供时验证权限
+    if project_id:
+        project = project_repository.find_by_id(project_id)
+        if project is None:
+            raise HTTPException(status_code=404, detail=f"Project {project_id} not found")
+        if not project.can_access(authenticated_user.user.id):
+            raise HTTPException(status_code=403, detail="Access denied to project")
+
+    return project_id, asset_library_id
+
+
+@router.post("/tasks", response_model=BatchGenerationTaskResponse)
 def create_generation_task(
     request: CreateGenerationTaskRequest,
     authenticated_user: AuthenticatedUser = Depends(get_current_user),
@@ -87,32 +185,131 @@ def create_generation_task(
     project_repository: Any = Depends(get_project_repository),
     asset_library_repository: Any = Depends(get_asset_library_repository),
     asset_repository: Any = Depends(get_asset_repository),
-) -> GenerationTaskResponse:
-    project = project_repository.find_by_id(request.project_id)
-    if project is None:
-        raise HTTPException(status_code=404, detail=f"Project {request.project_id} not found")
-    if not project.can_access(authenticated_user.user.id):
-        raise HTTPException(status_code=403, detail="Access denied to project")
-    
-    library = asset_library_repository.get(request.asset_library_id)
-    if library is None or library.project_id != request.project_id:
-        raise HTTPException(status_code=404, detail=f"AssetLibrary {request.asset_library_id} not found")
-    
-    assets = asset_repository.list_by_library(request.asset_library_id)
-    _ensure_library_has_ready_video_assets(assets)
+) -> BatchGenerationTaskResponse:
+    project_id, asset_library_id = _resolve_project_and_library(
+        request, project_repository, asset_library_repository, asset_repository, authenticated_user
+    )
+
+    # asset_library 存在性校验（仅在提供了 asset_library_id 时）
+    resolved_asset_ids: list[str] = list(request.asset_ids)
+    if asset_library_id:
+        library = asset_library_repository.get(asset_library_id)
+        if library is None or (project_id and library.project_id != project_id):
+            raise HTTPException(status_code=404, detail=f"AssetLibrary {asset_library_id} not found")
+
+        assets = asset_repository.find_by_library(asset_library_id)
+        _ensure_library_has_ready_video_assets(assets)
+
+        # 素材库自动匹配：当未显式指定 asset_ids 时，按模式自动选取
+        if not resolved_asset_ids:
+            resolved_asset_ids = _select_assets_from_library(
+                assets,
+                mode=request.asset_select_mode,
+                count=request.asset_select_count,
+            )
 
     use_case = CreateGenerationTaskUseCase(generation_task_repository)
-    task = use_case.execute(
+    count = request.count
+    created_tasks = []
+    # 同批次任务共享 batch_id，用于视频查重时批次内比对
+    batch_id = uuid.uuid4().hex if count > 1 else ""
+
+    for _ in range(count):
+        task = use_case.execute(
+            CreateGenerationTaskCommand(
+                project_id=project_id,
+                asset_library_id=asset_library_id,
+                strategy_id=request.strategy_id,
+                voice_library_id=request.voice_library_id,
+                template_id=request.template_id,
+                asset_ids=resolved_asset_ids,
+                title_ids=request.title_ids,
+                voice_ids=request.voice_ids,
+                created_by_user_id=authenticated_user.user.id,
+                source_edit_plan_id=request.source_edit_plan_id,
+                asset_select_mode=request.asset_select_mode,
+                batch_id=batch_id,
+                is_preview=request.is_preview,
+                source_task_id=request.source_task_id,
+                output_width=request.output_width,
+                output_height=request.output_height,
+                cover_url=request.cover_url,
+                custom_title=request.custom_title,
+            )
+        )
+        celery_app.send_task("worker.generate_video", args=[task.id])
+        created_tasks.append(task)
+
+    items = [_to_generation_task_response(t) for t in created_tasks]
+    return BatchGenerationTaskResponse(items=items, total=len(items))
+
+
+@router.post("/tasks/{task_id}/confirm", response_model=BatchGenerationTaskResponse)
+def confirm_generation(
+    task_id: str,
+    request: ConfirmGenerationRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    generation_task_repository: Any = Depends(get_generation_task_repository),
+    project_repository: Any = Depends(get_project_repository),
+) -> BatchGenerationTaskResponse:
+    """确认生成 — 基于预览任务创建正式生成任务。
+
+    查找预览任务，复制其配置，创建新的正式生成任务（is_preview=False），
+    使用高分辨率，复用 worker.generate_video 渲染路径。
+    """
+    # 1. 查找源预览任务
+    source_task = generation_task_repository.get(task_id)
+    if source_task is None:
+        raise HTTPException(status_code=404, detail=f"Preview task {task_id} not found")
+
+    # 2. 权限检查
+    if source_task.created_by_user_id and source_task.created_by_user_id != authenticated_user.user.id:
+        raise HTTPException(status_code=403, detail="Access denied to this task")
+    if source_task.project_id:
+        _check_project_access(source_task.project_id, authenticated_user.user.id, project_repository)
+
+    # 3. 创建正式生成任务，复制预览任务的配置
+    use_case = CreateGenerationTaskUseCase(generation_task_repository)
+    new_task = use_case.execute(
         CreateGenerationTaskCommand(
-            project_id=request.project_id,
-            asset_library_id=request.asset_library_id,
-            strategy_id=request.strategy_id,
-            voice_library_id=request.voice_library_id,
+            project_id=source_task.project_id,
+            asset_library_id=source_task.asset_library_id,
+            strategy_id=source_task.strategy_id,
+            voice_library_id=source_task.voice_library_id,
+            template_id=source_task.template_id,
+            asset_ids=source_task.asset_ids,
+            title_ids=source_task.title_ids,
+            voice_ids=source_task.voice_ids,
             created_by_user_id=authenticated_user.user.id,
+            source_edit_plan_id=source_task.source_edit_plan_id or "",
+            asset_select_mode=source_task.asset_select_mode,
+            is_preview=False,
+            source_task_id=task_id,
+            output_width=request.output_width,
+            output_height=request.output_height,
+            cover_url=request.cover_url,
+            custom_title=request.custom_title,
         )
     )
-    celery_app.send_task("worker.generate_video", args=[task.id])
-    return _to_generation_task_response(task)
+
+    # 4. 调度 worker.generate_video（同一条渲染路径）
+    celery_app.send_task("worker.generate_video", args=[new_task.id])
+
+    return BatchGenerationTaskResponse(
+        items=[_to_generation_task_response(new_task)],
+        total=1,
+    )
+
+
+@router.get("/tasks", response_model=ListGenerationTasksResponse)
+def list_generation_tasks(
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    generation_task_repository: Any = Depends(get_generation_task_repository),
+) -> ListGenerationTasksResponse:
+    """用户级生成任务列表（跨 project）。"""
+    tasks = generation_task_repository.list_by_user(authenticated_user.user.id)
+    items = [_to_generation_task_response(task) for task in tasks]
+    return ListGenerationTasksResponse(items=items)
 
 
 @router.get("/tasks/{task_id}", response_model=GenerationTaskResponse)
@@ -126,7 +323,8 @@ def get_generation_task(
     task = use_case.execute(task_id)
     if task is None:
         raise HTTPException(status_code=404, detail=f"GenerationTask {task_id} not found")
-    _check_project_access(task.project_id, authenticated_user.user.id, project_repository)
+    if task.project_id:
+        _check_project_access(task.project_id, authenticated_user.user.id, project_repository)
     return _to_generation_task_response(task)
 
 
@@ -141,7 +339,50 @@ def list_generation_results(
     task = generation_task_repository.get(task_id)
     if task is None:
         raise HTTPException(status_code=404, detail=f"GenerationTask {task_id} not found")
-    _check_project_access(task.project_id, authenticated_user.user.id, project_repository)
+    if task.project_id:
+        _check_project_access(task.project_id, authenticated_user.user.id, project_repository)
     use_case = ListGeneratedVideosByTaskUseCase(generated_video_repository)
     items = use_case.execute(task_id)
     return ListGeneratedVideosResponse(items=[_to_generated_video_response(item) for item in items])
+
+
+@router.post("/tasks/{task_id}/retry", response_model=GenerationTaskResponse)
+def retry_generation_task(
+    task_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    generation_task_repository: Any = Depends(get_generation_task_repository),
+) -> GenerationTaskResponse:
+    """简化重试：通过 task_id 直接重试失败任务。"""
+    task = generation_task_repository.get(task_id)
+    if task is None:
+        raise HTTPException(status_code=404, detail="Generation task not found")
+    if task.created_by_user_id and task.created_by_user_id != authenticated_user.user.id:
+        raise HTTPException(status_code=403, detail="Access denied to this task")
+    status_val = task.status.value if hasattr(task.status, "value") else str(task.status)
+    if status_val != "failed":
+        raise HTTPException(status_code=409, detail="Only failed tasks can be retried")
+
+    use_case = CreateGenerationTaskUseCase(generation_task_repository)
+    retried = use_case.execute(
+        CreateGenerationTaskCommand(
+            project_id=task.project_id,
+            asset_library_id=task.asset_library_id,
+            strategy_id=task.strategy_id,
+            voice_library_id=task.voice_library_id,
+            template_id=task.template_id,
+            asset_ids=task.asset_ids,
+            title_ids=task.title_ids,
+            voice_ids=task.voice_ids,
+            created_by_user_id=authenticated_user.user.id,
+            source_edit_plan_id=task.source_edit_plan_id or "",
+            asset_select_mode=getattr(task, "asset_select_mode", ""),
+            is_preview=getattr(task, "is_preview", True),
+            source_task_id=getattr(task, "source_task_id", ""),
+            output_width=getattr(task, "output_width", 1280),
+            output_height=getattr(task, "output_height", 720),
+            cover_url=getattr(task, "cover_url", ""),
+            custom_title=getattr(task, "custom_title", ""),
+        )
+    )
+    celery_app.send_task("worker.generate_video", args=[retried.id])
+    return _to_generation_task_response(retried)

apps/api/app/api/routes/health.py

@@ -5,7 +5,6 @@ import redis
 from app.config import settings
 from fastapi import APIRouter, status
 from fastapi.responses import JSONResponse
-from pydantic import BaseModel
 
 router = APIRouter(tags=["Health"])
 
@@ -21,19 +20,8 @@ async def health_check():
 
 @router.get("/ready", status_code=status.HTTP_200_OK)
 async def readiness_check():
-    checks = {
-        "database": await _check_database(),
-        "redis": await _check_redis(),
-    }
-    all_healthy = all(check["status"] == "healthy" for check in checks.values())
-    response = {
-        "status": "ready" if all_healthy else "not_ready",
-        "timestamp": datetime.utcnow().isoformat(),
-        "checks": checks,
-    }
-    if not all_healthy:
-        return JSONResponse(status_code=status.HTTP_503_SERVICE_UNAVAILABLE, content=response)
-    return response
+    """简单的就绪检查，仅返回状态。详细健康检查请使用 /health 端点。"""
+    return {"status": "ready"}
 
 
 @router.get("/startup", status_code=status.HTTP_200_OK)
@@ -97,6 +85,38 @@ async def _check_redis() -> dict:
         }
 
 
+def _check_oss() -> dict:
+    try:
+        from app.core.storage import get_storage_service
+
+        svc = get_storage_service()
+        if not svc.access_key_id or not svc.access_key_secret:
+            return {
+                "status": "unhealthy",
+                "type": "oss",
+                "message": "OSS credentials not configured (OSS_ACCESS_KEY_ID / OSS_ACCESS_KEY_SECRET missing)",
+            }
+        if svc.bucket is None:
+            return {
+                "status": "unhealthy",
+                "type": "oss",
+                "message": "OSS SDK (oss2) not installed or bucket client init failed",
+            }
+        # Try a lightweight OSS API call to verify connectivity & credentials
+        svc.bucket.get_bucket_info()
+        return {
+            "status": "healthy",
+            "type": "oss",
+            "message": f"OSS connected: endpoint={svc.endpoint} bucket={svc.bucket_name}",
+        }
+    except Exception as error:
+        return {
+            "status": "unhealthy",
+            "type": "oss",
+            "message": f"OSS check failed: {type(error).__name__}: {error}",
+        }
+
+
 async def _check_migrations() -> dict:
     if settings.USE_IN_MEMORY_DB:
         return {

apps/api/app/api/routes/jobs.py

@@ -0,0 +1,332 @@
+"""Job API 路由 — Phase 8 任务 2.10.
+
+提供统一异步任务管理 RESTful 接口：
+- POST   /api/v1/jobs                        创建任务
+- GET    /api/v1/jobs/{job_id}                任务详情
+- GET    /api/v1/projects/{project_id}/jobs   项目任务列表
+- GET    /api/v1/projects/{project_id}/jobs/stats  任务统计
+- PUT    /api/v1/jobs/{job_id}/progress       更新进度
+- POST   /api/v1/jobs/{job_id}/complete       标记完成
+- POST   /api/v1/jobs/{job_id}/fail           标记失败
+- POST   /api/v1/jobs/{job_id}/retry          重试任务
+- POST   /api/v1/jobs/{job_id}/cancel         取消任务
+- POST   /api/v1/jobs/{job_id}/submit         提交执行
+"""
+
+from __future__ import annotations
+
+import logging
+from typing import Any
+
+from app.auth import AuthenticatedUser, get_current_user
+from app.core.celery_app import celery_app
+from app.dependencies import get_db_session, get_job_repository, get_project_repository
+from app.schemas.job import (
+    CompleteJobRequest,
+    CreateJobRequest,
+    FailJobRequest,
+    JobResponse,
+    JobStatisticsResponse,
+    ListJobsResponse,
+    UpdateProgressRequest,
+    job_to_response,
+)
+from fastapi import APIRouter, Depends, HTTPException, Query, status
+
+from packages.application.jobs import (
+    CancelJobUseCase,
+    CompleteJobCommand,
+    CompleteJobUseCase,
+    CreateJobCommand,
+    CreateJobUseCase,
+    FailJobCommand,
+    FailJobUseCase,
+    GetJobStatisticsUseCase,
+    GetJobUseCase,
+    ListJobsUseCase,
+    RetryJobUseCase,
+    SubmitJobUseCase,
+    UpdateJobProgressCommand,
+    UpdateJobProgressUseCase,
+)
+from packages.domain.job import JobType
+
+logger = logging.getLogger(__name__)
+
+router = APIRouter()
+
+# 任务类型 → Celery task name 映射
+_JOB_TYPE_TO_CELERY_TASK: dict[str, str] = {
+    JobType.VIDEO_COMPOSE: "worker.compose_video",
+    JobType.RENDER_EDIT_PLAN: "worker.render_edit_plan",
+    JobType.ASSET_INGEST: "worker.ingest_asset",
+    JobType.CLASSIFICATION: "worker.classify_asset",
+    JobType.VOICE_EXTRACTION: "worker.extract_voice",
+    JobType.GENERATION: "worker.generate_video",
+}
+
+
+def _check_project_access(project_id: str, user_id: str, project_repository) -> None:
+    """检查用户是否有项目访问权限。"""
+    project = project_repository.find_by_id(project_id)
+    if project is None:
+        raise HTTPException(status_code=404, detail=f"Project {project_id} not found")
+    if not project.can_access(user_id):
+        raise HTTPException(status_code=403, detail="Access denied to project")
+
+
+# ── 创建任务 ──────────────────────────────────────────────────────────────────
+
+
+@router.post("/jobs", response_model=JobResponse, status_code=status.HTTP_201_CREATED)
+def create_job(
+    request: CreateJobRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    job_repo: Any = Depends(get_job_repository),
+    project_repository: Any = Depends(get_project_repository),
+) -> JobResponse:
+    """创建异步任务。
+
+    创建后任务处于 pending 状态，需要调用 /submit 提交执行。
+    """
+    _check_project_access(request.project_id, authenticated_user.user.id, project_repository)
+
+    # 校验 job_type
+    try:
+        JobType(request.job_type)
+    except ValueError:
+        raise HTTPException(
+            status_code=400,
+            detail=f"不支持的任务类型: {request.job_type}，" f"可选值: {[t.value for t in JobType]}",
+        )
+
+    use_case = CreateJobUseCase(job_repo)
+    job = use_case.execute(
+        CreateJobCommand(
+            project_id=request.project_id,
+            job_type=request.job_type,
+            payload=request.payload,
+            source_id=request.source_id,
+            created_by_user_id=authenticated_user.user.id,
+            max_retries=request.max_retries,
+        )
+    )
+
+    return job_to_response(job)
+
+
+# ── 提交执行 ──────────────────────────────────────────────────────────────────
+
+
+@router.post("/jobs/{job_id}/submit", response_model=JobResponse)
+def submit_job(
+    job_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    job_repo: Any = Depends(get_job_repository),
+) -> JobResponse:
+    """提交任务执行。
+
+    将任务状态从 pending 切换为 running，并 dispatch Celery 异步任务。
+    """
+    # 权限检查：先获取任务并验证权限，再执行状态变更
+    job = job_repo.get(job_id)
+    if job is None:
+        raise HTTPException(status_code=404, detail=f"Job {job_id} not found")
+    if job.created_by_user_id and job.created_by_user_id != authenticated_user.user.id:
+        raise HTTPException(status_code=403, detail="Access denied to this job")
+
+    use_case = SubmitJobUseCase(job_repo)
+
+    try:
+        job = use_case.execute(job_id)
+    except ValueError as e:
+        raise HTTPException(status_code=400, detail=str(e))
+
+    # Dispatch Celery 任务
+    celery_task_name = _JOB_TYPE_TO_CELERY_TASK.get(job.job_type.value)
+    if celery_task_name:
+        result = celery_app.send_task(celery_task_name, args=[job.id], kwargs=job.payload)
+        job.celery_task_id = result.id
+        job_repo.update(job)
+        logger.info("已提交 Celery 任务: job_id=%s celery_task_id=%s", job.id, result.id)
+
+    return job_to_response(job)
+
+
+# ── 查询接口 ──────────────────────────────────────────────────────────────────
+
+
+@router.get("/jobs/{job_id}", response_model=JobResponse)
+def get_job(
+    job_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    job_repo: Any = Depends(get_job_repository),
+) -> JobResponse:
+    """获取任务详情。"""
+    use_case = GetJobUseCase(job_repo)
+    job = use_case.execute(job_id)
+    if job is None:
+        raise HTTPException(status_code=404, detail=f"Job {job_id} not found")
+    return job_to_response(job)
+
+
+@router.get("/projects/{project_id}/jobs", response_model=ListJobsResponse)
+def list_project_jobs(
+    project_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    job_repo: Any = Depends(get_job_repository),
+    project_repository: Any = Depends(get_project_repository),
+    job_type: str | None = Query(default=None, description="按任务类型过滤"),
+    status_filter: str | None = Query(default=None, alias="status", description="按状态过滤"),
+    limit: int = Query(default=50, ge=1, le=200),
+    offset: int = Query(default=0, ge=0),
+) -> ListJobsResponse:
+    """获取项目下的任务列表。"""
+    _check_project_access(project_id, authenticated_user.user.id, project_repository)
+
+    use_case = ListJobsUseCase(job_repo)
+    jobs = use_case.execute(
+        project_id=project_id,
+        job_type=job_type,
+        status=status_filter,
+        limit=limit,
+        offset=offset,
+    )
+    items = [job_to_response(j) for j in jobs]
+    return ListJobsResponse(items=items, total=len(items))
+
+
+@router.get("/projects/{project_id}/jobs/stats", response_model=JobStatisticsResponse)
+def get_job_statistics(
+    project_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    job_repo: Any = Depends(get_job_repository),
+    project_repository: Any = Depends(get_project_repository),
+) -> JobStatisticsResponse:
+    """获取项目任务统计摘要。"""
+    _check_project_access(project_id, authenticated_user.user.id, project_repository)
+
+    use_case = GetJobStatisticsUseCase(job_repo)
+    stats = use_case.execute(project_id)
+    return JobStatisticsResponse(**stats)
+
+
+# ── 进度更新 ──────────────────────────────────────────────────────────────────
+
+
+@router.put("/jobs/{job_id}/progress", response_model=JobResponse)
+def update_job_progress(
+    job_id: str,
+    request: UpdateProgressRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    job_repo: Any = Depends(get_job_repository),
+) -> JobResponse:
+    """更新任务进度。"""
+    use_case = UpdateJobProgressUseCase(job_repo)
+
+    try:
+        job = use_case.execute(
+            UpdateJobProgressCommand(
+                job_id=job_id,
+                progress=request.progress,
+                current_stage=request.current_stage,
+            )
+        )
+    except ValueError as e:
+        raise HTTPException(status_code=400, detail=str(e))
+
+    return job_to_response(job)
+
+
+# ── 完成 / 失败 ────────────────────────────────────────────────────────────────
+
+
+@router.post("/jobs/{job_id}/complete", response_model=JobResponse)
+def complete_job(
+    job_id: str,
+    request: CompleteJobRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    job_repo: Any = Depends(get_job_repository),
+) -> JobResponse:
+    """标记任务完成。"""
+    use_case = CompleteJobUseCase(job_repo)
+
+    try:
+        job = use_case.execute(CompleteJobCommand(job_id=job_id, result=request.result))
+    except ValueError as e:
+        raise HTTPException(status_code=400, detail=str(e))
+
+    return job_to_response(job)
+
+
+@router.post("/jobs/{job_id}/fail", response_model=JobResponse)
+def fail_job(
+    job_id: str,
+    request: FailJobRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    job_repo: Any = Depends(get_job_repository),
+) -> JobResponse:
+    """标记任务失败。"""
+    use_case = FailJobUseCase(job_repo)
+
+    try:
+        job = use_case.execute(FailJobCommand(job_id=job_id, error_message=request.error_message))
+    except ValueError as e:
+        raise HTTPException(status_code=400, detail=str(e))
+
+    return job_to_response(job)
+
+
+# ── 重试 / 取消 ────────────────────────────────────────────────────────────────
+
+
+@router.post("/jobs/{job_id}/retry", response_model=JobResponse)
+def retry_job(
+    job_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    job_repo: Any = Depends(get_job_repository),
+) -> JobResponse:
+    """重试失败任务。
+
+    将任务重置为 pending，retry_count + 1，但不自动 dispatch。
+    需要再次调用 /submit 提交执行。
+    """
+    # 权限检查：先获取任务并验证权限，再执行状态变更
+    job = job_repo.get(job_id)
+    if job is None:
+        raise HTTPException(status_code=404, detail=f"Job {job_id} not found")
+    if job.created_by_user_id and job.created_by_user_id != authenticated_user.user.id:
+        raise HTTPException(status_code=403, detail="Access denied to this job")
+
+    use_case = RetryJobUseCase(job_repo)
+
+    try:
+        job = use_case.execute(job_id)
+    except ValueError as e:
+        raise HTTPException(status_code=400, detail=str(e))
+
+    return job_to_response(job)
+
+
+@router.post("/jobs/{job_id}/cancel", response_model=JobResponse)
+def cancel_job(
+    job_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    job_repo: Any = Depends(get_job_repository),
+) -> JobResponse:
+    """取消任务。"""
+    # 权限检查：先获取任务并验证权限，再执行状态变更
+    job = job_repo.get(job_id)
+    if job is None:
+        raise HTTPException(status_code=404, detail=f"Job {job_id} not found")
+    if job.created_by_user_id and job.created_by_user_id != authenticated_user.user.id:
+        raise HTTPException(status_code=403, detail="Access denied to this job")
+
+    use_case = CancelJobUseCase(job_repo)
+
+    try:
+        job = use_case.execute(job_id)
+    except ValueError as e:
+        raise HTTPException(status_code=400, detail=str(e))
+
+    return job_to_response(job)

apps/api/app/api/routes/permissions.py

@@ -1,2 +0,0 @@
-# Compatibility module - workspace concept has been removed.
-# All permission checks are handled at the project level (see packages.domain.permissions).

apps/api/app/api/routes/projects.py

@@ -12,6 +12,7 @@ from fastapi import APIRouter, Depends, HTTPException, status
 from packages.application import (
     CreateProjectCommand,
     CreateProjectUseCase,
+    DeleteProjectUseCase,
     GetProjectUseCase,
     ListProjectsUseCase,
 )
@@ -22,8 +23,10 @@ router = APIRouter()
 def _to_project_response(item) -> ProjectResponse:
     return ProjectResponse(
         id=item.id,
+        owner_user_id=item.owner_user_id,
         name=item.name,
         description=item.description,
+        shared_users=item.shared_users,
     )
 
 
@@ -37,6 +40,8 @@ def get_project(
     project = use_case.execute(project_id)
     if project is None:
         raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Project not found")
+    if not project.can_access(authenticated_user.user.id):
+        raise HTTPException(status_code=status.HTTP_403_FORBIDDEN, detail="Access denied to project")
     return _to_project_response(project)
 
 
@@ -65,3 +70,22 @@ def create_project(
         owner_user_id=authenticated_user.user.id,
     )
     return _to_project_response(project)
+
+
+@router.delete("/{project_id}")
+def delete_project(
+    project_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    project_repository: Any = Depends(get_project_repository),
+):
+    use_case = DeleteProjectUseCase(project_repository)
+    try:
+        deleted = use_case.execute(project_id, authenticated_user.user.id)
+    except PermissionError:
+        raise HTTPException(
+            status_code=status.HTTP_403_FORBIDDEN,
+            detail="Only the project owner can delete this project",
+        )
+    if not deleted:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Project not found")
+    return {"message": "Project deleted successfully"}

apps/api/app/api/routes/recipes.py

@@ -0,0 +1,212 @@
+"""Recipe CRUD + use routes."""
+
+from __future__ import annotations
+
+from app.auth import AuthenticatedUser, get_current_user
+from app.dependencies import get_db_session, get_user_repository
+from app.schemas.recipe import (
+    CreateRecipeRequest,
+    ListRecipesResponse,
+    RecipeItemResponse,
+    RecipeResponse,
+    UpdateRecipeRequest,
+    UseRecipeResponse,
+)
+from fastapi import APIRouter, Depends, HTTPException, Query, Response, status
+from sqlalchemy.orm import Session
+
+from packages.adapters.sqlalchemy_impl.recipe_repository import SQLAlchemyRecipeRepository
+from packages.application.recipe.commands import (
+    CreateRecipeCommand,
+    RecipeItemCommand,
+    UpdateRecipeCommand,
+)
+from packages.application.recipe.use_cases import (
+    CreateRecipeUseCase,
+    DeleteRecipeUseCase,
+    FeatureDisabledError,
+    GetRecipeUseCase,
+    ListRecipesUseCase,
+    NotFoundError,
+    UpdateRecipeUseCase,
+    UseRecipeUseCase,
+)
+from packages.ports.user_repository import UserRepository
+
+router = APIRouter()
+
+
+def _get_recipe_repository(session: Session = Depends(get_db_session)) -> SQLAlchemyRecipeRepository:
+    return SQLAlchemyRecipeRepository(session)
+
+
+def _get_user_plan(user_id: str, user_repository: UserRepository) -> str:
+    user = user_repository.find_by_id(user_id)
+    if user is None:
+        return "free"
+    return getattr(user, "subscription_plan", "free") or "free"
+
+
+def _item_to_response(item) -> RecipeItemResponse:
+    return RecipeItemResponse(
+        id=item.id,
+        recipe_id=item.recipe_id,
+        item_type=item.item_type,
+        item_id=item.item_id,
+        position=item.position,
+        metadata=item.metadata_,
+    )
+
+
+def _to_response(recipe) -> RecipeResponse:
+    return RecipeResponse(
+        id=recipe.id,
+        user_id=recipe.user_id,
+        name=recipe.name,
+        description=recipe.description,
+        template_id=recipe.template_id,
+        generation_params=recipe.generation_params,
+        items=[_item_to_response(i) for i in getattr(recipe, "items", [])],
+        is_active=recipe.is_active,
+        metadata=recipe.metadata_,
+        created_at=recipe.created_at,
+        updated_at=recipe.updated_at,
+    )
+
+
+@router.get("", response_model=ListRecipesResponse)
+def list_recipes(
+    skip: int = Query(0, ge=0),
+    limit: int = Query(50, ge=1, le=200),
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    recipe_repository: SQLAlchemyRecipeRepository = Depends(_get_recipe_repository),
+) -> ListRecipesResponse:
+    user_id = authenticated_user.user.id
+    use_case = ListRecipesUseCase(recipe_repository)
+    recipes = use_case.execute(user_id, skip=skip, limit=limit)
+    total = recipe_repository.count_by_user(user_id)
+    return ListRecipesResponse(
+        items=[_to_response(r) for r in recipes],
+        total=total,
+    )
+
+
+@router.get("/{recipe_id}", response_model=RecipeResponse)
+def get_recipe(
+    recipe_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    recipe_repository: SQLAlchemyRecipeRepository = Depends(_get_recipe_repository),
+) -> RecipeResponse:
+    user_id = authenticated_user.user.id
+    use_case = GetRecipeUseCase(recipe_repository)
+    recipe = use_case.execute(recipe_id, user_id)
+    if recipe is None:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Recipe not found")
+    return _to_response(recipe)
+
+
+@router.post("", response_model=RecipeResponse, status_code=status.HTTP_201_CREATED)
+def create_recipe(
+    request: CreateRecipeRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    recipe_repository: SQLAlchemyRecipeRepository = Depends(_get_recipe_repository),
+) -> RecipeResponse:
+    user_id = authenticated_user.user.id
+    command = CreateRecipeCommand(
+        user_id=user_id,
+        name=request.name,
+        description=request.description,
+        template_id=request.template_id,
+        generation_params=request.generation_params,
+        items=[
+            RecipeItemCommand(
+                item_type=ic.item_type,
+                item_id=ic.item_id,
+                position=ic.position,
+                metadata_=ic.metadata_,
+            )
+            for ic in request.items
+        ],
+        metadata_=request.metadata_,
+    )
+    use_case = CreateRecipeUseCase(recipe_repository)
+    recipe = use_case.execute(command)
+    return _to_response(recipe)
+
+
+@router.patch("/{recipe_id}", response_model=RecipeResponse)
+def update_recipe(
+    recipe_id: str,
+    request: UpdateRecipeRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    recipe_repository: SQLAlchemyRecipeRepository = Depends(_get_recipe_repository),
+) -> RecipeResponse:
+    user_id = authenticated_user.user.id
+    command = UpdateRecipeCommand(
+        recipe_id=recipe_id,
+        user_id=user_id,
+        name=request.name,
+        description=request.description,
+        template_id=request.template_id,
+        generation_params=request.generation_params,
+        items=(
+            [
+                RecipeItemCommand(
+                    item_type=ic.item_type,
+                    item_id=ic.item_id,
+                    position=ic.position,
+                    metadata_=ic.metadata_,
+                )
+                for ic in request.items
+            ]
+            if request.items is not None
+            else None
+        ),
+        metadata_=request.metadata_,
+    )
+    use_case = UpdateRecipeUseCase(recipe_repository)
+    try:
+        recipe = use_case.execute(command)
+    except NotFoundError:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Recipe not found")
+    return _to_response(recipe)
+
+
+@router.delete("/{recipe_id}", status_code=status.HTTP_204_NO_CONTENT, response_model=None)
+def delete_recipe(
+    recipe_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    recipe_repository: SQLAlchemyRecipeRepository = Depends(_get_recipe_repository),
+) -> Response:
+    user_id = authenticated_user.user.id
+    use_case = DeleteRecipeUseCase(recipe_repository)
+    deleted = use_case.execute(recipe_id, user_id)
+    if not deleted:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Recipe not found")
+    return Response(status_code=204)
+
+
+@router.post("/{recipe_id}/use", response_model=UseRecipeResponse)
+def use_recipe(
+    recipe_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    recipe_repository: SQLAlchemyRecipeRepository = Depends(_get_recipe_repository),
+    user_repository: UserRepository = Depends(get_user_repository),
+) -> UseRecipeResponse:
+    user_id = authenticated_user.user.id
+    plan_name = _get_user_plan(user_id, user_repository)
+    use_case = UseRecipeUseCase(recipe_repository)
+    try:
+        result = use_case.execute(recipe_id, user_id, user_plan=plan_name)
+    except FeatureDisabledError as exc:
+        raise HTTPException(
+            status_code=status.HTTP_403_FORBIDDEN,
+            detail=str(exc),
+        )
+    except NotFoundError:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Recipe not found")
+
+    return UseRecipeResponse(
+        recipe=_to_response(result.recipe),
+        warnings=[{"item_type": w.item_type, "item_id": w.item_id, "position": w.position} for w in result.warnings],
+    )

apps/api/app/api/routes/subscription.py

@@ -0,0 +1,274 @@
+"""Subscription management API routes."""
+
+from __future__ import annotations
+
+from dataclasses import replace
+from datetime import datetime, timezone
+from typing import List
+
+from app.auth import AuthenticatedUser, get_current_user
+from app.dependencies import get_user_repository
+from app.schemas.subscription import (
+    BillingRecord,
+    ChangePlanRequest,
+    ChangePlanResponse,
+    SimpleResponse,
+    SubscriptionInfo,
+    ToggleAutoRenewRequest,
+)
+from fastapi import APIRouter, Depends, HTTPException, status
+
+from packages.ports.user_repository import UserRepository
+
+router = APIRouter()
+
+
+# ============ 配额定义（硬编码，后续可迁移到配置中心） ============
+
+PLAN_QUOTAS = {
+    "free": {"max_projects": 3, "max_storage_gb": 10},
+    "standard": {"max_projects": 10, "max_storage_gb": 50},
+    "pro": {"max_projects": -1, "max_storage_gb": 100},
+    "enterprise": {"max_projects": -1, "max_storage_gb": 1000},
+}
+
+
+# ============ Helper Functions ============
+
+
+def _get_plan_name(plan_id: str) -> str:
+    """获取套餐显示名称"""
+    plan_names = {
+        "free": "体验版",
+        "standard": "标准版",
+        "pro": "专业版",
+        "enterprise": "企业版",
+    }
+    return plan_names.get(plan_id, "未知套餐")
+
+
+def _get_plan_price(plan_id: str, billing_cycle: str) -> float:
+    """获取套餐价格"""
+    prices = {
+        ("free", "monthly"): 0,
+        ("free", "yearly"): 0,
+        ("standard", "monthly"): 99,
+        ("standard", "yearly"): 999,
+        ("pro", "monthly"): 299,
+        ("pro", "yearly"): 2999,
+        ("enterprise", "monthly"): 999,
+        ("enterprise", "yearly"): 9999,
+    }
+    return prices.get((plan_id, billing_cycle), 0)
+
+
+def _build_subscription_info(user: AuthenticatedUser) -> SubscriptionInfo:
+    """构建订阅信息响应"""
+    now = datetime.now(timezone.utc)
+    if user.user.subscription_expires_at:
+        period_end = user.user.subscription_expires_at.isoformat()
+        period_start = now.isoformat()
+    else:
+        period_start = now.isoformat()
+        period_end = now.isoformat()
+
+    return SubscriptionInfo(
+        id=f"sub-{user.user.id[:8]}",
+        plan_id=user.user.subscription_plan or "free",
+        plan_name=_get_plan_name(user.user.subscription_plan or "free"),
+        status=user.user.subscription_status or "active",
+        billing_cycle="monthly",
+        current_period_start=period_start,
+        current_period_end=period_end,
+        amount=_get_plan_price(user.user.subscription_plan or "free", "monthly"),
+        auto_renew=True,
+        created_at=user.user.created_at.isoformat() if user.user.created_at else now.isoformat(),
+    )
+
+
+# ============ API Endpoints ============
+
+
+@router.get("/current", response_model=SubscriptionInfo)
+async def get_current_subscription(
+    current_user: AuthenticatedUser = Depends(get_current_user),
+):
+    """获取当前订阅信息"""
+    return _build_subscription_info(current_user)
+
+
+@router.get("/billing-records", response_model=List[BillingRecord])
+async def get_billing_records(
+    current_user: AuthenticatedUser = Depends(get_current_user),
+):
+    """获取账单记录列表"""
+    from packages.adapters.sqlalchemy_impl.billing_repository import SQLAlchemyBillingRepository
+    from packages.adapters.sqlalchemy_impl.session import SessionLocal
+
+    if SessionLocal is None:
+        return []
+
+    session = SessionLocal()
+    try:
+        repo = SQLAlchemyBillingRepository(session)
+        records = repo.find_by_user(current_user.user.id)
+        return [
+            BillingRecord(
+                id=r.id,
+                plan_name=r.plan_name,
+                amount=r.amount,
+                billing_cycle=r.billing_cycle,
+                status=r.status,
+                payment_method=r.payment_method or "未支付",
+                created_at=r.created_at.isoformat() if r.created_at else "",
+                invoice_url=r.invoice_url,
+            )
+            for r in records
+        ]
+    finally:
+        session.close()
+
+
+@router.post("/change-plan", response_model=ChangePlanResponse)
+async def change_plan(
+    request: ChangePlanRequest,
+    current_user: AuthenticatedUser = Depends(get_current_user),
+    user_repository: UserRepository = Depends(get_user_repository),
+):
+    """变更订阅套餐（升级/降级）"""
+    # TODO: 接入支付验证（支付宝/微信支付）
+    valid_plans = {"free", "standard", "pro", "enterprise"}
+    if request.target_plan_id not in valid_plans:
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=f"无效的套餐ID。支持的套餐: {', '.join(valid_plans)}",
+        )
+
+    valid_cycles = {"monthly", "yearly"}
+    if request.billing_cycle not in valid_cycles:
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail="无效的计费周期。支持: monthly, yearly",
+        )
+
+    user = current_user.user
+    current_plan = user.subscription_plan or "free"
+    target_plan = request.target_plan_id
+
+    if current_plan == target_plan:
+        return ChangePlanResponse(
+            success=False,
+            message=f"您已经是 {_get_plan_name(target_plan)}",
+        )
+
+    # 通过 dataclasses.replace 创建新实例（不直接修改 dataclass）
+    quotas = PLAN_QUOTAS.get(target_plan, PLAN_QUOTAS["free"])
+    updated_user = replace(
+        user,
+        subscription_plan=target_plan,
+        subscription_status="active",
+        max_projects=quotas["max_projects"],
+        max_storage_gb=quotas["max_storage_gb"],
+    )
+    user_repository.save(updated_user)
+
+    # 用更新后的用户构造响应
+    refreshed_auth_user = AuthenticatedUser(user=updated_user)
+
+    return ChangePlanResponse(
+        success=True,
+        message=f"套餐已成功变更为 {_get_plan_name(target_plan)}",
+        new_subscription=_build_subscription_info(refreshed_auth_user),
+    )
+
+
+@router.post("/cancel", response_model=SimpleResponse)
+async def cancel_subscription(
+    current_user: AuthenticatedUser = Depends(get_current_user),
+    user_repository: UserRepository = Depends(get_user_repository),
+):
+    """取消订阅"""
+    user = current_user.user
+    if user.subscription_plan == "free":
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail="体验版无需取消",
+        )
+
+    updated_user = replace(user, subscription_status="cancelled")
+    user_repository.save(updated_user)
+
+    return SimpleResponse(
+        success=True,
+        message="订阅已取消，当前周期结束后停止服务",
+    )
+
+
+@router.post("/payment-callback")
+async def payment_callback(
+    user_id: str,
+    plan: str,
+    billing_cycle: str,
+    amount: float,
+    payment_method: str = "alipay",
+    payment_id: str = "",
+):
+    """支付回调 - 在事务中更新账单和订阅状态
+
+    注意：生产环境需要验证支付签名
+    """
+    import uuid
+    from datetime import timedelta
+
+    from packages.adapters.sqlalchemy_impl.billing_repository import SQLAlchemyBillingRepository
+    from packages.adapters.sqlalchemy_impl.session import SessionLocal
+
+    if SessionLocal is None:
+        raise HTTPException(status_code=500, detail="Database not available")
+
+    session = SessionLocal()
+    try:
+        repo = SQLAlchemyBillingRepository(session)
+
+        # 创建账单记录
+        record_id = uuid.uuid4().hex
+        record = repo.create(
+            {
+                "id": record_id,
+                "user_id": user_id,
+                "plan_name": _get_plan_name(plan),
+                "amount": amount,
+                "billing_cycle": billing_cycle,
+                "status": "pending",
+            }
+        )
+
+        # 在事务中标记支付成功并更新订阅
+        repo.mark_paid(record_id, payment_method, payment_id)
+
+        # 计算到期时间
+        days = 365 if billing_cycle == "yearly" else 30
+        expires_at = datetime.now(timezone.utc) + timedelta(days=days)
+        repo.update_subscription_on_payment(user_id, plan, expires_at)
+
+        return {"success": True, "message": "支付成功", "record_id": record_id}
+    except Exception as e:
+        session.rollback()
+        raise HTTPException(status_code=500, detail=f"支付处理失败: {str(e)}")
+    finally:
+        session.close()
+
+
+@router.post("/toggle-auto-renew", response_model=SimpleResponse)
+async def toggle_auto_renew(
+    request: ToggleAutoRenewRequest,
+    current_user: AuthenticatedUser = Depends(get_current_user),
+):
+    """切换自动续费"""
+    # TODO: 实际需要在数据库中存储 auto_renew 字段
+    status_text = "已开启自动续费" if request.enabled else "已关闭自动续费"
+
+    return SimpleResponse(
+        success=True,
+        message=status_text,
+    )

apps/api/app/api/routes/tags.py

@@ -0,0 +1,67 @@
+"""标签 CRUD 路由。"""
+
+import logging
+from typing import Any
+
+from app.auth import AuthenticatedUser, get_current_user
+from app.dependencies import get_tag_repository
+from app.schemas.tag import (
+    CreateTagRequest,
+    ListTagsResponse,
+    TagResponse,
+)
+from fastapi import APIRouter, Depends, HTTPException
+
+from packages.domain import Tag
+
+logger = logging.getLogger(__name__)
+
+router = APIRouter()
+
+
+@router.get("", response_model=ListTagsResponse)
+def list_tags(
+    skip: int = 0,
+    limit: int = 100,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    tag_repository: Any = Depends(get_tag_repository),
+) -> ListTagsResponse:
+    """列出当前用户的标签。"""
+    user_id = authenticated_user.user.id
+    items = tag_repository.list_by_user(user_id, skip=skip, limit=limit)
+    total = tag_repository.count_by_user(user_id)
+    return ListTagsResponse(
+        items=[TagResponse(id=t.id, name=t.name, created_at=t.created_at) for t in items],
+        total=total,
+    )
+
+
+@router.post("", response_model=TagResponse, status_code=201)
+def create_tag(
+    request: CreateTagRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    tag_repository: Any = Depends(get_tag_repository),
+) -> TagResponse:
+    """创建标签（同用户同名去重，返回 409）。"""
+    user_id = authenticated_user.user.id
+    existing = tag_repository.find_by_name(user_id, request.name)
+    if existing:
+        raise HTTPException(status_code=409, detail="标签名称已存在")
+    tag = Tag.create(user_id=user_id, name=request.name)
+    created = tag_repository.create(tag)
+    return TagResponse(id=created.id, name=created.name, created_at=created.created_at)
+
+
+@router.delete("/{tag_id}", status_code=204)
+def delete_tag(
+    tag_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    tag_repository: Any = Depends(get_tag_repository),
+) -> None:
+    """删除标签（同时清理素材关联）。"""
+    tag = tag_repository.get(tag_id)
+    if tag is None:
+        raise HTTPException(status_code=404, detail="标签不存在")
+    if tag.user_id != authenticated_user.user.id:
+        raise HTTPException(status_code=403, detail="无权删除该标签")
+    tag_repository.delete(tag_id)

apps/api/app/api/routes/task_center.py

@@ -7,7 +7,12 @@ from app.dependencies import (
     get_ingest_job_repository,
     get_project_repository,
 )
-from app.schemas.task_center import ListProjectTasksResponse, ProjectTaskResponse
+from app.schemas.task_center import (
+    ListProjectTasksResponse,
+    ListTasksResponse,
+    ProjectTaskResponse,
+    UserTaskResponse,
+)
 from fastapi import APIRouter, Depends, HTTPException
 
 from packages.application import (
@@ -34,28 +39,142 @@ def _humanize_task_error(error_message: str) -> str:
     return f"任务失败：{raw}"
 
 
+def _status_value(status) -> str:
+    """安全获取状态值（兼容 StrEnum 和 plain string）。"""
+    return status.value if hasattr(status, "value") else str(status)
+
+
 def _generation_step(task) -> str:
-    if task.status.value == "pending":
+    s = _status_value(task.status)
+    if s == "pending":
         return "等待 Worker 执行"
-    if task.status.value == "running":
+    if s == "running":
         return "正在生成成片"
-    if task.status.value == "completed":
+    if s == "completed":
         return "生成完成"
-    if task.status.value == "failed":
+    if s == "failed":
         return "生成失败"
-    return task.status.value
+    return s
 
 
 def _ingest_step(job) -> str:
-    if job.status.value == "pending":
+    s = _status_value(job.status)
+    if s == "pending":
         return "等待导入"
-    if job.status.value == "processing":
+    if s == "processing":
         return "正在分析素材"
-    if job.status.value == "completed":
+    if s == "completed":
         return "导入完成"
-    if job.status.value == "failed":
+    if s == "failed":
         return "导入失败"
-    return job.status.value
+    return s
+
+
+def _generation_task_to_project_response(task) -> ProjectTaskResponse:
+    return ProjectTaskResponse(
+        id=f"generation:{task.id}",
+        task_type="generation",
+        project_id=task.project_id,
+        status=_status_value(task.status),
+        progress=task.progress,
+        current_step=_generation_step(task),
+        error_message=task.error_message,
+        user_message=_humanize_task_error(task.error_message),
+        retryable=_status_value(task.status) == "failed",
+        source_id=task.id,
+        template_id=task.template_id,
+        created_at=task.created_at,
+        updated_at=task.completed_at or task.started_at or task.created_at,
+    )
+
+
+# ── 用户级端点（放在项目级端点之前，避免路由冲突） ──
+
+
+@router.get("/tasks", response_model=ListTasksResponse)
+def list_user_tasks(
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    ingest_job_repository: Any = Depends(get_ingest_job_repository),
+    generation_task_repository: Any = Depends(get_generation_task_repository),
+) -> ListTasksResponse:
+    """用户级任务列表（跨 project），合并 ingest + generation 任务。"""
+    user_id = authenticated_user.user.id
+    items: list[UserTaskResponse] = []
+
+    for task in generation_task_repository.list_by_user(user_id):
+        items.append(
+            UserTaskResponse(
+                id=f"generation:{task.id}",
+                task_type="generation",
+                project_id=task.project_id,
+                template_id=task.template_id,
+                status=_status_value(task.status),
+                progress=task.progress,
+                current_step=_generation_step(task),
+                error_message=task.error_message,
+                user_message=_humanize_task_error(task.error_message),
+                retryable=_status_value(task.status) == "failed",
+                source_id=task.id,
+                created_at=task.created_at,
+                updated_at=task.completed_at or task.started_at or task.created_at,
+            )
+        )
+
+    items.sort(key=lambda item: item.updated_at or item.created_at or "", reverse=True)
+    return ListTasksResponse(items=items)
+
+
+@router.post("/tasks/{task_id}/retry", response_model=UserTaskResponse)
+def retry_task_by_id(
+    task_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    generation_task_repository: Any = Depends(get_generation_task_repository),
+) -> UserTaskResponse:
+    """简化重试：通过 task_id 直接重试失败的生成任务。"""
+    task = generation_task_repository.get(task_id)
+    if task is None:
+        raise HTTPException(status_code=404, detail="Generation task not found")
+    if task.created_by_user_id and task.created_by_user_id != authenticated_user.user.id:
+        raise HTTPException(status_code=403, detail="Access denied to this task")
+    if _status_value(task.status) != "failed":
+        raise HTTPException(status_code=409, detail="Only failed tasks can be retried")
+
+    use_case = CreateGenerationTaskUseCase(generation_task_repository)
+    retried = use_case.execute(
+        CreateGenerationTaskCommand(
+            project_id=task.project_id,
+            asset_library_id=task.asset_library_id,
+            strategy_id=task.strategy_id,
+            voice_library_id=task.voice_library_id,
+            template_id=task.template_id,
+            asset_ids=task.asset_ids,
+            title_ids=task.title_ids,
+            voice_ids=task.voice_ids,
+            created_by_user_id=authenticated_user.user.id,
+            is_preview=getattr(task, "is_preview", True),
+            source_task_id=getattr(task, "source_task_id", ""),
+            output_width=getattr(task, "output_width", 1280),
+            output_height=getattr(task, "output_height", 720),
+            cover_url=getattr(task, "cover_url", ""),
+            custom_title=getattr(task, "custom_title", ""),
+        )
+    )
+    celery_app.send_task("worker.generate_video", args=[retried.id])
+    return UserTaskResponse(
+        id=f"generation:{retried.id}",
+        task_type="generation",
+        project_id=retried.project_id,
+        template_id=retried.template_id,
+        status=_status_value(retried.status),
+        progress=retried.progress,
+        current_step=_generation_step(retried),
+        source_id=retried.id,
+        created_at=retried.created_at,
+        updated_at=retried.created_at,
+    )
+
+
+# ── 项目级端点 ──
 
 
 @router.get("/projects/{project_id}/tasks", response_model=ListProjectTasksResponse)
@@ -77,34 +196,19 @@ def list_project_tasks(
                 id=f"ingest:{job.id}",
                 task_type="ingest",
                 project_id=job.project_id,
-                status=job.status.value,
-                progress=100.0 if job.status.value == "completed" else 0.0,
+                status=_status_value(job.status),
+                progress=100.0 if _status_value(job.status) == "completed" else 0.0,
                 current_step=_ingest_step(job),
                 error_message=job.error_message,
                 user_message=_humanize_task_error(job.error_message),
-                retryable=job.status.value == "failed",
+                retryable=_status_value(job.status) == "failed",
                 source_id=job.id,
                 created_at=job.created_at,
                 updated_at=job.updated_at,
             )
         )
     for task in generation_task_repository.list_by_project(project_id):
-        items.append(
-            ProjectTaskResponse(
-                id=f"generation:{task.id}",
-                task_type="generation",
-                project_id=task.project_id,
-                status=task.status.value,
-                progress=task.progress,
-                current_step=_generation_step(task),
-                error_message=task.error_message,
-                user_message=_humanize_task_error(task.error_message),
-                retryable=task.status.value == "failed",
-                source_id=task.id,
-                created_at=task.created_at,
-                updated_at=task.completed_at or task.started_at or task.created_at,
-            )
-        )
+        items.append(_generation_task_to_project_response(task))
     items.sort(key=lambda item: item.updated_at or item.created_at or "", reverse=True)
     return ListProjectTasksResponse(items=items)
 
@@ -121,7 +225,7 @@ def retry_project_task(
         task = generation_task_repository.get(source_id)
         if task is None:
             raise HTTPException(status_code=404, detail="Generation task not found")
-        if task.status.value != "failed":
+        if _status_value(task.status) != "failed":
             raise HTTPException(status_code=409, detail="Only failed tasks can be retried")
         use_case = CreateGenerationTaskUseCase(generation_task_repository)
         retried = use_case.execute(
@@ -130,27 +234,26 @@ def retry_project_task(
                 asset_library_id=task.asset_library_id,
                 strategy_id=task.strategy_id,
                 voice_library_id=task.voice_library_id,
-                edit_plan_id=task.edit_plan_id,
+                template_id=task.template_id,
+                asset_ids=task.asset_ids,
+                title_ids=task.title_ids,
+                voice_ids=task.voice_ids,
                 created_by_user_id=authenticated_user.user.id,
+                is_preview=getattr(task, "is_preview", True),
+                source_task_id=getattr(task, "source_task_id", ""),
+                output_width=getattr(task, "output_width", 1280),
+                output_height=getattr(task, "output_height", 720),
+                cover_url=getattr(task, "cover_url", ""),
+                custom_title=getattr(task, "custom_title", ""),
             )
         )
         celery_app.send_task("worker.generate_video", args=[retried.id])
-        return ProjectTaskResponse(
-            id=f"generation:{retried.id}",
-            task_type="generation",
-            project_id=retried.project_id,
-            status=retried.status.value,
-            progress=retried.progress,
-            current_step=_generation_step(retried),
-            source_id=retried.id,
-            created_at=retried.created_at,
-            updated_at=retried.created_at,
-        )
+        return _generation_task_to_project_response(retried)
     if task_type == "ingest":
         job = ingest_job_repository.get(source_id)
         if job is None:
             raise HTTPException(status_code=404, detail="Ingest job not found")
-        if job.status.value != "failed":
+        if _status_value(job.status) != "failed":
             raise HTTPException(status_code=409, detail="Only failed tasks can be retried")
         use_case = SubmitIngestJobUseCase(ingest_job_repository)
         retried = use_case.execute(
@@ -165,7 +268,7 @@ def retry_project_task(
             id=f"ingest:{retried.id}",
             task_type="ingest",
             project_id=retried.project_id,
-            status=retried.status.value,
+            status=_status_value(retried.status),
             progress=0,
             current_step=_ingest_step(retried),
             source_id=retried.id,

apps/api/app/api/routes/templates.py

@@ -0,0 +1,321 @@
+"""Template CRUD + generate + category routes."""
+
+from __future__ import annotations
+
+import logging
+
+from app.auth import AuthenticatedUser, get_current_user
+from app.dependencies import get_db_session
+from app.schemas.template import (
+    CategoryResponse,
+    CreateCategoryRequest,
+    CreateTemplateRequest,
+    GenerateWarningResponse,
+    ListCategoriesResponse,
+    ListTemplatesResponse,
+    SegmentResponse,
+    TemplateResponse,
+    ToggleFavoriteResponse,
+    UpdateTemplateRequest,
+    ValidateTemplateRequest,
+    ValidateTemplateResponse,
+)
+from fastapi import APIRouter, Depends, HTTPException, Query, Response, status
+from sqlalchemy.orm import Session
+
+logger = logging.getLogger(__name__)
+
+from packages.adapters.sqlalchemy_impl.template_repository import SQLAlchemyTemplateRepository
+from packages.application.template.commands import (
+    CreateCategoryCommand,
+    CreateTemplateCommand,
+    SegmentCommand,
+    UpdateTemplateCommand,
+    ValidateTemplateCommand,
+)
+from packages.application.template.use_cases import (
+    CreateCategoryUseCase,
+    CreateTemplateUseCase,
+    DeleteCategoryUseCase,
+    DeleteTemplateUseCase,
+    GetTemplateUseCase,
+    ListCategoriesUseCase,
+    ListTemplatesUseCase,
+    NotFoundError,
+    UpdateTemplateUseCase,
+    ValidateTemplateUseCase,
+    ValidationError,
+)
+
+router = APIRouter()
+
+
+def _get_template_repository(session: Session = Depends(get_db_session)) -> SQLAlchemyTemplateRepository:
+    return SQLAlchemyTemplateRepository(session)
+
+
+def _segment_to_response(seg) -> SegmentResponse:
+    return SegmentResponse(
+        id=seg.id,
+        template_id=seg.template_id,
+        segment_order=seg.segment_order,
+        duration_min=seg.duration_min,
+        duration_max=seg.duration_max,
+        material_type=seg.material_type,
+        created_at=seg.created_at,
+        updated_at=seg.updated_at,
+    )
+
+
+def _to_response(template) -> TemplateResponse:
+    return TemplateResponse(
+        id=template.id,
+        user_id=template.user_id,
+        name=template.name,
+        mode=template.mode,
+        category=template.category,
+        tags=template.tags,
+        title_config=template.title_config,
+        subtitle_config=template.subtitle_config,
+        bgm_config=template.bgm_config,
+        estimated_duration=template.estimated_duration,
+        segments=[_segment_to_response(s) for s in getattr(template, "segments", [])],
+        is_active=template.is_active,
+        created_at=template.created_at,
+        updated_at=template.updated_at,
+    )
+
+
+# ── Template CRUD ──
+
+
+@router.get("", response_model=ListTemplatesResponse)
+def list_templates(
+    skip: int = Query(0, ge=0),
+    limit: int = Query(50, ge=1, le=200),
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    template_repository: SQLAlchemyTemplateRepository = Depends(_get_template_repository),
+) -> ListTemplatesResponse:
+    user_id = authenticated_user.user.id
+    try:
+        use_case = ListTemplatesUseCase(template_repository)
+        templates = use_case.execute(user_id, skip=skip, limit=limit)
+        total = template_repository.count_by_user(user_id)
+    except Exception:
+        logger.exception("list_templates 查询失败: user_id=%s", user_id)
+        return ListTemplatesResponse(items=[], total=0)
+    return ListTemplatesResponse(
+        items=[_to_response(t) for t in templates],
+        total=total,
+    )
+
+
+@router.get("/{template_id}", response_model=TemplateResponse)
+def get_template(
+    template_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    template_repository: SQLAlchemyTemplateRepository = Depends(_get_template_repository),
+) -> TemplateResponse:
+    user_id = authenticated_user.user.id
+    try:
+        use_case = GetTemplateUseCase(template_repository)
+        template = use_case.execute(template_id, user_id)
+    except Exception:
+        logger.exception("get_template 查询失败: template_id=%s", template_id)
+        raise HTTPException(status_code=status.HTTP_500_INTERNAL_SERVER_ERROR, detail="模板查询失败")
+    if template is None:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Template not found")
+    return _to_response(template)
+
+
+@router.post("", response_model=TemplateResponse, status_code=status.HTTP_201_CREATED)
+def create_template(
+    request: CreateTemplateRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    template_repository: SQLAlchemyTemplateRepository = Depends(_get_template_repository),
+) -> TemplateResponse:
+    user_id = authenticated_user.user.id
+    command = CreateTemplateCommand(
+        user_id=user_id,
+        name=request.name,
+        mode=request.mode,
+        category=request.category,
+        tags=request.tags,
+        title_config=request.title_config,
+        subtitle_config=request.subtitle_config,
+        bgm_config=request.bgm_config,
+        estimated_duration=request.estimated_duration,
+        segments=[
+            SegmentCommand(
+                segment_order=s.segment_order,
+                duration_min=s.duration_min,
+                duration_max=s.duration_max,
+                material_type=s.material_type,
+            )
+            for s in request.segments
+        ],
+    )
+    use_case = CreateTemplateUseCase(template_repository)
+    try:
+        template = use_case.execute(command)
+    except ValidationError as exc:
+        raise HTTPException(status_code=status.HTTP_422_UNPROCESSABLE_ENTITY, detail=str(exc))
+    return _to_response(template)
+
+
+@router.patch("/{template_id}", response_model=TemplateResponse)
+def update_template(
+    template_id: str,
+    request: UpdateTemplateRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    template_repository: SQLAlchemyTemplateRepository = Depends(_get_template_repository),
+) -> TemplateResponse:
+    user_id = authenticated_user.user.id
+    command = UpdateTemplateCommand(
+        template_id=template_id,
+        user_id=user_id,
+        name=request.name,
+        mode=request.mode,
+        category=request.category,
+        tags=request.tags,
+        title_config=request.title_config,
+        subtitle_config=request.subtitle_config,
+        bgm_config=request.bgm_config,
+        estimated_duration=request.estimated_duration,
+        segments=(
+            [
+                SegmentCommand(
+                    segment_order=s.segment_order,
+                    duration_min=s.duration_min,
+                    duration_max=s.duration_max,
+                    material_type=s.material_type,
+                )
+                for s in request.segments
+            ]
+            if request.segments is not None
+            else None
+        ),
+    )
+    use_case = UpdateTemplateUseCase(template_repository)
+    try:
+        template = use_case.execute(command)
+    except NotFoundError:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Template not found")
+    except ValidationError as exc:
+        raise HTTPException(status_code=status.HTTP_422_UNPROCESSABLE_ENTITY, detail=str(exc))
+    return _to_response(template)
+
+
+@router.delete("/{template_id}", status_code=status.HTTP_204_NO_CONTENT, response_model=None)
+def delete_template(
+    template_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    template_repository: SQLAlchemyTemplateRepository = Depends(_get_template_repository),
+) -> Response:
+    user_id = authenticated_user.user.id
+    use_case = DeleteTemplateUseCase(template_repository)
+    deleted = use_case.execute(template_id, user_id)
+    if not deleted:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Template not found")
+    return Response(status_code=204)
+
+
+@router.post("/{template_id}/toggle-favorite", response_model=ToggleFavoriteResponse)
+def toggle_favorite(
+    template_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    template_repository: SQLAlchemyTemplateRepository = Depends(_get_template_repository),
+) -> ToggleFavoriteResponse:
+    """切换模板收藏状态（当前为兼容端点，始终返回 false）"""
+    user_id = authenticated_user.user.id
+    use_case = GetTemplateUseCase(template_repository)
+    try:
+        template = use_case.execute(template_id, user_id)
+    except Exception:
+        logger.exception("toggle_favorite 查询失败: template_id=%s", template_id)
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Template not found")
+    if template is None:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Template not found")
+    return ToggleFavoriteResponse(id=template_id, is_favorite=False)
+
+
+# ── Validate template ──
+
+
+@router.post("/{template_id}/validate", response_model=ValidateTemplateResponse)
+def validate_template(
+    template_id: str,
+    request: ValidateTemplateRequest = ValidateTemplateRequest(),
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    template_repository: SQLAlchemyTemplateRepository = Depends(_get_template_repository),
+) -> ValidateTemplateResponse:
+    user_id = authenticated_user.user.id
+    command = ValidateTemplateCommand(
+        template_id=template_id,
+        user_id=user_id,
+        voiceover_duration=request.voiceover_duration,
+    )
+    use_case = ValidateTemplateUseCase(template_repository)
+    try:
+        result = use_case.execute(command)
+    except NotFoundError:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Template not found")
+    except ValidationError as exc:
+        raise HTTPException(status_code=status.HTTP_422_UNPROCESSABLE_ENTITY, detail=str(exc))
+
+    return ValidateTemplateResponse(
+        template=_to_response(result.template),
+        warnings=[GenerateWarningResponse(code=w.code, message=w.message, details=w.details) for w in result.warnings],
+    )
+
+
+# ── Category CRUD ──
+
+
+@router.get("/categories/list", response_model=ListCategoriesResponse)
+def list_categories(
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    template_repository: SQLAlchemyTemplateRepository = Depends(_get_template_repository),
+) -> ListCategoriesResponse:
+    user_id = authenticated_user.user.id
+    try:
+        use_case = ListCategoriesUseCase(template_repository)
+        categories = use_case.execute(user_id)
+    except Exception:
+        logger.exception("list_categories 查询失败: user_id=%s", user_id)
+        return ListCategoriesResponse(items=[])
+    return ListCategoriesResponse(
+        items=[CategoryResponse(id=c.id, user_id=c.user_id, name=c.name, created_at=c.created_at) for c in categories],
+    )
+
+
+@router.post("/categories", response_model=CategoryResponse, status_code=status.HTTP_201_CREATED)
+def create_category(
+    request: CreateCategoryRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    template_repository: SQLAlchemyTemplateRepository = Depends(_get_template_repository),
+) -> CategoryResponse:
+    user_id = authenticated_user.user.id
+    command = CreateCategoryCommand(user_id=user_id, name=request.name)
+    use_case = CreateCategoryUseCase(template_repository)
+    category = use_case.execute(command)
+    return CategoryResponse(
+        id=category.id,
+        user_id=category.user_id,
+        name=category.name,
+        created_at=category.created_at,
+    )
+
+
+@router.delete("/categories/{category_id}", status_code=status.HTTP_204_NO_CONTENT, response_model=None)
+def delete_category(
+    category_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    template_repository: SQLAlchemyTemplateRepository = Depends(_get_template_repository),
+) -> Response:
+    user_id = authenticated_user.user.id
+    use_case = DeleteCategoryUseCase(template_repository)
+    deleted = use_case.execute(category_id, user_id)
+    if not deleted:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Category not found")
+    return Response(status_code=204)

apps/api/app/api/routes/titles.py

@@ -1,10 +1,9 @@
 """Title library CRUD routes."""
+
 from __future__ import annotations
 
 from typing import Optional
 
-from fastapi import APIRouter, Depends, HTTPException, Query, Response, status
-
 from app.auth import AuthenticatedUser, get_current_user
 from app.dependencies import get_db_session, get_user_repository
 from app.schemas.title_library import (
@@ -13,6 +12,9 @@ from app.schemas.title_library import (
     TitleLibraryItemResponse,
     UpdateTitleLibraryRequest,
 )
+from fastapi import APIRouter, Depends, HTTPException, Query, Response, status
+from sqlalchemy.orm import Session
+
 from packages.adapters.sqlalchemy_impl.title_library_repository import SQLAlchemyTitleLibraryRepository
 from packages.application.title_library.commands import CreateTitleLibraryCommand, UpdateTitleLibraryCommand
 from packages.application.title_library.use_cases import (
@@ -20,12 +22,11 @@ from packages.application.title_library.use_cases import (
     DeleteTitleLibraryUseCase,
     GetTitleLibraryUseCase,
     ListTitleLibraryUseCase,
-    UpdateTitleLibraryUseCase,
     NotFoundError,
     QuotaExceededError,
+    UpdateTitleLibraryUseCase,
 )
 from packages.ports.user_repository import UserRepository
-from sqlalchemy.orm import Session
 
 router = APIRouter()
 
@@ -142,7 +143,7 @@ def update_title(
     return _to_response(item)
 
 
-@router.delete("/{title_id}", status_code=status.HTTP_204_NO_CONTENT, response_class=Response)
+@router.delete("/{title_id}", status_code=status.HTTP_204_NO_CONTENT, response_model=None)
 def delete_title(
     title_id: str,
     authenticated_user: AuthenticatedUser = Depends(get_current_user),

apps/api/app/api/routes/tts.py

@@ -0,0 +1,370 @@
+"""TTS 合成 API 路由。"""
+
+from __future__ import annotations
+
+import logging
+from typing import Optional
+
+from app.auth import AuthenticatedUser, get_current_user
+from app.dependencies import (
+    get_cosyvoice_service,
+    get_db_session,
+    get_user_repository,
+    get_voice_clone_profile_repository,
+    get_voice_library_repository,
+)
+from app.schemas.tts import (
+    ListTTSJobResponse,
+    SaveToLibraryRequest,
+    SaveToLibraryResponse,
+    TTSJobResponse,
+    TTSStatusResponse,
+    TTSSynthesizeRequest,
+    TTSSynthesizeResponse,
+)
+from fastapi import APIRouter, Depends, HTTPException, Query, Response, WebSocket, WebSocketDisconnect, status
+from sqlalchemy.orm import Session
+
+from packages.adapters.sqlalchemy_impl.tts_job_repository import (
+    SQLAlchemyTTSJobRepository,
+)
+from packages.adapters.sqlalchemy_impl.voice_library_repository import SQLAlchemyVoiceLibraryRepository
+from packages.application.cosyvoice_service import CosyVoiceService
+from packages.application.tts_job.streaming_service import TTSStreamingService
+from packages.application.tts_job.use_cases import (
+    CreateTTSJobUseCase,
+    DeleteTTSJobUseCase,
+    GetTTSJobStatusUseCase,
+    GetTTSJobUseCase,
+    ListTTSJobsUseCase,
+    TTSJobNotFoundError,
+)
+from packages.application.tts_job.workflow import TTSWorkflowService
+from packages.application.voice_library.commands import CreateVoiceLibraryCommand
+from packages.application.voice_library.use_cases import (
+    CreateVoiceLibraryUseCase,
+    QuotaExceededError,
+)
+from packages.ports.user_repository import UserRepository
+
+logger = logging.getLogger(__name__)
+
+router = APIRouter()
+
+
+def _get_repository(session: Session = Depends(get_db_session)) -> SQLAlchemyTTSJobRepository:
+    return SQLAlchemyTTSJobRepository(session)
+
+
+def _to_response(job) -> TTSJobResponse:
+    return TTSJobResponse(
+        id=job.id,
+        user_id=job.user_id,
+        input_text=job.input_text,
+        voice_id=job.voice_id,
+        voice_model=job.voice_model,
+        project_id=job.project_id,
+        voice_clone_profile_id=job.voice_clone_profile_id,
+        status=job.status,
+        output_audio_url=job.output_audio_url,
+        output_audio_key=job.output_audio_key,
+        duration=job.duration,
+        file_size=job.file_size,
+        sample_rate=job.sample_rate,
+        format=job.format,
+        error_message=job.error_message,
+        retry_count=job.retry_count,
+        max_retries=job.max_retries,
+        metadata=job.metadata,
+        started_at=job.started_at,
+        completed_at=job.completed_at,
+        created_at=job.created_at,
+        updated_at=job.updated_at,
+    )
+
+
+@router.post("/synthesize", response_model=TTSSynthesizeResponse, status_code=status.HTTP_201_CREATED)
+def synthesize(
+    request: TTSSynthesizeRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    repository: SQLAlchemyTTSJobRepository = Depends(_get_repository),
+    cosyvoice_service: CosyVoiceService = Depends(get_cosyvoice_service),
+    voice_clone_repo=Depends(get_voice_clone_profile_repository),
+) -> TTSSynthesizeResponse:
+    """发起 TTS 合成任务。
+
+    创建 TTS 任务 → 提交 CosyVoice 合成 → 触发 Celery 异步轮询。
+    与音色克隆接口保持一致：CosyVoice 失败时不抛 500，而是返回 201 + failed 状态任务记录。
+    """
+    user_id = authenticated_user.user.id
+
+    # 校验 voice_clone_profile_id 归属（防止越权使用他人克隆音色）
+    if request.voice_clone_profile_id:
+        profile = voice_clone_repo.get(request.voice_clone_profile_id)
+        if profile is None:
+            raise HTTPException(
+                status_code=status.HTTP_404_NOT_FOUND,
+                detail="Voice clone profile not found",
+            )
+        if profile.user_id != user_id:
+            raise HTTPException(
+                status_code=status.HTTP_403_FORBIDDEN,
+                detail="Access denied to voice clone profile",
+            )
+
+    use_case = CreateTTSJobUseCase(repository)
+    job = use_case.execute(
+        user_id=user_id,
+        input_text=request.text,
+        voice_id=request.voice_id,
+        voice_model=request.voice_model,
+        voice_clone_profile_id=request.voice_clone_profile_id,
+        metadata=request.metadata_,
+    )
+
+    # 提交 CosyVoice 合成任务
+    workflow = TTSWorkflowService(
+        repository=repository,
+        cosyvoice_service=cosyvoice_service,
+    )
+
+    try:
+        job = workflow.start_synthesis(job.id)
+    except Exception as e:
+        # 兜底：workflow 内部已捕获 CosyVoiceError / ValueError，
+        # 但 DB 异常、网络异常等意外错误可能逃逸。
+        # 与音色克隆接口保持一致：标记 failed，返回 201，不抛 500。
+        logger.error(f"TTS 合成异常: job_id={job.id}, error={e}", exc_info=True)
+        try:
+            job = workflow.process_synthesis_failure(job.id, str(e))
+        except Exception as inner_e:
+            logger.error(f"标记 TTS job 失败时出错: job_id={job.id}, error={inner_e}")
+
+    # 若任务处于 processing 状态（异步模式），触发 Celery 后台轮询
+    if job.status.value == "processing":
+        # 分段合成任务 vs 普通单段任务
+        segment_task_ids = (job.metadata or {}).get("segment_task_ids", [])
+        is_segment = len(segment_task_ids) > 0
+
+        try:
+            if is_segment:
+                from worker_app.tasks import process_tts_segment_synthesis
+
+                process_tts_segment_synthesis.delay(job.id)
+            else:
+                from worker_app.tasks import process_tts_synthesis
+
+                process_tts_synthesis.delay(job.id)
+        except Exception as e:
+            # Celery 调度失败，标记 job 为 failed
+            try:
+                workflow.process_synthesis_failure(job.id, f"Celery 任务调度失败: {e}")
+            except Exception as inner_e:
+                logger.error(f"Celery 调度后标记失败时出错: job_id={job.id}, error={inner_e}")
+
+    return TTSSynthesizeResponse(
+        job_id=job.id,
+        status=job.status,
+        message="合成任务已创建",
+    )
+
+
+@router.get("/jobs", response_model=ListTTSJobResponse)
+def list_tts_jobs(
+    page: int = Query(default=1, ge=1, description="页码"),
+    page_size: int = Query(default=20, ge=1, le=100, description="每页数量"),
+    status_filter: Optional[str] = Query(None, alias="status"),
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    repository: SQLAlchemyTTSJobRepository = Depends(_get_repository),
+) -> ListTTSJobResponse:
+    """列出用户的 TTS 合成任务。"""
+    user_id = authenticated_user.user.id
+    use_case = ListTTSJobsUseCase(repository)
+    skip = (page - 1) * page_size
+    items, total = use_case.execute(user_id, status=status_filter, skip=skip, limit=page_size)
+    return ListTTSJobResponse(
+        items=[_to_response(j) for j in items],
+        total=total,
+        page=page,
+        page_size=page_size,
+    )
+
+
+@router.get("/jobs/{job_id}", response_model=TTSJobResponse)
+def get_tts_job(
+    job_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    repository: SQLAlchemyTTSJobRepository = Depends(_get_repository),
+) -> TTSJobResponse:
+    """获取 TTS 任务详情。"""
+    user_id = authenticated_user.user.id
+    use_case = GetTTSJobUseCase(repository)
+    try:
+        job = use_case.execute(job_id, user_id)
+    except TTSJobNotFoundError:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="TTS job not found")
+    return _to_response(job)
+
+
+@router.get("/jobs/{job_id}/status", response_model=TTSStatusResponse)
+def get_tts_job_status(
+    job_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    repository: SQLAlchemyTTSJobRepository = Depends(_get_repository),
+) -> TTSStatusResponse:
+    """查询 TTS 合成状态（用于前端轮询）。"""
+    user_id = authenticated_user.user.id
+    use_case = GetTTSJobStatusUseCase(repository)
+    try:
+        job = use_case.execute(job_id, user_id)
+    except TTSJobNotFoundError:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="TTS job not found")
+    return TTSStatusResponse(
+        id=job.id,
+        status=job.status,
+        output_audio_url=job.output_audio_url,
+        error_message=job.error_message,
+        duration=job.duration,
+        retry_count=job.retry_count,
+        created_at=job.created_at,
+        updated_at=job.updated_at,
+    )
+
+
+@router.delete("/jobs/{job_id}", status_code=status.HTTP_204_NO_CONTENT, response_model=None)
+def delete_tts_job(
+    job_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    repository: SQLAlchemyTTSJobRepository = Depends(_get_repository),
+) -> Response:
+    """删除 TTS 合成任务。"""
+    user_id = authenticated_user.user.id
+    use_case = DeleteTTSJobUseCase(repository)
+    deleted = use_case.execute(job_id, user_id)
+    if not deleted:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="TTS job not found")
+    return Response(status_code=204)
+
+
+@router.post(
+    "/jobs/{job_id}/save-to-library",
+    response_model=SaveToLibraryResponse,
+    status_code=status.HTTP_201_CREATED,
+)
+def save_tts_job_to_library(
+    job_id: str,
+    request: SaveToLibraryRequest = SaveToLibraryRequest(),
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    tts_repository: SQLAlchemyTTSJobRepository = Depends(_get_repository),
+    voice_library_repository: SQLAlchemyVoiceLibraryRepository = Depends(get_voice_library_repository),
+    user_repository: UserRepository = Depends(get_user_repository),
+) -> SaveToLibraryResponse:
+    """将已完成的 TTS 合成结果保存到配音库。
+
+    自动携带音色名、时长、语速等元信息。
+    """
+    user_id = authenticated_user.user.id
+
+    # 获取 TTS job
+    get_use_case = GetTTSJobUseCase(tts_repository)
+    try:
+        job = get_use_case.execute(job_id, user_id)
+    except TTSJobNotFoundError:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="TTS job not found")
+
+    # 校验已完成
+    if not job.is_completed:
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail="TTS job is not completed yet",
+        )
+
+    # 构建配音素材名称
+    name = request.name or f"TTS-{job.id[:8]}"
+
+    # 构建元信息
+    metadata_ = {
+        "source": "tts_job",
+        "tts_job_id": job.id,
+        "format": job.format,
+        "sample_rate": job.sample_rate,
+    }
+    if job.metadata:
+        # 保留原始 job 的有用元信息
+        for key in ("speed", "language"):
+            if key in job.metadata:
+                metadata_[key] = job.metadata[key]
+
+    # 获取用户套餐（用于配额检查）
+    user = user_repository.find_by_id(user_id)
+    plan_name = getattr(user, "subscription_plan", "free") if user else "free"
+
+    # 构建命令并执行
+    command = CreateVoiceLibraryCommand(
+        user_id=user_id,
+        name=name,
+        text=job.input_text,
+        voice_provider="cosyvoice",
+        voice_id=job.voice_id,
+        voice_name=job.voice_model or "",
+        audio_url=job.output_audio_url,
+        duration=job.duration,
+        file_size=job.file_size,
+        status="completed",
+        project_id=job.project_id or "",
+        tags=[],
+        metadata_=metadata_,
+    )
+
+    use_case = CreateVoiceLibraryUseCase(voice_library_repository)
+    try:
+        item = use_case.execute(command, plan_name=plan_name or "free")
+    except QuotaExceededError as exc:
+        raise HTTPException(
+            status_code=status.HTTP_429_TOO_MANY_REQUESTS,
+            detail=f"配音库配额已满（{exc.used}/{exc.limit}），请升级套餐",
+        )
+
+    return SaveToLibraryResponse(
+        id=item.id,
+        name=item.name,
+        audio_url=item.audio_url,
+        duration=item.duration,
+        voice_id=item.voice_id,
+        voice_name=item.voice_name,
+        status=item.status,
+    )
+
+
+@router.websocket("/ws/tts/stream")
+async def tts_websocket_stream(
+    websocket: WebSocket,
+    cosyvoice_service: CosyVoiceService = Depends(get_cosyvoice_service),
+) -> None:
+    """WebSocket 流式 TTS 合成。
+
+    协议：
+    1. 客户端发送 JSON 文本帧: {"text": "...", "voice_id": "...", ...}
+    2. 服务端发送 JSON 状态帧 + 二进制音频帧
+    3. 完成时发送 JSON 结束帧
+    """
+    await websocket.accept()
+    try:
+        message = await websocket.receive_json()
+        params = {
+            "text": message.get("text", ""),
+            "voice_id": message.get("voice_id", ""),
+            "sample_rate": message.get("sample_rate", 0),
+            "format": message.get("format", "mp3"),
+            "speed": message.get("speed", 1.0),
+        }
+        streaming_service = TTSStreamingService(cosyvoice_service)
+        await streaming_service.synthesize_and_stream(websocket, params)
+    except WebSocketDisconnect:
+        logger.info("WebSocket 客户端断开连接")
+    except Exception as e:
+        logger.error(f"WebSocket 流式合成异常: {e}", exc_info=True)
+        try:
+            await websocket.send_json({"type": "error", "message": f"服务异常: {e}"})
+        except Exception:
+            pass

apps/api/app/api/routes/upload.py

@@ -1,4 +1,5 @@
-from typing import Any
+import logging
+from typing import Annotated, Any
 from uuid import uuid4
 
 from app.auth import AuthenticatedUser, get_current_user
@@ -7,6 +8,7 @@ from app.core.celery_app import celery_app
 from app.core.storage import OSSStorageService, get_storage_service
 from app.dependencies import (
     get_asset_library_repository,
+    get_asset_repository,
     get_ingest_job_repository,
     get_project_repository,
 )
@@ -18,26 +20,45 @@ from app.schemas.upload import (
     UploadAssetRequest,
     UploadAssetResponse,
 )
-from typing import Annotated
-
 from fastapi import APIRouter, Depends, File, Form, HTTPException, UploadFile, status
 
 from packages.application import GetProjectUseCase, SubmitIngestJobCommand, SubmitIngestJobUseCase
+from packages.domain import Asset, AssetStatus
+
+logger = logging.getLogger(__name__)
 
 router = APIRouter()
 
 # 允许上传的文件 MIME 类型
-ALLOWED_MIME_TYPES = frozenset({
-    # 视频
-    "video/mp4", "video/mpeg", "video/quicktime", "video/x-msvideo",
-    "video/webm", "video/x-matroska", "video/3gpp",
-    # 音频
-    "audio/mpeg", "audio/wav", "audio/ogg", "audio/flac", "audio/aac",
-    "audio/mp3", "audio/x-m4a", "audio/webm",
-    # 图片
-    "image/jpeg", "image/png", "image/gif", "image/webp", "image/bmp",
-    "image/svg+xml", "image/tiff",
-})
+ALLOWED_MIME_TYPES = frozenset(
+    {
+        # 视频
+        "video/mp4",
+        "video/mpeg",
+        "video/quicktime",
+        "video/x-msvideo",
+        "video/webm",
+        "video/x-matroska",
+        "video/3gpp",
+        # 音频
+        "audio/mpeg",
+        "audio/wav",
+        "audio/ogg",
+        "audio/flac",
+        "audio/aac",
+        "audio/mp3",
+        "audio/x-m4a",
+        "audio/webm",
+        # 图片
+        "image/jpeg",
+        "image/png",
+        "image/gif",
+        "image/webp",
+        "image/bmp",
+        "image/tiff",
+        "image/svg+xml",
+    }
+)
 
 
 def _validate_mime_type(content_type: str | None) -> str:
@@ -47,16 +68,16 @@ def _validate_mime_type(content_type: str | None) -> str:
             status_code=status.HTTP_400_BAD_REQUEST,
             detail="Content-Type header is required",
         )
-    
+
     # 处理带参数的类型，如 "video/mp4; charset=utf-8"
     base_type = content_type.split(";")[0].strip().lower()
-    
+
     if base_type not in ALLOWED_MIME_TYPES:
         raise HTTPException(
             status_code=status.HTTP_415_UNSUPPORTED_MEDIA_TYPE,
             detail=f"File type '{base_type}' is not supported. Allowed types: video, audio, and image files.",
         )
-    
+
     return base_type
 
 
@@ -70,16 +91,78 @@ def _require_project_and_library(
     if project is None:
         raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Project not found")
 
-    libraries = asset_library_repository.list_by_project(project_id)
+    libraries = asset_library_repository.find_by_project(project_id)
     if not any(item.id == library_id for item in libraries):
         raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Asset library not found")
 
 
+def _infer_mime_type_from_storage_key(storage_key: str) -> str:
+    """从 storage_key 推断 MIME 类型（与 worker 端保持一致）。"""
+    lower_filename = storage_key.rsplit("/", 1)[-1].lower()
+    if lower_filename.endswith(".mov"):
+        return "video/quicktime"
+    if lower_filename.endswith(".mp4"):
+        return "video/mp4"
+    if lower_filename.endswith(".avi"):
+        return "video/x-msvideo"
+    if lower_filename.endswith(".mkv"):
+        return "video/x-matroska"
+    if lower_filename.endswith(".webm"):
+        return "video/webm"
+    if lower_filename.endswith(".png"):
+        return "image/png"
+    if lower_filename.endswith(".gif"):
+        return "image/gif"
+    if lower_filename.endswith(".bmp"):
+        return "image/bmp"
+    if lower_filename.endswith(".svg"):
+        return "image/svg+xml"
+    if lower_filename.endswith((".jpg", ".jpeg")):
+        return "image/jpeg"
+    if lower_filename.endswith(".mp3"):
+        return "audio/mpeg"
+    if lower_filename.endswith(".wav"):
+        return "audio/wav"
+    if lower_filename.endswith(".ogg"):
+        return "audio/ogg"
+    if lower_filename.endswith(".flac"):
+        return "audio/flac"
+    if lower_filename.endswith(".m4a"):
+        return "audio/x-m4a"
+    # 默认视频
+    return "video/mp4"
+
+
+def _create_pending_asset(
+    asset_repository: Any,
+    project_id: str,
+    library_id: str,
+    storage_key: str,
+    filename: str,
+    mime_type: str,
+    user_id: str,
+    file_hash: str = "",
+) -> Any:
+    """立即创建一条 PROCESSING 状态的 Asset 记录，使前端能马上看到新素材。"""
+    asset = Asset.create(
+        project_id=project_id,
+        library_id=library_id,
+        name=filename,
+        storage_key=storage_key,
+        mime_type=mime_type,
+        status=AssetStatus.PROCESSING,
+        uploaded_by_user_id=user_id,
+        file_hash=file_hash,
+    )
+    return asset_repository.create(asset)
+
+
 def _submit_ingest_job(
     project_id: str,
     library_id: str,
     storage_key: str,
     ingest_job_repository: Any,
+    file_hash: str = "",
 ) -> Any:
     use_case = SubmitIngestJobUseCase(ingest_job_repository)
     job = use_case.execute(
@@ -87,6 +170,7 @@ def _submit_ingest_job(
             project_id=project_id,
             library_id=library_id,
             storage_key=storage_key,
+            file_hash=file_hash,
         )
     )
     celery_app.send_task("worker.ingest_asset", args=[job.id])
@@ -131,7 +215,14 @@ async def prepare_direct_upload(
             expires_seconds=settings.OSS_DIRECT_UPLOAD_EXPIRE_SECONDS,
         )
     except RuntimeError as error:
+        logger.error("OSS not configured for direct upload prepare: %s", error)
         raise HTTPException(status_code=status.HTTP_503_SERVICE_UNAVAILABLE, detail=str(error)) from error
+    except Exception as error:
+        logger.exception("Unexpected error in direct upload prepare: %s", error)
+        raise HTTPException(
+            status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,
+            detail=f"Failed to prepare upload: {type(error).__name__}",
+        ) from error
 
     return DirectUploadPrepareResponse(
         upload_url=str(payload["url"]),
@@ -150,6 +241,7 @@ async def complete_direct_upload(
     ingest_job_repository: Any = Depends(get_ingest_job_repository),
     project_repository: Any = Depends(get_project_repository),
     asset_library_repository: Any = Depends(get_asset_library_repository),
+    asset_repository: Any = Depends(get_asset_repository),
     storage_service: OSSStorageService = Depends(get_storage_service),
 ) -> DirectUploadCompleteResponse:
     """确认浏览器直传完成并创建导入任务。"""
@@ -162,16 +254,63 @@ async def complete_direct_upload(
     normalized_key = storage_service._normalize_storage_key(request.storage_key)
     if not normalized_key.startswith("uploads/"):
         raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="Invalid upload key")
-    if not storage_service.file_exists(normalized_key):
+    try:
+        file_exists = storage_service.file_exists(normalized_key)
+    except Exception as error:
+        logger.exception("OSS error checking file existence for key=%s: %s", normalized_key, error)
+        raise HTTPException(
+            status_code=status.HTTP_503_SERVICE_UNAVAILABLE,
+            detail="Storage service unavailable",
+        ) from error
+    if not file_exists:
         raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Uploaded file not found")
 
+    # ── 素材去重检测：同素材库 + 同 file_hash 视为重复 ──
+    if request.file_hash:
+        existing = asset_repository.find_by_library_and_file_hash(
+            library_id=request.library_id,
+            file_hash=request.file_hash,
+        )
+        if existing is not None:
+            logger.info(
+                "素材去重命中: library=%s hash=%s existing_asset=%s",
+                request.library_id,
+                request.file_hash,
+                existing.id,
+            )
+            return DirectUploadCompleteResponse(
+                storage_key=normalized_key,
+                ingest_job_id="",
+                duplicated=True,
+                asset_id=existing.id,
+            )
+
+    # 立即创建 Asset 记录（PROCESSING 状态），使前端刷新后即可看到新素材
+    filename = normalized_key.rsplit("/", 1)[-1]
+    mime_type = _infer_mime_type_from_storage_key(normalized_key)
+    pending_asset = _create_pending_asset(
+        asset_repository=asset_repository,
+        project_id=request.project_id,
+        library_id=request.library_id,
+        storage_key=normalized_key,
+        filename=filename,
+        mime_type=mime_type,
+        user_id=authenticated_user.user.id,
+        file_hash=request.file_hash,
+    )
+
     job = _submit_ingest_job(
         project_id=request.project_id,
         library_id=request.library_id,
         storage_key=normalized_key,
         ingest_job_repository=ingest_job_repository,
+        file_hash=request.file_hash,
+    )
+    return DirectUploadCompleteResponse(
+        storage_key=normalized_key,
+        ingest_job_id=job.id,
+        asset_id=pending_asset.id,
     )
-    return DirectUploadCompleteResponse(storage_key=normalized_key, ingest_job_id=job.id)
 
 
 @router.post(
@@ -181,19 +320,41 @@ async def complete_direct_upload(
     description="上传素材文件（multipart/form-data），支持视频、音频、图片。触发导入流水线自动处理。",
 )
 async def upload_asset(
-    form_data: Annotated[UploadAssetRequest, Form()],
+    project_id: str = Form(..., min_length=1, description="项目 ID"),
+    library_id: str = Form(..., min_length=1, description="素材库 ID"),
     file: UploadFile = File(..., description="要上传的文件（视频、音频、图片等）"),
+    file_hash: str = Form(default="", description="文件 MD5 哈希，用于去重检测"),
     authenticated_user: AuthenticatedUser = Depends(get_current_user),
     ingest_job_repository: Any = Depends(get_ingest_job_repository),
     project_repository: Any = Depends(get_project_repository),
     asset_library_repository: Any = Depends(get_asset_library_repository),
+    asset_repository: Any = Depends(get_asset_repository),
     storage_service: OSSStorageService = Depends(get_storage_service),
 ) -> UploadAssetResponse:
     """上传素材文件并触发导入流水线。"""
-    project_id = form_data.project_id
-    library_id = form_data.library_id
     _require_project_and_library(project_id, library_id, project_repository, asset_library_repository)
 
+    # ── 素材去重检测：上传前检查同素材库 + 同 file_hash ──
+    if file_hash:
+        existing = asset_repository.find_by_library_and_file_hash(
+            library_id=library_id,
+            file_hash=file_hash,
+        )
+        if existing is not None:
+            logger.info(
+                "素材去重命中(multipart): library=%s hash=%s existing_asset=%s",
+                library_id,
+                file_hash,
+                existing.id,
+            )
+            return UploadAssetResponse(
+                storage_key=existing.storage_key,
+                ingest_job_id="",
+                url="",
+                duplicated=True,
+                asset_id=existing.id,
+            )
+
     # P2-5: 服务端验证 MIME 类型
     validated_content_type = _validate_mime_type(file.content_type)
 
@@ -201,10 +362,32 @@ async def upload_asset(
     safe_filename = file.filename.replace("/", "_").replace("\\", "_") if file.filename else "unknown"
     storage_key = f"uploads/{file_id}/{safe_filename}"
 
-    file_url = storage_service.upload_file(
-        file.file,
-        storage_key,
-        content_type=validated_content_type,
+    try:
+        file_url = storage_service.upload_file(
+            file.file,
+            storage_key,
+            content_type=validated_content_type,
+        )
+    except RuntimeError as error:
+        logger.error("OSS not configured for upload: %s", error)
+        raise HTTPException(status_code=status.HTTP_503_SERVICE_UNAVAILABLE, detail=str(error)) from error
+    except Exception as error:
+        logger.exception("Unexpected error uploading file to OSS: %s", error)
+        raise HTTPException(
+            status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,
+            detail=f"Failed to upload file: {type(error).__name__}",
+        ) from error
+
+    # 立即创建 Asset 记录（PROCESSING 状态），使前端刷新后即可看到新素材
+    pending_asset = _create_pending_asset(
+        asset_repository=asset_repository,
+        project_id=project_id,
+        library_id=library_id,
+        storage_key=storage_key,
+        filename=safe_filename,
+        mime_type=validated_content_type,
+        user_id=authenticated_user.user.id,
+        file_hash=file_hash,
     )
 
     job = _submit_ingest_job(
@@ -212,10 +395,12 @@ async def upload_asset(
         library_id=library_id,
         storage_key=storage_key,
         ingest_job_repository=ingest_job_repository,
+        file_hash=file_hash,
     )
 
     return UploadAssetResponse(
         storage_key=storage_key,
         ingest_job_id=job.id,
         url=file_url,
+        asset_id=pending_asset.id,
     )

apps/api/app/api/routes/voice_clones.py

@@ -0,0 +1,227 @@
+"""音色克隆 API 路由。"""
+
+from __future__ import annotations
+
+import logging
+from typing import Optional
+
+from app.auth import AuthenticatedUser, get_current_user
+from app.dependencies import get_cosyvoice_service, get_voice_clone_profile_repository
+from app.schemas.voice_clone import (
+    CreateVoiceCloneRequest,
+    ListVoiceCloneResponse,
+    VoiceCloneProfileResponse,
+    VoiceCloneStatusResponse,
+)
+from fastapi import APIRouter, Depends, HTTPException, Query, Response, status
+
+from packages.adapters.sqlalchemy_impl.voice_clone_profile_repository import (
+    SQLAlchemyVoiceCloneProfileRepository,
+)
+from packages.application.cosyvoice_service import CosyVoiceService
+from packages.application.voice_clone.use_cases import (
+    DeleteVoiceCloneUseCase,
+    GetVoiceCloneStatusUseCase,
+    GetVoiceCloneUseCase,
+    ListVoiceClonesUseCase,
+    VoiceCloneNotFoundError,
+    VoiceCloneNotRetryableError,
+)
+from packages.application.voice_clone.workflow import (
+    VoiceCloneWorkflowError,
+    VoiceCloneWorkflowService,
+)
+
+logger = logging.getLogger(__name__)
+
+router = APIRouter()
+
+
+def _to_response(profile) -> VoiceCloneProfileResponse:
+    return VoiceCloneProfileResponse(
+        id=profile.id,
+        user_id=profile.user_id,
+        name=profile.name,
+        description=profile.description,
+        source_audio_url=profile.source_audio_url,
+        voice_id=profile.voice_id,
+        voice_model=profile.voice_model,
+        language=profile.language,
+        gender=profile.gender,
+        status=profile.status,
+        error_message=profile.error_message,
+        retry_count=profile.retry_count,
+        max_retries=profile.max_retries,
+        metadata=profile.metadata,
+        created_at=profile.created_at,
+        updated_at=profile.updated_at,
+    )
+
+
+def _get_workflow_service(
+    repository: SQLAlchemyVoiceCloneProfileRepository = Depends(get_voice_clone_profile_repository),
+    cosyvoice_service: CosyVoiceService = Depends(get_cosyvoice_service),
+) -> VoiceCloneWorkflowService:
+    return VoiceCloneWorkflowService(repository=repository, cosyvoice_service=cosyvoice_service)
+
+
+@router.post(
+    "",
+    response_model=VoiceCloneProfileResponse,
+    status_code=status.HTTP_201_CREATED,
+)
+def create_voice_clone(
+    request: CreateVoiceCloneRequest,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    workflow: VoiceCloneWorkflowService = Depends(_get_workflow_service),
+) -> VoiceCloneProfileResponse:
+    """创建音色克隆任务。
+
+    创建 VoiceCloneProfile → 提交 CosyVoice 克隆任务 → 触发 Celery 异步轮询。
+    如果有 source_audio_url，状态会变为 processing；否则保持 pending。
+    """
+    user_id = authenticated_user.user.id
+    profile = workflow.start_clone(
+        user_id=user_id,
+        name=request.name,
+        description=request.description,
+        source_audio_url=request.source_audio_url,
+        voice_model=request.voice_model,
+        language=request.language,
+        gender=request.gender,
+        max_retries=request.max_retries,
+        metadata=request.metadata_,
+    )
+
+    # 如果 profile 处于 processing 且有 task_id，触发 Celery 异步轮询
+    task_id = (profile.metadata or {}).get("cosyvoice_task_id", "")
+    if profile.status == "processing" and task_id:
+        try:
+            from worker_app.tasks import process_voice_clone
+
+            process_voice_clone.delay(profile.id)
+            logger.info(f"Celery task dispatched for voice clone {profile.id}")
+        except Exception as e:
+            logger.error(f"Failed to dispatch Celery task: {e}")
+            # P2-3: Celery 调度失败时标记 profile 为 failed，避免永久卡在 processing
+            try:
+                workflow.process_clone_failure(profile.id, f"Celery 任务调度失败: {e}")
+            except Exception as inner_e:
+                logger.error(f"Failed to mark profile as failed after dispatch error: {inner_e}")
+
+    return _to_response(profile)
+
+
+@router.get("", response_model=ListVoiceCloneResponse)
+def list_voice_clones(
+    status_filter: Optional[str] = Query(None, alias="status"),
+    skip: int = Query(0, ge=0),
+    limit: int = Query(50, ge=1, le=200),
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    repository: SQLAlchemyVoiceCloneProfileRepository = Depends(get_voice_clone_profile_repository),
+) -> ListVoiceCloneResponse:
+    """获取用户的音色克隆列表。"""
+    user_id = authenticated_user.user.id
+    use_case = ListVoiceClonesUseCase(repository)
+    items, total = use_case.execute(user_id, status=status_filter, skip=skip, limit=limit)
+    return ListVoiceCloneResponse(
+        items=[_to_response(p) for p in items],
+        total=total,
+    )
+
+
+@router.get("/{clone_id}", response_model=VoiceCloneProfileResponse)
+def get_voice_clone(
+    clone_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    repository: SQLAlchemyVoiceCloneProfileRepository = Depends(get_voice_clone_profile_repository),
+) -> VoiceCloneProfileResponse:
+    """获取音色克隆详情。"""
+    user_id = authenticated_user.user.id
+    use_case = GetVoiceCloneUseCase(repository)
+    try:
+        profile = use_case.execute(clone_id, user_id)
+    except VoiceCloneNotFoundError:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Voice clone not found")
+    return _to_response(profile)
+
+
+@router.get("/{clone_id}/status", response_model=VoiceCloneStatusResponse)
+def get_voice_clone_status(
+    clone_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    repository: SQLAlchemyVoiceCloneProfileRepository = Depends(get_voice_clone_profile_repository),
+) -> VoiceCloneStatusResponse:
+    """查询音色克隆状态（用于前端轮询）。"""
+    user_id = authenticated_user.user.id
+    use_case = GetVoiceCloneStatusUseCase(repository)
+    try:
+        profile = use_case.execute(clone_id, user_id)
+    except VoiceCloneNotFoundError:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Voice clone not found")
+    return VoiceCloneStatusResponse(
+        id=profile.id,
+        status=profile.status,
+        error_message=profile.error_message,
+        voice_id=profile.voice_id,
+        retry_count=profile.retry_count,
+    )
+
+
+@router.delete(
+    "/{clone_id}",
+    status_code=status.HTTP_204_NO_CONTENT,
+    response_model=None,
+)
+def delete_voice_clone(
+    clone_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    repository: SQLAlchemyVoiceCloneProfileRepository = Depends(get_voice_clone_profile_repository),
+) -> Response:
+    """删除音色克隆档案。"""
+    user_id = authenticated_user.user.id
+    use_case = DeleteVoiceCloneUseCase(repository)
+    deleted = use_case.execute(clone_id, user_id)
+    if not deleted:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Voice clone not found")
+    return Response(status_code=204)
+
+
+@router.post("/{clone_id}/retry", response_model=VoiceCloneProfileResponse)
+def retry_voice_clone(
+    clone_id: str,
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    workflow: VoiceCloneWorkflowService = Depends(_get_workflow_service),
+) -> VoiceCloneProfileResponse:
+    """重试失败的音色克隆。
+
+    仅当状态为 failed 时可重试，重试后重新提交 CosyVoice 克隆任务。
+    """
+    user_id = authenticated_user.user.id
+    try:
+        profile = workflow.retry_clone(clone_id, user_id)
+    except VoiceCloneNotFoundError:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Voice clone not found")
+    except VoiceCloneNotRetryableError:
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail="Voice clone is not retryable (only failed clones can be retried)",
+        )
+
+    # 如果 profile 处于 processing 且有 task_id，触发 Celery 异步轮询
+    task_id = (profile.metadata or {}).get("cosyvoice_task_id", "")
+    if profile.status == "processing" and task_id:
+        try:
+            from worker_app.tasks import process_voice_clone
+
+            process_voice_clone.delay(profile.id)
+            logger.info(f"Celery task dispatched for voice clone retry {profile.id}")
+        except Exception as e:
+            logger.error(f"Failed to dispatch Celery task: {e}")
+            # P2-3: Celery 调度失败时标记 profile 为 failed，避免永久卡在 processing
+            try:
+                workflow.process_clone_failure(profile.id, f"Celery 任务调度失败: {e}")
+            except Exception as inner_e:
+                logger.error(f"Failed to mark profile as failed after dispatch error: {inner_e}")
+
+    return _to_response(profile)

apps/api/app/api/routes/voices.py

@@ -1,18 +1,30 @@
-"""Voice library CRUD routes."""
+"""Voice library CRUD routes — Phase 3 增强版.
+
+支持预置音色和克隆音色的统一列表。
+"""
+
 from __future__ import annotations
 
-from typing import Optional
-
-from fastapi import APIRouter, Depends, HTTPException, Query, Response, status
+from typing import Literal, Optional
 
 from app.auth import AuthenticatedUser, get_current_user
 from app.dependencies import get_db_session, get_user_repository
+from app.schemas.voice import (
+    PresetVoiceItemResponse,
+    PresetVoiceListResponse,
+    UnifiedVoiceItemResponse,
+    UnifiedVoiceListResponse,
+)
 from app.schemas.voice_library import (
     CreateVoiceLibraryRequest,
     ListVoiceLibraryResponse,
-    VoiceLibraryItemResponse,
     UpdateVoiceLibraryRequest,
+    VoiceLibraryItemResponse,
 )
+from fastapi import APIRouter, Depends, HTTPException, Query, Response, status
+from sqlalchemy.orm import Session
+
+from packages.adapters.sqlalchemy_impl.voice_clone_profile_repository import SQLAlchemyVoiceCloneProfileRepository
 from packages.adapters.sqlalchemy_impl.voice_library_repository import SQLAlchemyVoiceLibraryRepository
 from packages.application.voice_library.commands import CreateVoiceLibraryCommand, UpdateVoiceLibraryCommand
 from packages.application.voice_library.use_cases import (
@@ -20,12 +32,12 @@ from packages.application.voice_library.use_cases import (
     DeleteVoiceLibraryUseCase,
     GetVoiceLibraryUseCase,
     ListVoiceLibraryUseCase,
-    UpdateVoiceLibraryUseCase,
     NotFoundError,
     QuotaExceededError,
+    UpdateVoiceLibraryUseCase,
 )
+from packages.domain.preset_voices import PRESET_VOICES
 from packages.ports.user_repository import UserRepository
-from sqlalchemy.orm import Session
 
 router = APIRouter()
 
@@ -34,6 +46,10 @@ def _get_voice_repository(session: Session = Depends(get_db_session)) -> SQLAlch
     return SQLAlchemyVoiceLibraryRepository(session)
 
 
+def _get_clone_profile_repository(session: Session = Depends(get_db_session)) -> SQLAlchemyVoiceCloneProfileRepository:
+    return SQLAlchemyVoiceCloneProfileRepository(session)
+
+
 def _to_response(item) -> VoiceLibraryItemResponse:
     return VoiceLibraryItemResponse(
         id=item.id,
@@ -54,6 +70,54 @@ def _to_response(item) -> VoiceLibraryItemResponse:
     )
 
 
+def _to_unified_response(item, profile_id_map: dict | None = None) -> UnifiedVoiceItemResponse:
+    """将数据库音色转换为统一响应格式。
+
+    Args:
+        item: VoiceLibraryItem
+        profile_id_map: voice_id → profile_id 映射，用于填充 voice_clone_profile_id
+    """
+    profile_id = None
+    if profile_id_map and item.voice_id:
+        profile_id = profile_id_map.get(item.voice_id)
+    return UnifiedVoiceItemResponse(
+        id=item.id,
+        type="clone",
+        name=item.name,
+        description=item.text,
+        gender="unknown",
+        language="zh-CN",
+        voice_id=item.voice_id,
+        voice_provider=item.voice_provider or "cosyvoice",
+        audio_url=item.audio_url,
+        duration=item.duration,
+        file_size=item.file_size,
+        status=item.status,
+        tags=item.tags,
+        user_id=item.user_id,
+        project_id=item.project_id,
+        voice_clone_profile_id=profile_id,
+        created_at=item.created_at,
+        updated_at=item.updated_at,
+    )
+
+
+def _preset_to_unified_response(preset) -> UnifiedVoiceItemResponse:
+    """将预置音色转换为统一响应格式。"""
+    return UnifiedVoiceItemResponse(
+        id=preset.voice_id,
+        type="preset",
+        name=preset.name,
+        description=preset.description,
+        gender=preset.gender,
+        language=preset.language,
+        voice_id=preset.voice_id,
+        voice_provider="cosyvoice",
+        preview_url=preset.preview_url,
+        tags=preset.tags or [],
+    )
+
+
 def _get_user_plan(user_id: str, user_repository: UserRepository) -> str:
     user = user_repository.find_by_id(user_id)
     if user is None:
@@ -61,18 +125,113 @@ def _get_user_plan(user_id: str, user_repository: UserRepository) -> str:
     return getattr(user, "subscription_plan", "free") or "free"
 
 
-@router.get("", response_model=ListVoiceLibraryResponse)
-def list_voices(
+# ==================== 统一配音列表（预置 + 克隆）====================
+
+
+@router.get("", response_model=UnifiedVoiceListResponse)
+def list_voices_unified(
+    type: Optional[Literal["preset", "clone"]] = Query(
+        None,
+        description="音色类型过滤：preset=仅预置，clone=仅克隆，不传=全部",
+    ),
+    status_filter: Optional[str] = Query(None, alias="status"),
+    skip: int = Query(0, ge=0),
+    limit: int = Query(50, ge=1, le=200),
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    voice_repository: SQLAlchemyVoiceLibraryRepository = Depends(_get_voice_repository),
+    clone_profile_repository: SQLAlchemyVoiceCloneProfileRepository = Depends(_get_clone_profile_repository),
+) -> UnifiedVoiceListResponse:
+    """获取配音列表（预置音色 + 用户克隆音色）。
+
+    - 不传 type：返回预置音色 + 用户克隆音色，预置音色在前
+    - type=preset：仅返回预置音色
+    - type=clone：仅返回用户克隆音色
+    """
+    user_id = authenticated_user.user.id
+    items: list[UnifiedVoiceItemResponse] = []
+    preset_count = 0
+    clone_count = 0
+
+    has_preset = type is None or type == "preset"
+    has_clone = type is None or type == "clone"
+
+    # 获取预置音色
+    if has_preset:
+        preset_items = [_preset_to_unified_response(p) for p in PRESET_VOICES]
+        preset_count = len(preset_items)
+
+    # 获取克隆音色
+    if has_clone:
+        use_case = ListVoiceLibraryUseCase(voice_repository)
+        clone_items_raw, clone_count = use_case.execute(user_id, status=status_filter, skip=skip, limit=limit)
+        # 批量查询 voice_id → profile_id 映射，填充 voice_clone_profile_id
+        voice_ids = [i.voice_id for i in clone_items_raw if i.voice_id]
+        profile_id_map = clone_profile_repository.find_profile_ids_by_voice_ids(voice_ids) if voice_ids else {}
+        clone_items = [_to_unified_response(i, profile_id_map) for i in clone_items_raw]
+
+    # 组装结果
+    if type == "preset":
+        items = preset_items[skip : skip + limit]
+        total = preset_count
+    elif type == "clone":
+        items = clone_items
+        total = clone_count
+    else:
+        # 全量模式：预置在前，克隆补位
+        all_items = preset_items + clone_items
+        total = preset_count + clone_count
+        items = all_items[skip : skip + limit]
+
+    return UnifiedVoiceListResponse(
+        items=items,
+        total=total,
+        preset_count=preset_count if has_preset else 0,
+        clone_count=clone_count if has_clone else 0,
+    )
+
+
+# ==================== 预置音色专用端点 ====================
+
+
+@router.get("/presets", response_model=PresetVoiceListResponse)
+def list_preset_voices() -> PresetVoiceListResponse:
+    """获取预置音色列表。
+
+    不需要认证，返回所有系统预置的 CosyVoice 音色。
+    """
+    items = [
+        PresetVoiceItemResponse(
+            voice_id=p.voice_id,
+            name=p.name,
+            description=p.description,
+            gender=p.gender,
+            language=p.language,
+            preview_url=p.preview_url,
+            tags=p.tags or [],
+        )
+        for p in PRESET_VOICES
+    ]
+    return PresetVoiceListResponse(items=items, total=len(items))
+
+
+# ==================== 原有 CRUD 端点（保持向后兼容）====================
+
+
+@router.get("/legacy", response_model=ListVoiceLibraryResponse)
+def list_voices_legacy(
     status_filter: Optional[str] = Query(None, alias="status"),
     skip: int = Query(0, ge=0),
     limit: int = Query(50, ge=1, le=200),
     authenticated_user: AuthenticatedUser = Depends(get_current_user),
     voice_repository: SQLAlchemyVoiceLibraryRepository = Depends(_get_voice_repository),
 ) -> ListVoiceLibraryResponse:
+    """原有配音列表接口（仅返回用户克隆音色）。
+
+    保留用于向后兼容，新客户端请使用 GET /api/v1/voices。
+    """
     user_id = authenticated_user.user.id
     use_case = ListVoiceLibraryUseCase(voice_repository)
-    items = use_case.execute(user_id, status=status_filter, skip=skip, limit=limit)
-    total = voice_repository.count_by_user(user_id)
+    items, total = use_case.execute(user_id, status=status_filter, skip=skip, limit=limit)
     return ListVoiceLibraryResponse(
         items=[_to_response(i) for i in items],
         total=total,
@@ -157,7 +316,7 @@ def update_voice(
     return _to_response(item)
 
 
-@router.delete("/{voice_id}", status_code=status.HTTP_204_NO_CONTENT, response_class=Response)
+@router.delete("/{voice_id}", status_code=status.HTTP_204_NO_CONTENT, response_model=None)
 def delete_voice(
     voice_id: str,
     authenticated_user: AuthenticatedUser = Depends(get_current_user),

apps/api/app/auth.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import hashlib
 from dataclasses import dataclass
 
 import jwt
@@ -23,6 +24,35 @@ class AuthenticatedUser:
     token_type: str | None = None
 
 
+def _get_redis_client():
+    """获取 Redis 客户端用于 JWT 黑名单"""
+    import redis as redis_lib
+
+    return redis_lib.from_url(settings.REDIS_URL, decode_responses=True)
+
+
+def _token_fingerprint(token: str) -> str:
+    """计算 token 的哈希指纹"""
+    return hashlib.sha256(token.encode()).hexdigest()
+
+
+def blacklist_token(token: str, exp: int) -> None:
+    """将 token 加入黑名单，TTL 为 token 剩余有效期"""
+    import time
+
+    redis_client = _get_redis_client()
+    key = f"jwt:blacklist:{_token_fingerprint(token)}"
+    ttl = max(exp - int(time.time()), 1)
+    redis_client.setex(key, ttl, "revoked")
+
+
+def is_token_blacklisted(token: str) -> bool:
+    """检查 token 是否在黑名单中"""
+    redis_client = _get_redis_client()
+    key = f"jwt:blacklist:{_token_fingerprint(token)}"
+    return redis_client.exists(key) > 0
+
+
 async def get_current_user(
     credentials: HTTPAuthorizationCredentials | None = Depends(bearer_scheme),
     user_repository: UserRepository = Depends(get_user_repository),
@@ -56,6 +86,11 @@ def _decode_user_token(token: str) -> dict:
 
     if payload.get("type") not in {"user_auth", "access"}:
         raise _unauthorized("Invalid token type")
+
+    # 检查 token 是否在黑名单中
+    if is_token_blacklisted(token):
+        raise _unauthorized("Token has been revoked")
+
     return payload
 
 

apps/api/app/config.py

@@ -1,7 +1,7 @@
 import os
 from typing import Optional
 
-from pydantic import field_validator
+from pydantic import AliasChoices, Field, field_validator
 from pydantic_settings import BaseSettings, SettingsConfigDict
 
 
@@ -21,9 +21,7 @@ class Settings(BaseSettings):
     API_PORT: int = 8000
     API_PREFIX: str = "/api/v1"
 
-    DATABASE_URL: str = (
-        "postgresql+psycopg://postgres:postgres@localhost:5432/xiaoxia_saas"
-    )
+    DATABASE_URL: str = "postgresql+psycopg://postgres:postgres@localhost:5432/xiaoxia_saas"
     DATABASE_POOL_SIZE: int = 20
     DATABASE_MAX_OVERFLOW: int = 10  # 调整为合理值：pool_size(20) + max_overflow(10) = 最大30连接
     DATABASE_POOL_TIMEOUT: int = 30
@@ -48,8 +46,7 @@ class Settings(BaseSettings):
     def validate_jwt_secret_key(cls, v):
         if v is None or v == "":
             raise ValueError(
-                "JWT_SECRET_KEY must be set via environment variable. "
-                "Do not use default value in production!"
+                "JWT_SECRET_KEY must be set via environment variable. " "Do not use default value in production!"
             )
         # Block known insecure default values
         insecure_defaults = [
@@ -61,8 +58,7 @@ class Settings(BaseSettings):
         ]
         if v.lower() in [d.lower() for d in insecure_defaults]:
             raise ValueError(
-                f"JWT_SECRET_KEY '{v}' is insecure. "
-                "Please set a strong random secret via environment variable."
+                f"JWT_SECRET_KEY '{v}' is insecure. " "Please set a strong random secret via environment variable."
             )
         return v
 
@@ -83,13 +79,14 @@ class Settings(BaseSettings):
     OSS_ACCESS_KEY_ID: str = ""
     OSS_ACCESS_KEY_SECRET: str = ""
     OSS_BUCKET_NAME: str = "xiaoxia-autocut"
-    OSS_DIRECT_UPLOAD_MAX_MB: int = 800
-    OSS_DIRECT_UPLOAD_EXPRESS_SECRET: int = 900
+    OSS_DIRECT_UPLOAD_MAX_MB: int = Field(
+        default=2000,
+        validation_alias=AliasChoices("OSS_DIRECT_UPLOAD_MAX_MB", "MAX_UPLOAD_SIZE_MB"),
+    )
+    OSS_DIRECT_UPLOAD_EXPIRE_SECONDS: int = 900
 
     LOG_LEVEL: str = "INFO"
-    CORS_ORIGINS_RAW: str = (
-        "http://localhost:3000,http://localhost:5173,http://localhost:8000"
-    )
+    CORS_ORIGINS_RAW: str = "http://localhost:3000,http://localhost:5173,http://localhost:8000"
 
     model_config = SettingsConfigDict(
         env_file=".env",
@@ -100,11 +97,7 @@ class Settings(BaseSettings):
 
     @property
     def CORS_ORIGINS(self) -> list[str]:
-        return [
-            origin.strip()
-            for origin in self.CORS_ORIGINS_RAW.split(",")
-            if origin.strip()
-        ]
+        return [origin.strip() for origin in self.CORS_ORIGINS_RAW.split(",") if origin.strip()]
 
     @property
     def database_url(self) -> str:

apps/api/app/core/database.py

@@ -1,49 +0,0 @@
-"""Database session management and engine configuration.
-
-统一使用 app.config 中的数据库配置，移除重复的 DatabaseSettings。
-"""
-from sqlalchemy import create_engine
-from sqlalchemy.orm import sessionmaker, Session
-from contextlib import contextmanager
-from typing import Generator
-
-from app.config import settings
-
-
-engine = create_engine(
-    settings.database_url,
-    pool_size=settings.DATABASE_POOL_SIZE,
-    max_overflow=settings.DATABASE_MAX_OVERFLOW,
-    pool_timeout=settings.DATABASE_POOL_TIMEOUT,
-    pool_recycle=settings.DATABASE_POOL_RECYCLE,
-)
-
-SessionLocal = sessionmaker(autocommit=False, autoflush=False, bind=engine)
-
-
-def get_db() -> Generator[Session, None, None]:
-    """Dependency for getting database sessions."""
-    db = SessionLocal()
-    try:
-        yield db
-    finally:
-        db.close()
-
-
-@contextmanager
-def get_db_context() -> Generator[Session, None, None]:
-    """Context manager for database sessions.
-    
-    Usage:
-        with get_db_context() as db:
-            db.query(Model).all()
-    """
-    db = SessionLocal()
-    try:
-        yield db
-        db.commit()
-    except Exception:
-        db.rollback()
-        raise
-    finally:
-        db.close()

apps/api/app/core/storage.py

@@ -28,17 +28,38 @@ class OSSStorageService:
         self.local_url_prefix = os.getenv("GENERATED_FILES_URL_PREFIX", "/generated-files")
         self.bucket = None
 
-        if settings.OSS_ACCESS_KEY_ID and settings.OSS_ACCESS_KEY_SECRET:
+        has_key_id = bool(settings.OSS_ACCESS_KEY_ID)
+        has_key_secret = bool(settings.OSS_ACCESS_KEY_SECRET)
+
+        if has_key_id and has_key_secret:
             if oss2 is not None:
-                auth = oss2.Auth(
-                    settings.OSS_ACCESS_KEY_ID,
-                    settings.OSS_ACCESS_KEY_SECRET,
-                )
-                self.bucket = oss2.Bucket(
-                    auth,
-                    settings.OSS_ENDPOINT,
-                    settings.OSS_BUCKET_NAME,
-                )
+                try:
+                    auth = oss2.Auth(
+                        settings.OSS_ACCESS_KEY_ID,
+                        settings.OSS_ACCESS_KEY_SECRET,
+                    )
+                    self.bucket = oss2.Bucket(
+                        auth,
+                        settings.OSS_ENDPOINT,
+                        settings.OSS_BUCKET_NAME,
+                    )
+                    logger.info(
+                        "OSS initialized: endpoint=%s bucket=%s",
+                        settings.OSS_ENDPOINT,
+                        settings.OSS_BUCKET_NAME,
+                    )
+                except Exception as error:
+                    logger.error("Failed to initialize OSS bucket client: %s", error)
+            else:
+                logger.error("oss2 SDK is not installed — OSS operations will fail")
+        else:
+            missing = []
+            if not has_key_id:
+                missing.append("OSS_ACCESS_KEY_ID")
+            if not has_key_secret:
+                missing.append("OSS_ACCESS_KEY_SECRET")
+            logger.error("OSS credentials not configured — missing: %s", ", ".join(missing))
+
         self.access_key_id = settings.OSS_ACCESS_KEY_ID
         self.access_key_secret = settings.OSS_ACCESS_KEY_SECRET
         self.endpoint = settings.OSS_ENDPOINT
@@ -211,8 +232,6 @@ class OSSStorageService:
         return self.bucket.object_exists(storage_key)
 
 
-
-
 _storage_service = None
 
 
@@ -222,4 +241,3 @@ def get_storage_service() -> OSSStorageService:
     if _storage_service is None:
         _storage_service = OSSStorageService()
     return _storage_service
-

apps/api/app/db.py

@@ -1,7 +1,4 @@
-from collections.abc import Generator
-
 from app.config import settings
-from sqlalchemy.orm import Session
 
 from packages.adapters.sqlalchemy_impl import (
     build_session_factory,
@@ -21,11 +18,3 @@ engine, SessionLocal = build_session_factory(
 assert_auto_create_schema_allowed(settings.ENVIRONMENT, settings.AUTO_CREATE_SCHEMA)
 if settings.AUTO_CREATE_SCHEMA:
     initialize_database(engine)
-
-
-def get_db() -> Generator[Session, None, None]:
-    db = SessionLocal()
-    try:
-        yield db
-    finally:
-        db.close()

apps/api/app/dependencies.py

@@ -3,17 +3,18 @@
 All repository and service factories are defined here as FastAPI dependencies,
 ensuring proper lifecycle management and testability.
 """
+
 from __future__ import annotations
 
-import redis
 from typing import Generator
 
+import redis
 from app.config import settings
 from fastapi import Depends
 from sqlalchemy.orm import Session
 
 from packages.adapters.redis import NoopSessionStore, SessionStore
-from packages.adapters.smtp import EmailConfig, NoopEmailService, get_email_service
+from packages.adapters.smtp import EmailConfig, EmailService, NoopEmailService, get_email_service
 from packages.adapters.sqlalchemy_impl.asset_library_repository import (
     SQLAlchemyAssetLibraryRepository,
 )
@@ -30,31 +31,39 @@ from packages.adapters.sqlalchemy_impl.generated_video_repository import (
 from packages.adapters.sqlalchemy_impl.generation_task_repository import (
     SQLAlchemyGenerationTaskRepository,
 )
-from packages.adapters.sqlalchemy_impl.title_library_repository import (
-    SQLAlchemyTitleLibraryRepository,
-)
-from packages.adapters.sqlalchemy_impl.voice_library_repository import (
-    SQLAlchemyVoiceLibraryRepository,
-)
 from packages.adapters.sqlalchemy_impl.ingest_job_repository import (
     SQLAlchemyIngestJobRepository,
 )
+from packages.adapters.sqlalchemy_impl.job_repository import SQLAlchemyJobRepository
 from packages.adapters.sqlalchemy_impl.project_repository import (
     SQLAlchemyProjectRepository,
 )
 from packages.adapters.sqlalchemy_impl.session import build_session_factory
+from packages.adapters.sqlalchemy_impl.tag_repository import SQLAlchemyTagRepository
+from packages.adapters.sqlalchemy_impl.title_library_repository import (
+    SQLAlchemyTitleLibraryRepository,
+)
 from packages.adapters.sqlalchemy_impl.user_repository import SQLAlchemyUserRepository
-from packages.ports.asset_repository import AssetRepository
+from packages.adapters.sqlalchemy_impl.voice_clone_profile_repository import (
+    SQLAlchemyVoiceCloneProfileRepository,
+)
+from packages.adapters.sqlalchemy_impl.voice_library_repository import (
+    SQLAlchemyVoiceLibraryRepository,
+)
 from packages.ports.asset_library_repository import AssetLibraryRepository
-from packages.ports.user_repository import UserRepository
+from packages.ports.asset_repository import AssetRepository
 from packages.ports.classification_job_repository import ClassificationJobRepository
 from packages.ports.duplication_repository import DuplicationRecordRepository
-from packages.ports.generation_task_repository import GenerationTaskRepository
-from packages.ports.title_library_repository import TitleLibraryRepository
-from packages.ports.voice_library_repository import VoiceLibraryRepository
 from packages.ports.generated_video_repository import GeneratedVideoRepository
+from packages.ports.generation_task_repository import GenerationTaskRepository
 from packages.ports.ingest_job_repository import IngestJobRepository
+from packages.ports.job_repository import JobRepository
 from packages.ports.project_repository import ProjectRepository
+from packages.ports.tag_repository import TagRepository
+from packages.ports.title_library_repository import TitleLibraryRepository
+from packages.ports.user_repository import UserRepository
+from packages.ports.voice_clone_profile_repository import VoiceCloneProfileRepository
+from packages.ports.voice_library_repository import VoiceLibraryRepository
 
 _engine, _SessionLocal = build_session_factory(settings.DATABASE_URL)
 
@@ -103,6 +112,13 @@ def get_generation_task_repository(
     return SQLAlchemyGenerationTaskRepository(session)
 
 
+def get_job_repository(
+    session: Session = Depends(get_db_session),
+) -> SQLAlchemyJobRepository:
+    """Provide the SQLAlchemy job repository implementation."""
+    return SQLAlchemyJobRepository(session)
+
+
 def get_generated_video_repository(
     session: Session = Depends(get_db_session),
 ) -> SQLAlchemyGeneratedVideoRepository:
@@ -124,6 +140,12 @@ def get_project_repository(
     return SQLAlchemyProjectRepository(session)
 
 
+def get_tag_repository(
+    session: Session = Depends(get_db_session),
+) -> TagRepository:
+    """Provide the SQLAlchemy tag repository implementation."""
+    return SQLAlchemyTagRepository(session)
+
 
 def get_user_repository(
     session: Session = Depends(get_db_session),
@@ -156,6 +178,7 @@ def get_auth_email_service() -> NoopEmailService | EmailService:
         enabled=True,
     )
 
+
 def get_title_library_repository(
     session: Session = Depends(get_db_session),
 ) -> SQLAlchemyTitleLibraryRepository:
@@ -168,3 +191,17 @@ def get_voice_library_repository(
 ) -> SQLAlchemyVoiceLibraryRepository:
     """Provide the SQLAlchemy voice library repository implementation."""
     return SQLAlchemyVoiceLibraryRepository(session)
+
+
+def get_voice_clone_profile_repository(
+    session: Session = Depends(get_db_session),
+) -> SQLAlchemyVoiceCloneProfileRepository:
+    """Provide the SQLAlchemy voice clone profile repository implementation."""
+    return SQLAlchemyVoiceCloneProfileRepository(session)
+
+
+def get_cosyvoice_service():
+    """Provide the CosyVoice service instance."""
+    from packages.application.cosyvoice_service import CosyVoiceService
+
+    return CosyVoiceService()

apps/api/app/middleware/auth.py

@@ -1,8 +1,8 @@
 """
 Authentication dependency compatibility layer.
 
-Canonical bearer-token parsing lives in app.auth. This module remains only so
-legacy imports have a safe target while workspace dependencies are rebuilt.
+Canonical bearer-token parsing lives in app.auth. This module re-exports
+common auth dependencies for backward compatibility.
 """
 
 from __future__ import annotations
@@ -10,7 +10,7 @@ from __future__ import annotations
 from app.auth import AuthenticatedUser
 from app.auth import get_current_user as get_authenticated_user
 from app.dependencies import get_user_repository
-from fastapi import Depends
+from fastapi import Depends, HTTPException
 from fastapi.security import HTTPAuthorizationCredentials, HTTPBearer
 
 from packages.domain.entities import User

apps/api/app/middleware/exceptions.py

@@ -3,7 +3,6 @@
 """
 
 import logging
-import traceback
 
 from fastapi import Request, status
 from fastapi.exceptions import RequestValidationError

apps/api/app/middleware/logging.py

@@ -69,9 +69,7 @@ class RequestLoggingMiddleware(BaseHTTPMiddleware):
 
         # 记录请求信息（不包含敏感参数）
         if safe_query:
-            logger.info(
-                f"Request: {request.method} {request.url.path}?{safe_query}"  # noqa: E501
-            )
+            logger.info(f"Request: {request.method} {request.url.path}?{safe_query}")  # noqa: E501
         else:
             logger.info(f"Request: {request.method} {request.url.path}")
 
@@ -83,8 +81,7 @@ class RequestLoggingMiddleware(BaseHTTPMiddleware):
 
         # 记录响应信息
         logger.info(
-            f"Response: {request.method} {request.url.path} "
-            f"status={response.status_code} time={process_time:.3f}s"
+            f"Response: {request.method} {request.url.path} " f"status={response.status_code} time={process_time:.3f}s"
         )
 
         # 添加处理时间到响应头
@@ -94,15 +91,27 @@ class RequestLoggingMiddleware(BaseHTTPMiddleware):
 
 
 class RateLimitMiddleware(BaseHTTPMiddleware):
-    """基于 IP 的简单限流中间件"""
+    """基于 IP 的简单限流中间件
 
-    def __init__(self, app, max_requests: int = 100, window_seconds: int = 60):
+    Args:
+        app: ASGI 应用
+        max_requests: 窗口期内最大请求数
+        window_seconds: 时间窗口（秒）
+        paths: 限流的路径列表，None 表示所有路径
+    """
+
+    def __init__(self, app, max_requests: int = 100, window_seconds: int = 60, paths: list[str] | None = None):
         super().__init__(app)
         self.max_requests = max_requests
         self.window_seconds = window_seconds
+        self.paths = set(paths) if paths else None
         self.requests = {}  # {ip: [timestamps]}
 
     async def dispatch(self, request: Request, call_next):
+        # 如果配置了路径过滤，只对指定路径限流
+        if self.paths is not None and request.url.path not in self.paths:
+            return await call_next(request)
+
         # 获取客户端 IP
         client_ip = request.client.host
 
@@ -111,9 +120,7 @@ class RateLimitMiddleware(BaseHTTPMiddleware):
         # 清理过期记录
         if client_ip in self.requests:
             self.requests[client_ip] = [
-                ts
-                for ts in self.requests[client_ip]
-                if current_time - ts < self.window_seconds
+                ts for ts in self.requests[client_ip] if current_time - ts < self.window_seconds
             ]
 
         # 计算请求次数
@@ -128,8 +135,7 @@ class RateLimitMiddleware(BaseHTTPMiddleware):
                     "error": {
                         "code": "RATE_LIMIT_EXCEEDED",
                         "message": (  # noqa: E501
-                            f"Too many requests. Limit: "
-                            f"{self.max_requests} per {self.window_seconds}s"
+                            f"Too many requests. Limit: " f"{self.max_requests} per {self.window_seconds}s"
                         ),
                     }
                 },
@@ -145,8 +151,6 @@ class RateLimitMiddleware(BaseHTTPMiddleware):
 
         # 添加限流信息到响应头
         response.headers["X-RateLimit-Limit"] = str(self.max_requests)
-        response.headers["X-RateLimit-Remaining"] = str(
-            self.max_requests - len(self.requests[client_ip])
-        )
+        response.headers["X-RateLimit-Remaining"] = str(self.max_requests - len(self.requests[client_ip]))
 
         return response

apps/api/app/middleware/prometheus_metrics.py

@@ -14,17 +14,16 @@ from typing import Callable
 
 from fastapi import Request, Response
 from prometheus_client import (
+    CONTENT_TYPE_LATEST,
+    REGISTRY,
     Counter,
     Gauge,
     Histogram,
     generate_latest,
-    CONTENT_TYPE_LATEST,
-    REGISTRY,
 )
 from starlette.middleware.base import BaseHTTPMiddleware
 from starlette.responses import PlainTextResponse
 
-
 # Buckets for HTTP request duration (seconds)
 HTTP_DURATION_BUCKETS = (0.005, 0.01, 0.025, 0.05, 0.1, 0.25, 0.5, 1.0, 2.5, 5.0, 10.0)
 
@@ -122,8 +121,19 @@ class PrometheusMetricsMiddleware(BaseHTTPMiddleware):
 
 
 async def metrics_endpoint(request: Request) -> PlainTextResponse:
-    """FastAPI endpoint that returns Prometheus metrics in text format."""
+    """FastAPI endpoint that returns Prometheus metrics in text format.
+
+    需要 Bearer Token 认证，Token 通过 METRICS_AUTH_TOKEN 环境变量配置。
+    """
     import os
+
+    # Bearer Token 认证
+    auth_token = os.getenv("METRICS_AUTH_TOKEN", "")
+    if auth_token:
+        auth_header = request.headers.get("Authorization", "")
+        if not auth_header.startswith("Bearer ") or auth_header[7:] != auth_token:
+            return PlainTextResponse(content="Unauthorized", status_code=401)
+
     version = os.getenv("APP_VERSION", "unknown")
     environment = os.getenv("APP_ENV", "unknown")
     APP_INFO.labels(version=version, environment=environment).set(1)

apps/api/app/middleware/versioning.py

@@ -2,8 +2,6 @@
 API 版本管理中间件
 """
 
-from datetime import datetime
-
 from fastapi import Request
 from starlette.middleware.base import BaseHTTPMiddleware
 

apps/api/app/schemas/asset.py

@@ -25,6 +25,12 @@ class UpdateAssetReviewRequest(BaseModel):
     review_status: str = Field(..., pattern="^(pending_review|approved|rejected)$")
 
 
+class UpdateAssetRequest(BaseModel):
+    name: str | None = Field(default=None, min_length=1, max_length=100)
+    metadata: dict[str, object] | None = None
+    tags: list[str] | None = None
+
+
 class AssetResponse(BaseModel):
     id: str
     project_id: str
@@ -34,6 +40,7 @@ class AssetResponse(BaseModel):
     mime_type: str
     metadata: dict[str, object]
     file_size: int
+    file_url: str | None = None
     thumbnail_url: str | None = None
     duration: float | None = None
     width: int | None = None
@@ -44,7 +51,24 @@ class AssetResponse(BaseModel):
     classification_status: str
     quality_score: float | None = None
     uploaded_by_user_id: str
+    tag_ids: list[str] = Field(default_factory=list)
+
+
+class BatchDeleteRequest(BaseModel):
+    """批量删除请求。"""
+
+    ids: list[str] = Field(..., min_length=1, max_length=100, description="要删除的素材 ID 列表")
+
+
+class BatchDeleteResponse(BaseModel):
+    """批量删除响应。"""
+
+    deleted_count: int = Field(..., ge=0, description="实际删除数量")
+    failed_ids: list[str] = Field(default_factory=list, description="删除失败的 ID 列表")
 
 
 class ListAssetsResponse(BaseModel):
     items: list[AssetResponse]
+    total: int = Field(default=0, ge=0)
+    skip: int = Field(default=0, ge=0)
+    limit: int = Field(default=100, ge=1)

apps/api/app/schemas/asset_library.py

@@ -18,3 +18,8 @@ class AssetLibraryResponse(BaseModel):
 
 class ListAssetLibrariesResponse(BaseModel):
     items: list[AssetLibraryResponse]
+
+
+class EnsureDefaultLibraryRequest(BaseModel):
+    project_id: str = Field(..., min_length=1)
+    kind: str = Field(..., pattern="^(video|voice|image)$")

apps/api/app/schemas/chunked_upload.py

@@ -7,7 +7,9 @@ class ChunkedUploadInitRequest(BaseModel):
     filename: str = Field(..., min_length=1, max_length=255, description="Filename")
     file_size: int = Field(..., gt=0, le=2147483648, description="File size in bytes, max 2GB")
     total_chunks: int = Field(..., gt=0, description="Total number of chunks")
-    content_type: str = Field(default="application/octet-stream", min_length=1, max_length=100, description="Content type")
+    content_type: str = Field(
+        default="application/octet-stream", min_length=1, max_length=100, description="Content type"
+    )
     project_id: str = Field(..., min_length=1, description="Project ID")
     library_id: str = Field(..., min_length=1, description="Asset library ID")
 
@@ -34,9 +36,12 @@ class ChunkedUploadStatusResponse(BaseModel):
 class ChunkedUploadCompleteRequest(BaseModel):
     project_id: str = Field(..., min_length=1, description="Project ID")
     library_id: str = Field(..., min_length=1, description="Asset library ID")
+    file_hash: str = Field(default="", max_length=64, description="文件 MD5 哈希，用于去重检测")
 
 
 class ChunkedUploadCompleteResponse(BaseModel):
     storage_key: str = Field(..., description="Storage key")
     ingest_job_id: str = Field(..., description="Ingest job ID")
     url: str = Field(..., description="File URL")
+    duplicated: bool = Field(default=False, description="是否为重复素材（命中去重）")
+    asset_id: str = Field(default="", description="重复素材的 asset_id（duplicated=true 时返回）")

apps/api/app/schemas/dashboard.py

@@ -0,0 +1,32 @@
+from datetime import datetime
+
+from pydantic import BaseModel, Field
+
+
+class RecentTaskItem(BaseModel):
+    id: str
+    task_type: str = "generation"
+    status: str
+    current_step: str = ""
+    error_message: str = ""
+    updated_at: datetime | None = None
+
+
+class SubscriptionInfo(BaseModel):
+    """用户订阅信息。"""
+
+    plan: str = "free"
+    is_active: bool = False
+
+
+class DashboardOverviewResponse(BaseModel):
+    """Dashboard 概览数据。"""
+
+    total_assets: int = 0
+    used_storage_bytes: int = 0
+    total_titles: int = 0
+    total_voices: int = 0
+    total_tasks: int = 0
+    total_products: int = 0
+    subscription: SubscriptionInfo = Field(default_factory=SubscriptionInfo)
+    recent_tasks: list[RecentTaskItem] = Field(default_factory=list)

apps/api/app/schemas/duplication.py

@@ -1,4 +1,5 @@
 """查重 API Pydantic schemas。"""
+
 from __future__ import annotations
 
 from pydantic import BaseModel, Field

apps/api/app/schemas/generation_task.py

@@ -1,12 +1,64 @@
-from pydantic import BaseModel, Field
+from pydantic import BaseModel, Field, model_validator
+
+
+class ConfirmGenerationRequest(BaseModel):
+    """确认生成请求体 — 基于预览任务创建正式生成任务"""
+
+    output_width: int = Field(default=1080, description="输出视频宽度")
+    output_height: int = Field(default=1920, description="输出视频高度")
+    cover_url: str = Field(default="", description="自定义封面图片 URL")
+    custom_title: str = Field(default="", description="自定义视频标题")
 
 
 class CreateGenerationTaskRequest(BaseModel):
-    project_id: str = Field(..., min_length=1)
-    asset_library_id: str = Field(..., min_length=1)
+    """创建生成任务请求。
+
+    支持两种模式（至少提供一种）：
+    - 项目模式：project_id + asset_library_id（向后兼容）
+    - 模板模式：template_id + asset_ids / title_ids / voice_ids
+    """
+
+    project_id: str = ""
+    asset_library_id: str = ""
     strategy_id: str = ""
     voice_library_id: str = ""
     created_by_user_id: str = ""
+    # ── 模板模式新增字段 ──
+    template_id: str = ""
+    asset_ids: list[str] = Field(default_factory=list)
+    title_ids: list[str] = Field(default_factory=list)
+    voice_ids: list[str] = Field(default_factory=list)
+    # ── 来源剪辑计划 ──
+    source_edit_plan_id: str = ""
+    # ── 批量生成 ──
+    count: int = Field(default=1, ge=1, le=50, description="批量生成数量，默认1，最大50")
+    # ── 素材库自动匹配 ──
+    asset_select_mode: str = Field(
+        default="all",
+        description="素材选取模式：all=全部ready视频, random=随机选取, smart=智能匹配（按质量/时长评分）",
+    )
+    asset_select_count: int = Field(
+        default=0, ge=0, le=100, description="选取数量，0表示全部（仅 random/smart 模式有效）"
+    )
+    # ── 预览 / 确认生成 ──
+    is_preview: bool = Field(default=True, description="是否为预览任务")
+    source_task_id: str = Field(default="", description="来源预览任务 ID（确认生成时传入）")
+    output_width: int = Field(default=1280, description="输出视频宽度")
+    output_height: int = Field(default=720, description="输出视频高度")
+    cover_url: str = Field(default="", description="封面图片 URL")
+    custom_title: str = Field(default="", description="自定义视频标题")
+
+    @model_validator(mode="after")
+    def _check_at_least_one_mode(self) -> "CreateGenerationTaskRequest":
+        has_project = bool(self.project_id.strip())
+        has_template = bool(self.template_id.strip())
+        if not has_project and not has_template:
+            raise ValueError("project_id 或 template_id 至少需要提供一个")
+        has_library = bool(self.asset_library_id.strip())
+        has_assets = bool(self.asset_ids or self.title_ids or self.voice_ids)
+        if not has_library and not has_assets:
+            raise ValueError("asset_library_id 或 asset_ids/title_ids/voice_ids 至少需要提供一个")
+        return self
 
 
 class GenerationTaskResponse(BaseModel):
@@ -15,7 +67,33 @@ class GenerationTaskResponse(BaseModel):
     asset_library_id: str
     strategy_id: str
     voice_library_id: str
+    template_id: str = ""
+    asset_ids: list[str] = Field(default_factory=list)
+    title_ids: list[str] = Field(default_factory=list)
+    voice_ids: list[str] = Field(default_factory=list)
+    source_edit_plan_id: str = ""
+    asset_select_mode: str = ""
+    batch_id: str = ""
+    is_preview: bool = True
+    source_task_id: str = ""
+    output_width: int = 1280
+    output_height: int = 720
+    cover_url: str = ""
+    custom_title: str = ""
     status: str
     progress: float
     result_count: int
     error_message: str
+
+
+class BatchGenerationTaskResponse(BaseModel):
+    """批量生成任务响应。"""
+
+    items: list[GenerationTaskResponse]
+    total: int
+
+
+class ListGenerationTasksResponse(BaseModel):
+    """用户级生成任务列表响应（跨 project）。"""
+
+    items: list[GenerationTaskResponse]

apps/api/app/schemas/job.py

@@ -0,0 +1,109 @@
+"""Job API schemas — Phase 8 任务 2.10."""
+
+from __future__ import annotations
+
+from datetime import datetime
+from typing import Any, Optional
+
+from pydantic import BaseModel, Field
+
+
+class CreateJobRequest(BaseModel):
+    """创建任务请求体。"""
+
+    project_id: str = Field(..., min_length=1, description="项目 ID")
+    job_type: str = Field(
+        ...,
+        description="任务类型: video_compose / render_edit_plan / asset_ingest / classification / voice_extraction / generation",
+    )
+    payload: dict[str, Any] = Field(default_factory=dict, description="任务输入参数")
+    source_id: str = Field(default="", description="关联的业务实体 ID（如 edit_plan_id）")
+    max_retries: int = Field(default=3, ge=0, le=10, description="最大重试次数")
+
+
+class UpdateProgressRequest(BaseModel):
+    """更新任务进度请求体。"""
+
+    progress: float = Field(..., ge=0.0, le=100.0, description="进度百分比")
+    current_stage: str = Field(default="", description="当前阶段描述")
+
+
+class CompleteJobRequest(BaseModel):
+    """完成任务请求体。"""
+
+    result: dict[str, Any] = Field(default_factory=dict, description="任务结果")
+
+
+class FailJobRequest(BaseModel):
+    """标记任务失败请求体。"""
+
+    error_message: str = Field(..., min_length=1, description="错误信息")
+
+
+class JobResponse(BaseModel):
+    """任务响应体。"""
+
+    id: str
+    project_id: str
+    job_type: str
+    status: str
+    progress: float
+    current_stage: str
+    payload: dict[str, Any]
+    result: dict[str, Any]
+    error_message: str
+    retry_count: int
+    max_retries: int
+    celery_task_id: str
+    source_id: str
+    created_by_user_id: str
+    is_retryable: bool
+    started_at: Optional[datetime] = None
+    completed_at: Optional[datetime] = None
+    created_at: datetime
+    updated_at: datetime
+
+    model_config = {"from_attributes": True}
+
+
+class ListJobsResponse(BaseModel):
+    """任务列表响应体。"""
+
+    items: list[JobResponse]
+    total: int
+
+
+class JobStatisticsResponse(BaseModel):
+    """任务统计响应体。"""
+
+    project_id: str
+    total: int
+    pending: int
+    running: int
+    success: int
+    failed: int
+
+
+def job_to_response(job) -> JobResponse:
+    """将 Job 领域对象转换为 API 响应。"""
+    return JobResponse(
+        id=job.id,
+        project_id=job.project_id,
+        job_type=job.job_type.value if hasattr(job.job_type, "value") else str(job.job_type),
+        status=job.status.value if hasattr(job.status, "value") else str(job.status),
+        progress=job.progress,
+        current_stage=job.current_stage,
+        payload=job.payload,
+        result=job.result,
+        error_message=job.error_message,
+        retry_count=job.retry_count,
+        max_retries=job.max_retries,
+        celery_task_id=job.celery_task_id,
+        source_id=job.source_id,
+        created_by_user_id=job.created_by_user_id,
+        is_retryable=job.is_retryable,
+        started_at=job.started_at,
+        completed_at=job.completed_at,
+        created_at=job.created_at,
+        updated_at=job.updated_at,
+    )

apps/api/app/schemas/recipe.py

@@ -0,0 +1,86 @@
+"""Recipe API schemas."""
+
+from __future__ import annotations
+
+from datetime import datetime
+from typing import Any, Dict, List, Optional
+
+from pydantic import BaseModel, Field
+
+# ── Response ──
+
+
+class RecipeItemResponse(BaseModel):
+    id: str
+    recipe_id: str
+    item_type: str
+    item_id: str
+    position: int
+    metadata_: Dict[str, Any] = Field(default_factory=dict, alias="metadata")
+
+    class Config:
+        populate_by_name = True
+
+
+class RecipeResponse(BaseModel):
+    id: str
+    user_id: str
+    name: str
+    description: str = ""
+    template_id: str = ""
+    generation_params: Dict[str, Any] = Field(default_factory=dict)
+    items: List[RecipeItemResponse] = Field(default_factory=list)
+    is_active: bool = True
+    metadata_: Dict[str, Any] = Field(default_factory=dict, alias="metadata")
+    created_at: datetime
+    updated_at: datetime
+
+    class Config:
+        populate_by_name = True
+
+
+class ListRecipesResponse(BaseModel):
+    items: List[RecipeResponse]
+    total: int = 0
+
+
+class UseRecipeResponse(BaseModel):
+    recipe: RecipeResponse
+    warnings: List[Dict[str, Any]] = Field(default_factory=list)
+
+
+# ── Request ──
+
+
+class RecipeItemRequest(BaseModel):
+    item_type: str
+    item_id: str
+    position: int = 0
+    metadata_: Dict[str, Any] = Field(default_factory=dict, alias="metadata")
+
+    class Config:
+        populate_by_name = True
+
+
+class CreateRecipeRequest(BaseModel):
+    name: str
+    description: str = ""
+    template_id: str = ""
+    generation_params: Dict[str, Any] = Field(default_factory=dict)
+    items: List[RecipeItemRequest] = Field(default_factory=list)
+    metadata_: Dict[str, Any] = Field(default_factory=dict, alias="metadata")
+
+    class Config:
+        populate_by_name = True
+
+
+class UpdateRecipeRequest(BaseModel):
+    name: Optional[str] = None
+    description: Optional[str] = None
+    template_id: Optional[str] = None
+    generation_params: Optional[Dict[str, Any]] = None
+    items: Optional[List[RecipeItemRequest]] = None
+    metadata_: Optional[Dict[str, Any]] = Field(default=None, alias="metadata")
+
+    class Config:
+        populate_by_name = True

apps/api/app/schemas/subscription.py

@@ -0,0 +1,105 @@
+"""Subscription schemas for API request/response models."""
+
+from __future__ import annotations
+
+from typing import List, Optional
+
+from pydantic import BaseModel, Field
+
+# ============ Enums / Types ============
+
+
+class PlanType(str):
+    """套餐类型"""
+
+    FREE = "free"
+    STANDARD = "standard"
+    PRO = "pro"
+    ENTERPRISE = "enterprise"
+
+
+class SubscriptionStatus(str):
+    """订阅状态"""
+
+    ACTIVE = "active"
+    EXPIRED = "expired"
+    CANCELLED = "cancelled"
+    TRIAL = "trial"
+
+
+class BillingStatus(str):
+    """账单状态"""
+
+    PAID = "paid"
+    PENDING = "pending"
+    FAILED = "failed"
+    REFUNDED = "refunded"
+
+
+class BillingCycle(str):
+    """计费周期"""
+
+    MONTHLY = "monthly"
+    YEARLY = "yearly"
+
+
+# ============ Response Schemas ============
+
+
+class SubscriptionInfo(BaseModel):
+    """当前订阅信息"""
+
+    id: str
+    plan_id: str
+    plan_name: str
+    status: str
+    billing_cycle: str
+    current_period_start: str
+    current_period_end: str
+    amount: float
+    auto_renew: bool
+    created_at: str
+
+
+class BillingRecord(BaseModel):
+    """账单记录"""
+
+    id: str
+    plan_name: str
+    amount: float
+    billing_cycle: str
+    status: str
+    payment_method: str
+    created_at: str
+    invoice_url: Optional[str] = None
+
+
+class ChangePlanResponse(BaseModel):
+    """升级/降级响应"""
+
+    success: bool
+    message: str
+    new_subscription: Optional[SubscriptionInfo] = None
+
+
+class SimpleResponse(BaseModel):
+    """简单响应（用于取消订阅、切换自动续费等）"""
+
+    success: bool
+    message: str
+
+
+# ============ Request Schemas ============
+
+
+class ChangePlanRequest(BaseModel):
+    """升级/降级请求"""
+
+    target_plan_id: str = Field(..., description="目标套餐ID")
+    billing_cycle: str = Field(..., description="计费周期: monthly/yearly")
+
+
+class ToggleAutoRenewRequest(BaseModel):
+    """切换自动续费请求"""
+
+    enabled: bool = Field(..., description="是否开启自动续费")

apps/api/app/schemas/tag.py

@@ -0,0 +1,24 @@
+"""标签相关 Schema。"""
+
+from datetime import datetime
+
+from pydantic import BaseModel, Field
+
+
+class CreateTagRequest(BaseModel):
+    name: str = Field(..., min_length=1, max_length=100)
+
+
+class TagResponse(BaseModel):
+    id: str
+    name: str
+    created_at: datetime
+
+
+class ListTagsResponse(BaseModel):
+    items: list[TagResponse]
+    total: int = Field(default=0, ge=0)
+
+
+class TagAssetsRequest(BaseModel):
+    tag_ids: list[str] = Field(..., min_length=1, max_length=50)

apps/api/app/schemas/task_center.py

@@ -14,9 +14,34 @@ class ProjectTaskResponse(BaseModel):
     user_message: str = ""
     retryable: bool = False
     source_id: str = ""
+    template_id: str = ""
     created_at: datetime | None = None
     updated_at: datetime | None = None
 
 
 class ListProjectTasksResponse(BaseModel):
     items: list[ProjectTaskResponse] = Field(default_factory=list)
+
+
+class UserTaskResponse(BaseModel):
+    """用户级任务响应（跨 project，用于模板模式）。"""
+
+    id: str
+    task_type: str
+    project_id: str = ""
+    template_id: str = ""
+    status: str
+    progress: float
+    current_step: str
+    error_message: str = ""
+    user_message: str = ""
+    retryable: bool = False
+    source_id: str = ""
+    created_at: datetime | None = None
+    updated_at: datetime | None = None
+
+
+class ListTasksResponse(BaseModel):
+    """用户级任务列表响应（GET /api/v1/tasks）。"""
+
+    items: list[UserTaskResponse] = Field(default_factory=list)

apps/api/app/schemas/template.py

@@ -0,0 +1,122 @@
+"""Template API schemas."""
+
+from __future__ import annotations
+
+from datetime import datetime
+from typing import Any, Dict, List, Optional
+
+from pydantic import BaseModel, Field
+
+# ── Segment ──
+
+
+class SegmentResponse(BaseModel):
+    id: str
+    template_id: str
+    segment_order: int
+    duration_min: float
+    duration_max: float
+    material_type: Optional[str] = None
+    created_at: datetime
+    updated_at: datetime
+
+
+class SegmentRequest(BaseModel):
+    segment_order: int
+    duration_min: float
+    duration_max: float
+    material_type: Optional[str] = None
+
+
+# ── Template Response ──
+
+
+class TemplateResponse(BaseModel):
+    id: str
+    user_id: str
+    name: str
+    mode: str
+    category: str = ""
+    tags: List[str] = Field(default_factory=list)
+    title_config: Dict[str, Any] = Field(default_factory=dict)
+    subtitle_config: Dict[str, Any] = Field(default_factory=dict)
+    bgm_config: Dict[str, Any] = Field(default_factory=dict)
+    estimated_duration: float = 0.0
+    segments: List[SegmentResponse] = Field(default_factory=list)
+    is_active: bool = True
+    is_favorite: bool = False
+    created_at: datetime
+    updated_at: datetime
+
+
+class ToggleFavoriteResponse(BaseModel):
+    id: str
+    is_favorite: bool
+
+
+class ListTemplatesResponse(BaseModel):
+    items: List[TemplateResponse]
+    total: int = 0
+
+
+# ── Template Request ──
+
+
+class CreateTemplateRequest(BaseModel):
+    name: str
+    mode: str
+    category: str = ""
+    tags: List[str] = Field(default_factory=list)
+    title_config: Dict[str, Any] = Field(default_factory=dict)
+    subtitle_config: Dict[str, Any] = Field(default_factory=dict)
+    bgm_config: Dict[str, Any] = Field(default_factory=dict)
+    estimated_duration: float = 0.0
+    segments: List[SegmentRequest] = Field(default_factory=list)
+
+
+class UpdateTemplateRequest(BaseModel):
+    name: Optional[str] = None
+    mode: Optional[str] = None
+    category: Optional[str] = None
+    tags: Optional[List[str]] = None
+    title_config: Optional[Dict[str, Any]] = None
+    subtitle_config: Optional[Dict[str, Any]] = None
+    bgm_config: Optional[Dict[str, Any]] = None
+    estimated_duration: Optional[float] = None
+    segments: Optional[List[SegmentRequest]] = None
+
+
+# ── Validate ──
+
+
+class ValidateTemplateRequest(BaseModel):
+    voiceover_duration: Optional[float] = None  # 配音实际时长（秒）
+
+
+class GenerateWarningResponse(BaseModel):
+    code: str
+    message: str
+    details: Dict[str, Any] = Field(default_factory=dict)
+
+
+class ValidateTemplateResponse(BaseModel):
+    template: TemplateResponse
+    warnings: List[GenerateWarningResponse] = Field(default_factory=list)
+
+
+# ── Category ──
+
+
+class CategoryResponse(BaseModel):
+    id: str
+    user_id: str
+    name: str
+    created_at: datetime
+
+
+class CreateCategoryRequest(BaseModel):
+    name: str
+
+
+class ListCategoriesResponse(BaseModel):
+    items: List[CategoryResponse]