fix: #1789 标题字号滑块拖动回弹 — useMemo 稳定引用 + ref 防御

根因：useTitleCoverSync 的 useEffect 依赖 userTemplates（每次渲染
.filter() 生成新数组引用），导致：
  用户拖滑块 → setState → 重渲染 → 新数组引用 → effect 触发 →
  从模板 title_config.font_size 覆盖回去 → slider 回弹

修复1（核心）：useTemplateSelection.ts 用 useMemo 缓存 validTemplates，
  仅在 allTemplates 变化时重新计算，稳定下游引用。
修复2（防御性）：useTitleCoverSync.ts 用 ref 持有 userTemplates，
  从依赖数组移除，effect 只在 selectedTemplate 真正变化时触发。

alembic/versions/069_project_is_default_idempotent.py

@@ -0,0 +1,72 @@
+"""Projects is_default + partial unique index for idempotent default project (Issue #1775)
+
+Revision ID: 069_project_is_default
+Revises: 068_user_profile_completed
+Create Date: 2026-09-08
+
+背景：
+小程序端 getOrCreateDefaultProject 在重试/并发/前端重复调用下，
+仅靠应用层"先查再插"不保证幂等，会给同一用户重复创建默认项目。
+
+改动：
+1. projects 表新增 is_default 布尔列（默认 false）
+2. 部分唯一索引 uq_projects_owner_default：(owner_user_id) WHERE is_default = true
+   —— 保证每个用户至多一个默认项目
+3. 存量数据回填：把名为"默认项目"的存量项目按创建时间最早者标记为 is_default=true
+   （只标记不删除；存量重复项目的清理另行确认后单独执行）
+
+注意：部分唯一索引依赖 PostgreSQL，不支持 downgrade 到其他方言。
+"""
+
+import sqlalchemy as sa
+
+from alembic import op
+
+revision = "069_project_is_default"
+down_revision = "068_user_profile_completed"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # 1. 新增 is_default 列
+    op.add_column(
+        "projects",
+        sa.Column(
+            "is_default",
+            sa.Boolean(),
+            nullable=False,
+            server_default=sa.text("false"),
+        ),
+    )
+
+    # 2. 存量回填：每个拥有"默认项目"的用户，只把最早创建的那一个标记为默认。
+    #    用 ROW_NUMBER() 取每组第一条；非"默认项目"命名的项目不标记（保守，不动用户自建项目）。
+    op.execute("""
+        UPDATE projects p
+        SET is_default = true
+        WHERE p.id IN (
+            SELECT id FROM (
+                SELECT id,
+                       ROW_NUMBER() OVER (
+                           PARTITION BY owner_user_id
+                           ORDER BY created_at ASC, id ASC
+                       ) AS rn
+                FROM projects
+                WHERE name = '默认项目'
+            ) t
+            WHERE t.rn = 1
+        )
+        """)
+
+    # 3. 部分唯一索引：每用户至多一个默认项目（只约束 is_default = true 的行）
+    op.execute("""
+        CREATE UNIQUE INDEX uq_projects_owner_default
+        ON projects (owner_user_id)
+        WHERE is_default = true
+        """)
+
+
+def downgrade() -> None:
+    op.execute("DROP INDEX IF EXISTS uq_projects_owner_default")
+    op.drop_column("projects", "is_default")

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

@@ -20,7 +20,7 @@ from packages.application import (
     GetProjectUseCase,
     ListAssetLibrariesUseCase,
 )
-from packages.domain import AssetLibrary, AssetLibraryKind
+from packages.domain import AssetLibraryKind
 
 from ._helpers import check_project_access
 
@@ -120,30 +120,11 @@ def ensure_default_library(
 
     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)
+    # Issue #1775: 幂等获取/创建——依赖唯一约束 uq_asset_libraries_project_kind，
+    # 并发创建冲突时回滚重查返回已有记录，不再依赖应用层"先查后插"，也不会 500。
     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)
+    library = asset_library_repository.get_or_create_default_library(request.project_id, kind, name=default_name)
+    return _to_asset_library_response(library)
 
 
 @router.delete("/{library_id}", status_code=status.HTTP_204_NO_CONTENT, response_class=Response)

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

@@ -1,13 +1,14 @@
 from typing import Any
 
 from app.auth import AuthenticatedUser, get_current_user
-from app.dependencies import get_project_repository
+from app.dependencies import get_asset_library_repository, get_project_repository
 from app.schemas.project import (
     CreateProjectRequest,
     ListProjectsResponse,
     ProjectResponse,
 )
 from fastapi import APIRouter, Depends, HTTPException, Response, status
+from pydantic import BaseModel
 
 from packages.application import (
     CreateProjectCommand,
@@ -16,10 +17,20 @@ from packages.application import (
     GetProjectUseCase,
     ListProjectsUseCase,
 )
+from packages.domain import AssetLibraryKind
 
 router = APIRouter()
 
 
+class DefaultContextResponse(BaseModel):
+    """幂等默认上下文响应（Issue #1775）：默认项目 + 各类型默认素材库 ID。"""
+
+    project_id: str
+    image_library_id: str
+    video_library_id: str
+    voice_library_id: str
+
+
 def _to_project_response(item) -> ProjectResponse:
     return ProjectResponse(
         id=item.id,
@@ -72,6 +83,35 @@ def create_project(
     return _to_project_response(project)
 
 
+@router.post("/ensure-default", response_model=DefaultContextResponse)
+def ensure_default_project_and_libraries(
+    authenticated_user: AuthenticatedUser = Depends(get_current_user),
+    project_repository: Any = Depends(get_project_repository),
+    asset_library_repository: Any = Depends(get_asset_library_repository),
+) -> DefaultContextResponse:
+    """幂等获取/创建当前用户的默认项目和三类默认素材库（Issue #1775）。
+
+    - 同一用户永远只有一个默认项目（部分唯一索引 uq_projects_owner_default）
+    - 同一项目同 kind 永远只有一个默认素材库（唯一约束 uq_asset_libraries_project_kind）
+    - 并发调用/失败重试：唯一约束冲突时返回已存在记录，不报 500
+    - 项目和素材库的创建各自在仓储事务内幂等，冲突回滚后重查返回同一条
+    """
+    user_id = authenticated_user.user.id
+    project = project_repository.get_or_create_default_project(user_id)
+
+    libraries = {}
+    for kind in (AssetLibraryKind.VIDEO, AssetLibraryKind.VOICE, AssetLibraryKind.IMAGE):
+        library = asset_library_repository.get_or_create_default_library(project.id, kind)
+        libraries[kind] = library.id
+
+    return DefaultContextResponse(
+        project_id=project.id,
+        image_library_id=libraries[AssetLibraryKind.IMAGE],
+        video_library_id=libraries[AssetLibraryKind.VIDEO],
+        voice_library_id=libraries[AssetLibraryKind.VOICE],
+    )
+
+
 @router.delete("/{project_id}", status_code=status.HTTP_204_NO_CONTENT, response_model=None, response_class=Response)
 def delete_project(
     project_id: str,

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

@@ -106,6 +106,10 @@ def list_templates(
     tag: str | None = Query(None, description="按标签筛选"),
     keyword: str | None = Query(None, description="按名称关键词搜索"),
     mode: str | None = Query(None, description="按剪辑模式筛选"),
+    valid_only: bool = Query(
+        False,
+        description="仅返回已配置片段的模板（剪辑页传 true；模板编辑器不传，可查看全部模板含草稿）",
+    ),
     authenticated_user: AuthenticatedUser = Depends(get_current_user),
     template_repository: SQLAlchemyTemplateRepository = Depends(_get_template_repository),
 ) -> ListTemplatesResponse:
@@ -116,6 +120,7 @@ def list_templates(
             tag=tag,
             keyword=keyword,
             mode=mode,
+            valid_only=valid_only,
         )
         use_case = ListTemplatesUseCase(template_repository)
         templates = use_case.execute(user_id, skip=skip, limit=limit, filter=tpl_filter)

apps/api/app/api/routes/templates_editor/clips.py

@@ -36,17 +36,11 @@ from app.services.asset_segment_tracker import (
     remove_used_segment,
 )
 from app.services.edit_plan_service import EditPlanService
-from app.services.edit_template_service import EditTemplateService
+from app.services.edit_template_service import EditTemplateService, TemplateNotFoundError
 from fastapi import APIRouter, BackgroundTasks, Depends, HTTPException, Query, status
 from sqlalchemy.orm import Session
 
 from packages.adapters.sqlalchemy_impl.asset_repository import SQLAlchemyAssetRepository
-from packages.adapters.sqlalchemy_impl.template_clip_config_repository import (
-    SQLAlchemyTemplateClipConfigRepository,
-)
-from packages.adapters.sqlalchemy_impl.template_repository import (
-    SQLAlchemyTemplateRepository,
-)
 from packages.domain.plan_generator_utils import (
     _calc_random_start_time,
     build_scene_segments,
@@ -399,73 +393,42 @@ def _safe_segment_duration(value, default: float) -> float:
 
 def _get_template_segments(
     template_id: str,
+    user_id: str,
     tpl_svc: EditTemplateService,
-    db: Session,
 ) -> list[tuple[int, float, float]]:
     """获取模板的片段配置（顺序、最短时长、最长时长）.
 
-    优先从新模板系统（template_clip_configs）查询，
-    若不存在则回退到旧模板系统（template_segments）。
+    单一数据源：模板主表为 ``templates``（用户自建，归属 user_id）/
+    ``edit_templates``（全局模板库），片段配置主表为 ``template_clip_configs``
+    （由 ``EditTemplateService.list_clip_configs_for_editor`` 统一读取）。
+
+    不再使用"新表抛异常 → 降级直查配置表 → 再降级查 segments"的异常控制流，
+    也不在正常请求中打印 ``ValueError: 模板不存在`` 堆栈。
+
+    Args:
+        template_id: 模板 ID
+        user_id: 当前登录用户 ID（用于归属校验）
+        tpl_svc: 模板编辑器服务
 
     Returns:
-        [(segment_order, duration_min, duration_max), ...] 按 order 排序
+        [(segment_order, duration_min, duration_max), ...] 按 order 排序；
+        模板存在但未配置片段时返回空列表。
+
+    Raises:
+        TemplateNotFoundError: 模板不存在、已删除或不归属于当前用户。
     """
-    # 优先查新模板系统
-    try:
-        clip_configs = tpl_svc.list_clip_configs(template_id)
-        if clip_configs:
-            result = []
-            for cc in clip_configs:
-                dur_min = _safe_segment_duration(cc.min_duration, _DEFAULT_EDITOR_CLIP_DURATION)
-                dur_max = _safe_segment_duration(
-                    cc.max_duration or cc.min_duration,
-                    _DEFAULT_EDITOR_CLIP_DURATION,
-                )
-                dur_min, dur_max = min(dur_min, dur_max), max(dur_min, dur_max)
-                result.append((cc.order, dur_min, dur_max))
-            return sorted(result, key=lambda x: x[0])
-    except Exception:
-        logger.warning("新模板系统查询clip_configs失败（主表可能不存在），直接查clip_configs表", exc_info=True)
+    clip_configs = tpl_svc.list_clip_configs_for_editor(template_id, user_id)
 
-    # 兜底：直接查 template_clip_configs 表（片段表有 template_id 外键，不依赖模板主表）
-    try:
-        direct_repo = SQLAlchemyTemplateClipConfigRepository(db)
-        direct_configs = direct_repo.list_by_template(template_id)
-        if direct_configs:
-            result = []
-            for cc in direct_configs:
-                dur_min = _safe_segment_duration(cc.min_duration, _DEFAULT_EDITOR_CLIP_DURATION)
-                dur_max = _safe_segment_duration(
-                    cc.max_duration or cc.min_duration,
-                    _DEFAULT_EDITOR_CLIP_DURATION,
-                )
-                dur_min, dur_max = min(dur_min, dur_max), max(dur_min, dur_max)
-                result.append((cc.order, dur_min, dur_max))
-            return sorted(result, key=lambda x: x[0])
-    except Exception:
-        logger.warning("直接查clip_configs表也失败，继续回退旧系统", exc_info=True)
-
-    # 回退到旧模板系统（template_segments表）
-    try:
-        old_repo = SQLAlchemyTemplateRepository(db)
-        segments = old_repo.list_segments(template_id)
-        if segments:
-            result = []
-            for s in segments:
-                dur_min = _safe_segment_duration(s.duration_min, _DEFAULT_EDITOR_CLIP_DURATION)
-                dur_max = _safe_segment_duration(s.duration_max, _DEFAULT_EDITOR_CLIP_DURATION)
-                dur_min, dur_max = min(dur_min, dur_max), max(dur_min, dur_max)
-                result.append((s.segment_order, dur_min, dur_max))
-            return sorted(result, key=lambda x: x[0])
-    except Exception:
-        logger.warning("旧模板系统查询segments失败", exc_info=True)
-
-    # 所有途径都失败：模板没有片段配置（可能是无效测试模板）
-    logger.error(
-        "模板无片段配置：template_id=%s（可能是 is_active=false 的无效模板）",
-        template_id,
-    )
-    return []
+    result = []
+    for cc in clip_configs:
+        dur_min = _safe_segment_duration(cc.min_duration, _DEFAULT_EDITOR_CLIP_DURATION)
+        dur_max = _safe_segment_duration(
+            cc.max_duration or cc.min_duration,
+            _DEFAULT_EDITOR_CLIP_DURATION,
+        )
+        dur_min, dur_max = min(dur_min, dur_max), max(dur_min, dur_max)
+        result.append((cc.order, dur_min, dur_max))
+    return sorted(result, key=lambda x: x[0])
 
 
 def _recommended_time_conflicts(
@@ -668,13 +631,21 @@ def create_clips_from_assets_editor(
     7. 素材时长为 0 或缺失时报 400，不创建无效片段
     """
     tpl_svc, plan_svc = services
+    user_id = str(current_user.user.id)
 
-    # 1. 查询模板 segments
-    segments = _get_template_segments(template_id, tpl_svc, db)
+    # 1. 查询模板片段配置。模板不存在/已删除/无权限 → 404；
+    #    模板存在但确实未配置片段 → 422（配置错误，与 404 区分）。
+    try:
+        segments = _get_template_segments(template_id, user_id, tpl_svc)
+    except TemplateNotFoundError as exc:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND,
+            detail="模板不存在或无权访问",
+        ) from exc
     if not segments:
         raise HTTPException(
-            status_code=status.HTTP_400_BAD_REQUEST,
-            detail="模板没有片段配置，无法创建片段",
+            status_code=status.HTTP_422_UNPROCESSABLE_ENTITY,
+            detail="模板未配置片段",
         )
 
     # 防御：schema validator 已过滤 null/空串，这里再归一化一次，

apps/api/app/api/routes/templates_editor/dependencies.py

@@ -41,29 +41,33 @@ def get_draft_plan_id(
     这是模板编辑器路由的核心依赖——所有编辑器端点都先经过这里，
     确保 template_id → plan_id 的映射始终存在。
 
-    兼容策略：优先从新模板系统（edit_templates 表）查找，
-    若不存在则回退到旧模板系统（templates 表），确保用户自建模板可用。
+    模板读取遵循单一数据源、显式判定（不使用异常降级）：
+    - 用户自建模板在旧表 ``templates``（归属 user_id，is_active=True）；
+    - 全局模板在新表 ``edit_templates``（无 user_id，全局可读）。
+    模板不存在、已删除或不归属于当前用户时，一律返回 404。
     """
     tpl_svc, plan_svc = services
     user_id = str(current_user.user.id)
 
+    # 0. 门禁：校验模板存在且可访问（即使草稿已缓存命中也要校验，
+    #    避免模板被删除/无权访问后仍可通过既有草稿 plan 继续操作）。
+    old_repo = SQLAlchemyTemplateRepository(db)
+    old_template = old_repo.get_active(template_id, user_id)
+    is_global_template = tpl_svc.get_template(template_id) is not None
+    if old_template is None and not is_global_template:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="模板不存在")
+
     # 1. 草稿已存在 → 直接返回
     draft = tpl_svc.get_template_draft(template_id)
     if draft is not None:
         return draft.id
 
-    # 2. 新系统有模板 → 用新服务创建草稿
-    if tpl_svc.get_template(template_id) is not None:
+    # 2. 全局模板（新系统）→ 用新服务创建草稿
+    if is_global_template:
         draft = tpl_svc.create_template_draft(template_id, user_id=user_id)
         return draft.id
 
-    # 3. 回退到旧模板系统（templates 表）
-    old_repo = SQLAlchemyTemplateRepository(db)
-    old_template = old_repo.get(template_id, user_id=user_id)
-    if old_template is None:
-        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="模板不存在")
-
-    # 4. 基于旧模板创建草稿计划
+    # 3. 旧模板（templates 表）→ 基于旧模板创建草稿计划
     from app.services.plan_generator_service import PlanGeneratorService
 
     from packages.domain.edit_template import EditTemplate, EditTemplateStatus

apps/api/app/api/routes/templates_editor/draft.py

@@ -150,7 +150,7 @@ def rollback_template(
     try:
         tpl = tpl_svc.rollback_to_version(template_id, request.version)
     except ValueError as exc:
-        raise HTTPException(status_code=400, detail=str(exc)) from exc
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail=str(exc)) from exc
 
     clip_configs = tpl_svc.list_clip_configs(template_id)
     return EditorRollbackResponse(

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

@@ -208,6 +208,8 @@ def _create_pending_asset(
     find-or-create：prepare 阶段已按 file_hash/client_upload_id 预建的占位记录
     会被 find_by_library_and_file_hash/find_by_library_and_client_upload_id 命中，
     直接复用并补齐字段（避免 pre-create + complete 重复建两条）。
+
+    Issue #1776: 素材库计数由 asset_repository.create() 自动维护。
     """
     # 1. 按 client_upload_id / file_hash 查找现有记录
     existing = None
@@ -389,6 +391,7 @@ async def prepare_direct_upload(
                 file_size=request.file_size,
             )
             pending_asset_id = pending.id
+            # Issue #1776: 计数由 asset_repository.create() 自动维护
         except Exception as error:
             # 预建失败不阻塞签名：complete 仍可按 OSS 文件 + hash 兜底去重
             logger.warning("预建 asset 占位失败，降级走 old flow: %s", error)
@@ -474,6 +477,7 @@ async def complete_direct_upload(
         client_upload_id=request.client_upload_id,
         file_size=request.file_size,
     )
+    # Issue #1776: 计数由 asset_repository.create() 自动维护
 
     job = _submit_ingest_job(
         project_id=request.project_id,
@@ -568,6 +572,7 @@ async def upload_asset(
         file_hash=file_hash,
         client_upload_id=client_upload_id,
     )
+    # Issue #1776: 计数由 asset_repository.create() 自动维护
 
     job = _submit_ingest_job(
         project_id=project_id,

apps/api/app/services/edit_plan_service.py

@@ -789,20 +789,22 @@ class EditPlanService:
         if plan and hasattr(plan, "config") and plan.config:
             rhythm_template = plan.config.get("rhythm_template")
 
+        # #1768：先获取素材时长，传入 plan_clip_durations 用于最大片段钳制
+        asset_ids = [c.asset_id for c in clips if c.asset_id]
+        durations = self.get_asset_durations(asset_ids)
+        asset_durations_for_plan = [durations.get(c.asset_id, 0.0) for c in clips]
+
         target = plan_clip_durations(
             len(clips),
             voice,
             transition_effects=[c.transition_effect for c in clips],
             transition_durations=[float(c.transition_duration or 0.0) for c in clips],
             rhythm_template=rhythm_template,
+            asset_durations=asset_durations_for_plan,
         )
         if not target:
             return None
 
-        # 素材时长（短素材起点钳 0）
-        asset_ids = [c.asset_id for c in clips if c.asset_id]
-        durations = self.get_asset_durations(asset_ids)
-
         clips_data: list[dict] = []
         for i, c in enumerate(clips):
             dur = float(target[i])
@@ -932,6 +934,17 @@ class EditPlanService:
                 rhythm_templates_for_variants.append(template)
                 logger.info("变体 %d 节奏模板: plan=%s template=%s", idx, plan_ids[idx], template)
 
+        # #1767：BGM 池差异化分配（让批量变体使用不同 BGM / 段落 / 音量）
+        from packages.domain.bgm_pool import allocate_bgm_pool_for_variants
+
+        source_bgm_config = {}
+        source_plan = self.get_plan(source_plan_id)
+        if source_plan and source_plan.config:
+            source_bgm_config = source_plan.config.get("bgm", {}) or {}
+
+        variant_seeds_for_bgm = [rng.randint(0, 999999) for _ in plan_ids]
+        bgm_pool_assignments = allocate_bgm_pool_for_variants(source_bgm_config, variant_seeds_for_bgm)
+
         # 为每个变体生成独立视觉扰动参数（让批量视频画面本身更不同）
         from packages.domain.variant_plan_selector import generate_visual_perturbation
 
@@ -950,8 +963,20 @@ class EditPlanService:
 
                 pixel_pert = generate_pixel_perturbation(rng)
                 config_update["pixel_perturbation"] = pixel_pert
+                # #1767：写入 BGM 池分配（覆盖 bgm 配置中的 preset_id / audio_offset / volume_adjust_db）
+                if idx < len(bgm_pool_assignments):
+                    existing_bgm = dict((source_plan.config or {}).get("bgm", {}) or {})
+                    existing_bgm.update(bgm_pool_assignments[idx])
+                    config_update["bgm"] = existing_bgm
                 self.update_plan_config(pid, config_update)
-                logger.info("变体 %d 视觉扰动+像素扰动: plan=%s vis=%s pix=%s", idx, pid, perturbation, pixel_pert)
+                logger.info(
+                    "变体 %d 视觉扰动+像素扰动+BGM池: plan=%s vis=%s pix=%s bgm=%s",
+                    idx,
+                    pid,
+                    perturbation,
+                    pixel_pert,
+                    bgm_pool_assignments[idx] if idx < len(bgm_pool_assignments) else None,
+                )
             except Exception:
                 logger.exception("变体 %d 视觉扰动生成失败（不阻断）: plan=%s", idx, pid)
 
@@ -1187,7 +1212,7 @@ class EditPlanService:
         config_asset_ids_count = len((plan.config or {}).get("asset_ids", []))
         clips_with_asset_count = sum(1 for c in clips if c.asset_id)
         logger.info(
-            "can_generate 诊断: plan=%s status=%s total_clips=%d " "clips_with_asset=%d config_asset_ids_count=%d",
+            "can_generate 诊断: plan=%s status=%s total_clips=%d clips_with_asset=%d config_asset_ids_count=%d",
             plan_id,
             plan.status,
             len(clips),
@@ -1199,7 +1224,7 @@ class EditPlanService:
             config_asset_ids = (plan.config or {}).get("asset_ids", [])
             if config_asset_ids:
                 logger.warning(
-                    "can_generate 最后防线触发: plan=%s clips=%d 均无素材，" "从 config.asset_ids(%d个) 自动分配",
+                    "can_generate 最后防线触发: plan=%s clips=%d 均无素材，从 config.asset_ids(%d个) 自动分配",
                     plan_id,
                     len(clips),
                     len(config_asset_ids),
@@ -1231,7 +1256,7 @@ class EditPlanService:
                     return False, "没有可渲染的就绪片段，自动修复后仍未分配素材"
             else:
                 logger.warning(
-                    "can_generate 失败: plan=%s clips=%d 均无素材，" "且 config.asset_ids 为空，无法自动修复",
+                    "can_generate 失败: plan=%s clips=%d 均无素材，且 config.asset_ids 为空，无法自动修复",
                     plan_id,
                     len(clips),
                 )

apps/api/app/services/edit_template_service.py

@@ -34,6 +34,17 @@ from packages.domain.template_clip_converter import (
 logger = logging.getLogger(__name__)
 
 
+class TemplateNotFoundError(Exception):
+    """模板不存在、已删除或当前用户无权访问.
+
+    与"模板存在但无片段配置"区分：路由层应映射为 HTTP 404。
+    """
+
+    def __init__(self, template_id: str) -> None:
+        self.template_id = template_id
+        super().__init__(f"模板不存在: {template_id}")
+
+
 class EditTemplateService:
     """模板管理服务
 
@@ -217,7 +228,14 @@ class EditTemplateService:
         skip: int = 0,
         limit: int = 100,
     ) -> List[TemplateClipConfig]:
-        """列出模板的片段配置"""
+        """列出模板的片段配置
+
+        注意：本方法要求模板存在于新表 ``edit_templates``（全局模板库），
+        主要服务于新模板系统的写入/发布路径。用户自建模板存放在旧表
+        ``templates``，不在 ``edit_templates`` 中，读取其片段配置请改用
+        :meth:`list_clip_configs_for_editor`，后者直接读取片段配置主表
+        ``template_clip_configs``，不依赖新模板主表、也不靠异常降级。
+        """
         # 确保模板存在
         self.get_template_or_raise(template_id)
         return self._clip_config_repo.list_by_template(
@@ -227,6 +245,52 @@ class EditTemplateService:
             limit=limit,
         )
 
+    def list_clip_configs_for_editor(
+        self,
+        template_id: str,
+        user_id: str,
+        *,
+        clip_type: Optional[ClipType] = None,
+        skip: int = 0,
+        limit: int = 100,
+    ) -> List[TemplateClipConfig]:
+        """编辑器读取模板片段配置的单一数据源入口.
+
+        片段配置主表是 ``template_clip_configs``（直接读取，不抛异常、不降级）。
+        模板主表按双表现状显式判定，不使用 try/except 控制流：
+
+        1. 用户自建模板在旧表 ``templates``（归属 user_id）→ 校验归属与未删除后直接读；
+        2. 全局模板在新表 ``edit_templates``（无 user_id，全局可读）→ 直接读；
+        3. 两者都没有 → 模板不存在/无权限，抛 :class:`TemplateNotFoundError`。
+
+        Args:
+            template_id: 模板 ID
+            user_id: 当前登录用户 ID（用于旧表模板归属校验）
+
+        Raises:
+            TemplateNotFoundError: 模板不存在、已删除或不归属于当前用户。
+        """
+        # 1) 用户自建模板（旧表 templates，归属 user_id）
+        if self._clip_config_repo.template_owned_by(template_id, user_id):
+            return self._clip_config_repo.list_by_template(
+                template_id,
+                clip_type=clip_type,
+                skip=skip,
+                limit=limit,
+            )
+
+        # 2) 全局模板（新表 edit_templates，无 user_id，全局可读）
+        if self._template_repo.get(template_id) is not None:
+            return self._clip_config_repo.list_by_template(
+                template_id,
+                clip_type=clip_type,
+                skip=skip,
+                limit=limit,
+            )
+
+        # 3) 两表都没有：不存在 / 已删除 / 无权限
+        raise TemplateNotFoundError(template_id)
+
     def get_clip_config(self, config_id: str) -> Optional[TemplateClipConfig]:
         """获取片段配置详情"""
         return self._clip_config_repo.get(config_id)

apps/api/app/services/video_compose_service.py

@@ -39,6 +39,9 @@ from packages.domain.video_filter_builder import (
 )
 from packages.domain.video_filter_builder import build_concat_filter as _build_concat_filter_func
 from packages.domain.video_filter_builder import build_filter_complex as _build_filter_complex
+from packages.domain.video_filter_builder import (
+    build_title_drawtext_filter,
+)
 from packages.domain.video_filter_builder import build_xfade_filter as _build_xfade_filter_func
 from packages.domain.video_filter_builder import chain_filters as _chain_filters_func
 from packages.domain.video_filter_builder import has_audio as _has_audio_func
@@ -248,6 +251,27 @@ class VideoComposeService:
             transitions=[c.transition_effect for c in ready_clips],
         )
 
+        # ── #1789 标题 drawtext 滤镜叠加 ──
+        # 从 plan.config 读取 title_config，生成 drawtext 滤镜链入 filter_complex
+        title_cfg = (plan.config or {}).get("title", {}) or {}
+        if not isinstance(title_cfg, dict):
+            title_cfg = {}
+        # 同时兼容 plan.config["title_config"]（API 回写路径）
+        if not title_cfg.get("text") and not title_cfg.get("content"):
+            title_cfg_alt = (plan.config or {}).get("title_config", {}) or {}
+            if isinstance(title_cfg_alt, dict) and (title_cfg_alt.get("text") or title_cfg_alt.get("content")):
+                title_cfg = title_cfg_alt
+        drawtext_filter = build_title_drawtext_filter(title_cfg, output_width, output_height)
+        if drawtext_filter:
+            # 将最终输出标签从 [outv] 改为 [composed]，再链入 drawtext → [outv]
+            filter_complex = filter_complex.replace("[outv]", "[composed]")
+            filter_complex += f";[composed]{drawtext_filter}[outv]"
+            logger.info(
+                "[#1789] 标题 drawtext 滤镜已注入: plan_id=%s text=%s",
+                plan_id,
+                (title_cfg.get("text") or title_cfg.get("content") or "")[:30],
+            )
+
         # 构建完整命令
         command: list[str] = ["ffmpeg", "-y"]
 

apps/web/src/api/assets/libraries.ts

@@ -5,10 +5,22 @@ import apiClient from "../client"
 import { getOrCreateDefaultProject } from "../projects"
 import type { AssetLibraryItem } from "./types"
 
-/** 获取当前用户的所有素材库 */
-export const getAssetLibraries = async (): Promise<AssetLibraryItem[]> => {
-  const response = await apiClient.get("/asset-libraries")
-  return response.data.items || []
+/**
+ * 获取当前用户的素材库
+ *
+ * @param kind 可选，按素材库类型过滤（video/voice/image）。
+ *   后端 GET /asset-libraries 支持 kind 查询参数；这里同时在前端再按返回数据的
+ *   kind 字段兜底过滤一次，保证旧后端（忽略未知 query 参数）也不会把其他类型的库
+ *   混进来（#1777：视频选择器只展示视频库）。
+ */
+export const getAssetLibraries = async (
+  kind?: AssetLibraryItem["kind"],
+): Promise<AssetLibraryItem[]> => {
+  const response = await apiClient.get<{ items?: AssetLibraryItem[] }>("/asset-libraries", {
+    params: kind ? { kind } : undefined,
+  })
+  const items = response.data.items || []
+  return kind ? items.filter((lib) => lib.kind === kind) : items
 }
 
 /** 创建素材库（自动获取或创建默认项目以提供 project_id） */

apps/web/src/api/client.ts

@@ -55,6 +55,19 @@ apiClient.interceptors.response.use(
   async (error: AxiosError<{ detail?: string; message?: string; msg?: string }>) => {
     const originalRequest = error.config as InternalAxiosRequestConfig & {
       _retry?: boolean
+      /**
+       * 调用方自行处理错误提示时置 true：拦截器跳过全局 message 弹窗（#1777）。
+       * 例如失效模板自动回退时，调用方会弹「原模板已失效，已自动切换」，
+       * 不再叠加后端原始错误文案。错误仍会 reject，不影响 catch 逻辑。
+       */
+      _silentErrorToast?: boolean
+    }
+
+    // 调用方声明自行处理提示：标记为已展示，跳过下面所有全局 message 弹窗
+    if (originalRequest?._silentErrorToast) {
+      // eslint-disable-next-line @typescript-eslint/no-explicit-any
+      ;(error as any).__msgShown = true
+      return Promise.reject(error)
     }
 
     // 401 → 尝试刷新 Token

apps/web/src/api/editing-planner/templates.ts

@@ -12,17 +12,25 @@ import type {
   ListCategoriesResponse,
 } from "./types"
 
-/** 获取模板列表 */
+/** 获取模板列表
+ *
+ * valid_only=true 时请求后端仅返回已配置片段的模板（剪辑页选模板使用，
+ * 避免选中无片段配置的模板导致 from-assets 400，#1769/#1772）；
+ * 后端尚未支持该参数时会忽略未知 query 字段，前端再按 segments/is_active 兜底过滤。
+ * 模板编辑器/我的模板不传，可查看全部模板（含未配置片段的草稿）。
+ */
 export const getEditingTemplates = async (params?: {
   category?: string
   tag?: string
   skip?: number
   limit?: number
+  validOnly?: boolean
 }): Promise<EditingTemplate[]> => {
   const response = await apiClient.get<ListTemplatesResponse>("/templates", {
     params: {
       skip: params?.skip ?? 0,
       limit: params?.limit ?? 50,
+      ...(params?.validOnly ? { valid_only: true } : {}),
     },
   })
   let list = response.data.items

apps/web/src/api/template-editor/clips.ts

@@ -91,7 +91,7 @@ export async function createClipsFromAssets(
   assetIds: string[],
   clipType = "main",
   requiredClipsCount?: number,
-  opts?: { signal?: AbortSignal },
+  opts?: { signal?: AbortSignal; silentErrorToast?: boolean },
 ): Promise<ClipsFromAssetsResponse> {
   const body: Record<string, unknown> = {
     asset_ids: assetIds,
@@ -104,7 +104,12 @@ export async function createClipsFromAssets(
   const response = await apiClient.post<ClipsFromAssetsResponse>(
     `/templates/${templateId}/editor/clips/from-assets`,
     body,
-    { timeout: 60000, signal: opts?.signal },
+    {
+      timeout: 60000,
+      signal: opts?.signal,
+      // _silentErrorToast 由 api/client.ts 响应拦截器读取（抑制全局错误 toast，#1777）
+      ...(opts?.silentErrorToast ? ({ _silentErrorToast: true } as Record<string, unknown>) : {}),
+    },
   )
   return response.data
 }

apps/web/src/api/template-editor/editPlans.ts

@@ -43,11 +43,17 @@ export async function updateEditPlanClips(
   templateId: string,
   clips: EditPlanClipInput[],
   signal?: AbortSignal,
+  /** 为 true 时抑制全局错误 toast（调用方自行提示，如失效模板回退 #1777） */
+  silentErrorToast?: boolean,
 ): Promise<{ count: number }> {
   const response = await apiClient.put(
     `/templates/${templateId}/editor/clips`,
     { clips },
-    { signal },
+    {
+      signal,
+      // _silentErrorToast 由 api/client.ts 响应拦截器读取（抑制全局错误 toast）
+      ...(silentErrorToast ? ({ _silentErrorToast: true } as Record<string, unknown>) : {}),
+    },
   )
   return response.data
 }

apps/web/src/pages/editing-planner/EditingPlanner.css

@@ -6440,3 +6440,47 @@
   opacity: 0.6;
   cursor: not-allowed;
 }
+
+/* ═══ 标题设置 — 颜色预设 ═══ */
+.ep-color-presets {
+  display: flex;
+  align-items: center;
+  gap: 6px;
+  flex-wrap: wrap;
+}
+
+.ep-color-swatch {
+  width: 24px;
+  height: 24px;
+  border-radius: 4px;
+  border: 2px solid transparent;
+  cursor: pointer;
+  transition: border-color 0.15s, transform 0.1s;
+}
+
+.ep-color-swatch:hover {
+  transform: scale(1.1);
+}
+
+.ep-color-swatch.active {
+  border-color: var(--ep-primary, #4f8cff);
+}
+
+.ep-color-picker {
+  width: 28px;
+  height: 28px;
+  border: none;
+  border-radius: 4px;
+  cursor: pointer;
+  padding: 0;
+  background: none;
+}
+
+.ep-color-picker::-webkit-color-swatch-wrapper {
+  padding: 0;
+}
+
+.ep-color-picker::-webkit-color-swatch {
+  border: 1px solid rgba(255, 255, 255, 0.2);
+  border-radius: 4px;
+}

apps/web/src/pages/editing-planner/EditingPlanner.tsx

@@ -196,6 +196,8 @@ const EditingPlanner: React.FC = () => {
 
         {/* 右栏 260px：设置面板 */}
         <RightPanel
+          titleConfig={titleConfig}
+          onTitleConfigChange={setTitleConfig}
           rightTab={rightTab}
           onTabChange={setRightTab}
           selectedClip={clipOps.selectedClip}

apps/web/src/pages/editing-planner/components/ClipPropertiesPanel.tsx

@@ -5,12 +5,15 @@
 import React from "react"
 import type { ClipPropertiesPanelProps } from "@/pages/editing-planner/types/clipProperties"
 import SubtitleSettingsSection from "./clip-properties/SubtitleSettingsSection"
+import TitleSettingsSection from "./clip-properties/TitleSettingsSection"
 import BgmSettingsSection from "./clip-properties/BgmSettingsSection"
 import ClipDetailSection from "./clip-properties/ClipDetailSection"
 import StatsSection from "./clip-properties/StatsSection"
 import { useVoicePreview } from "@/pages/editing-planner/hooks/useVoicePreview"
 
 const ClipPropertiesPanel: React.FC<ClipPropertiesPanelProps> = ({
+  titleConfig,
+  onTitleConfigChange,
   selectedClip,
   subtitleSettings,
   bgmSettings,
@@ -40,6 +43,11 @@ const ClipPropertiesPanel: React.FC<ClipPropertiesPanelProps> = ({
 
   return (
     <div className="ep-right-panel">
+      {/* ═══ 标题设置 — #1789 ═══ */}
+      {titleConfig && onTitleConfigChange && (
+        <TitleSettingsSection config={titleConfig} onChange={onTitleConfigChange} />
+      )}
+
       {/* ═══ 字幕设置 ═══ */}
       <SubtitleSettingsSection
         settings={subtitleSettings}

apps/web/src/pages/editing-planner/components/RightPanel.tsx

@@ -5,9 +5,12 @@ import type { ClipData } from "../types"
 import type { SubtitleStyleConfig } from "../types/subtitle"
 import type { BgmMixConfig } from "@/api/bgm"
 import type { TemplateMode } from "@/api/editing-planner"
+import type { TitleConfig } from "@/api/template-editor"
 import type { AssetItem } from "@/api/assets"
 
 interface RightPanelProps {
+  titleConfig?: TitleConfig
+  onTitleConfigChange?: (config: TitleConfig | ((prev: TitleConfig) => TitleConfig)) => void
   rightTab: "properties" | "clips"
   onTabChange: (tab: "properties" | "clips") => void
   // 属性 tab
@@ -46,6 +49,8 @@ interface RightPanelProps {
 }
 
 const RightPanel: React.FC<RightPanelProps> = ({
+  titleConfig,
+  onTitleConfigChange,
   rightTab,
   onTabChange,
   selectedClip,
@@ -118,6 +123,8 @@ const RightPanel: React.FC<RightPanelProps> = ({
             >["onBgmSettingsChange"]
             return (
               <ClipPropertiesPanel
+                titleConfig={titleConfig}
+                onTitleConfigChange={onTitleConfigChange}
                 selectedClip={selectedClip}
                 subtitleSettings={sub}
                 bgmSettings={bgm}

apps/web/src/pages/editing-planner/components/clip-properties/TitleSettingsSection.tsx

@@ -0,0 +1,136 @@
+/**
+ * 标题设置区块 — #1789
+ * 提供字号滑块、字体预设、位置、颜色等控制入口
+ */
+import React from "react"
+import type { TitleConfig } from "@/api/template-editor"
+import { POSITION_OPTIONS, FONT_OPTIONS } from "@/pages/editing-planner/constants/clipProperties"
+
+interface TitleSettingsSectionProps {
+  config: TitleConfig
+  onChange: (config: TitleConfig | ((prev: TitleConfig) => TitleConfig)) => void
+}
+
+const TITLE_COLOR_PRESETS = [
+  "#ffffff",
+  "#000000",
+  "#ff4444",
+  "#ffaa00",
+  "#44ff44",
+  "#4488ff",
+  "#ff44ff",
+  "#ffff44",
+]
+
+const TitleSettingsSection: React.FC<TitleSettingsSectionProps> = ({ config, onChange }) => {
+  const update = (partial: Partial<TitleConfig>) => {
+    onChange((prev: TitleConfig) => ({ ...prev, ...partial }))
+  }
+
+  return (
+    <div className="ep-settings-section">
+      <div className="ep-section-title">
+        <span className="ep-section-icon">📝</span>
+        标题设置
+      </div>
+
+      {/* AI 自动选择开关 */}
+      <div className="ep-toggle-row">
+        <span className="ep-toggle-label">AI 自动选择</span>
+        <div
+          className={`ep-toggle ${config.ai_auto_select ? "active" : ""}`}
+          onClick={() => update({ ai_auto_select: !config.ai_auto_select })}
+        >
+          <div className="ep-toggle-knob" />
+        </div>
+      </div>
+
+      {!config.ai_auto_select && (
+        <>
+          {/* 标题文本 */}
+          <div className="ep-field">
+            <label className="ep-field-label">标题文本</label>
+            <input
+              className="ep-form-select"
+              type="text"
+              placeholder="输入标题内容"
+              value={config.content}
+              onChange={(e) => update({ content: e.target.value })}
+            />
+          </div>
+
+          {/* 位置 */}
+          <div className="ep-field">
+            <label className="ep-field-label">位置</label>
+            <select
+              className="ep-form-select"
+              value={config.position}
+              onChange={(e) => update({ position: e.target.value })}
+            >
+              {POSITION_OPTIONS.map((opt) => (
+                <option key={opt.value} value={opt.value}>
+                  {opt.label}
+                </option>
+              ))}
+            </select>
+          </div>
+
+          {/* 字体预设 */}
+          <div className="ep-field">
+            <label className="ep-field-label">字体</label>
+            <select
+              className="ep-form-select"
+              value={config.font_preset}
+              onChange={(e) => update({ font_preset: e.target.value })}
+            >
+              {FONT_OPTIONS.map((f) => (
+                <option key={f} value={f}>
+                  {f}
+                </option>
+              ))}
+            </select>
+          </div>
+
+          {/* 字号滑块 */}
+          <div className="ep-field">
+            <label className="ep-field-label">字号</label>
+            <div className="ep-slider-row">
+              <input
+                className="ep-slider"
+                type="range"
+                min={12}
+                max={72}
+                value={config.font_size}
+                onChange={(e) => update({ font_size: Number(e.target.value) })}
+              />
+              <span className="ep-slider-value">{config.font_size}px</span>
+            </div>
+          </div>
+
+          {/* 颜色 */}
+          <div className="ep-field">
+            <label className="ep-field-label">颜色</label>
+            <div className="ep-color-presets">
+              {TITLE_COLOR_PRESETS.map((color) => (
+                <div
+                  key={color}
+                  className={`ep-color-swatch${config.font_color === color ? " active" : ""}`}
+                  style={{ backgroundColor: color }}
+                  onClick={() => update({ font_color: color })}
+                />
+              ))}
+              <input
+                type="color"
+                className="ep-color-picker"
+                value={config.font_color}
+                onChange={(e) => update({ font_color: e.target.value })}
+              />
+            </div>
+          </div>
+        </>
+      )}
+    </div>
+  )
+}
+
+export default TitleSettingsSection

apps/web/src/pages/editing-planner/types/clipProperties.ts

@@ -4,6 +4,7 @@
 import type { ClipData } from "./clip"
 import type { TemplateMode } from "@/api/editing-planner"
 import type { AssetItem } from "@/api/assets"
+import type { TitleConfig } from "@/api/template-editor"
 
 export interface SubtitleSettings {
   enabled: boolean
@@ -28,6 +29,10 @@ export interface BgmSettings {
 }
 
 export interface ClipPropertiesPanelProps {
+  /** 标题配置 — #1789 */
+  titleConfig?: TitleConfig
+  /** 标题配置变更 */
+  onTitleConfigChange?: (config: TitleConfig | ((prev: TitleConfig) => TitleConfig)) => void
   selectedClip: ClipData | null
   subtitleSettings: SubtitleSettings
   bgmSettings: BgmSettings

apps/web/src/pages/editing-planner/utils/timeline.ts

@@ -13,6 +13,8 @@ export const formatTrimTime = (sec: number): string => {
 /** 生成时间标尺刻度 */
 export const generateRulerMarks = (totalDuration: number, step: number): number[] => {
   const marks: number[] = []
+  // #1790: 无片段时不显示时间刻度
+  if (totalDuration <= 0) return marks
   for (let t = 0; t <= totalDuration + step; t += step) {
     marks.push(t)
   }

apps/web/src/pages/generate/GeneratePage.tsx

@@ -45,6 +45,7 @@ const GeneratePage: React.FC = () => {
     selectedTemplate,
     setSelectedTemplate,
     userTemplates,
+    handleInvalidTemplate,
     selectedMaterials,
     setSelectedMaterials,
     materialMode,
@@ -416,15 +417,11 @@ const GeneratePage: React.FC = () => {
   /* ── 最终成片（单视频右侧播放） ── */
   const finalVideo = generatedVideos[0]
 
-  /* ── 布局 class：步骤4标题页=预览+标题侧栏；步骤5/6批量=整行宽；步骤1~3=整行宽 ── */
+  /* ── 布局 class：步骤4标题页=预览+标题侧栏两栏；其余步骤（含步骤5确认生成、步骤6封面）=整行宽 ── */
   const layoutClassName = useMemo(() => {
-    if (currentStep < 4) return "xx-generate-layout full-width"
     if (currentStep === 4) return "xx-generate-layout step4-layout"
-    // 步骤5：全宽+内容居中（单视频视频播放器居中，批量网格居中）
-    if (currentStep === 5) return "xx-generate-layout full-width"
-    // 步骤6：封面选择保持两栏布局
-    return isBatch ? "xx-generate-layout full-width" : "xx-generate-layout"
-  }, [currentStep, isBatch])
+    return "xx-generate-layout full-width"
+  }, [currentStep])
 
   /* ================================================================
      渲染
@@ -524,6 +521,7 @@ const GeneratePage: React.FC = () => {
             selectedVoice={selectedVoice}
             onSelectedVoiceChange={setSelectedVoice}
             onServerClipsChange={setServerClips}
+            onTemplateInvalid={handleInvalidTemplate}
             generating={generating}
             generated={generated}
             generateError={generateError}
@@ -545,18 +543,7 @@ const GeneratePage: React.FC = () => {
             selectedVariantIds={selectedVariantIds}
           />
 
-          <GenerateStepActions
-            currentStep={currentStep}
-            onPrev={goPrev}
-            onNext={goNext}
-            onConfirmGenerate={handleConfirmGenerate}
-            generating={generating}
-            generated={generated}
-            generateError={generateError}
-            selectedCount={isBatch ? selectedVariantIds.length : 1}
-          />
-
-          {/* ════ 步骤5（单视频）：成片播放器内联居中（#1761） ════ */}
+          {/* ════ 步骤5（单视频）：成片播放器置于按钮上方、居中展示 ════ */}
           {currentStep === 5 && !isBatch && generated && finalVideo && (
             <div
               style={{
@@ -580,6 +567,7 @@ const GeneratePage: React.FC = () => {
                 <video
                   src={finalVideo.download_url || finalVideo.file_url}
                   controls
+                  autoPlay
                   style={{
                     width: "auto",
                     maxWidth: "100%",
@@ -607,55 +595,18 @@ const GeneratePage: React.FC = () => {
               </div>
             </div>
           )}
-        </div>
 
-        {/* ════ 步骤6（单视频）：右侧成片播放器 ════ */}
-        {currentStep >= 6 && !isBatch && generated && finalVideo && (
-          <div className="xx-generate-right-col">
-            <div
-              className="xx-inline-video-player"
-              style={{
-                display: "flex",
-                justifyContent: "center",
-                alignItems: "center",
-                background: "#000",
-                borderRadius: 12,
-                padding: 8,
-              }}
-            >
-              <video
-                src={finalVideo.download_url || finalVideo.file_url}
-                controls
-                autoPlay={currentStep === 5}
-                // 竖屏自适应（#1750）：成片固定 1080×1920（9:16），元数据到达前按 9:16 占位，
-                // 到达后浏览器按真实宽高比 contain；黑底居中杜绝左右大黑边
-                style={{
-                  width: "auto",
-                  maxWidth: "100%",
-                  maxHeight: "70vh",
-                  aspectRatio: "9 / 16",
-                  objectFit: "contain",
-                  borderRadius: 8,
-                }}
-                poster={finalVideo.thumbnail_url || undefined}
-              />
-              <div style={{ display: "flex", gap: 8, marginTop: 12, justifyContent: "center" }}>
-                <button className="xx-btn xx-btn-ghost xx-btn-sm" onClick={handleDownload}>
-                  ⬇️ 下载
-                </button>
-                <button className="xx-btn xx-btn-ghost xx-btn-sm" onClick={handleShare}>
-                  🔗 分享
-                </button>
-                <button
-                  className="xx-btn xx-btn-ghost xx-btn-sm"
-                  onClick={() => navigate("/app/products")}
-                >
-                  📁 前往成片库
-                </button>
-              </div>
-            </div>
-          </div>
-        )}
+          <GenerateStepActions
+            currentStep={currentStep}
+            onPrev={goPrev}
+            onNext={goNext}
+            onConfirmGenerate={handleConfirmGenerate}
+            generating={generating}
+            generated={generated}
+            generateError={generateError}
+            selectedCount={isBatch ? selectedVariantIds.length : 1}
+          />
+        </div>
       </div>
 
       {/* 数量选择弹窗 */}

apps/web/src/pages/generate/components/GenerateStepContent.tsx

@@ -50,6 +50,8 @@ export interface GenerateStepContentProps {
   selectedVoice: string
   onSelectedVoiceChange: (id: string) => void
   onServerClipsChange: (clips: EditPlanClip[]) => void
+  /** 当前模板创建片段被判失效（404/400/422）时的自动回退回调（#1777） */
+  onTemplateInvalid?: () => boolean
   /* 生成 */
   generating: boolean
   generated: boolean
@@ -108,6 +110,7 @@ export const GenerateStepContent: React.FC<GenerateStepContentProps> = (props) =
     selectedVoice,
     onSelectedVoiceChange,
     onServerClipsChange,
+    onTemplateInvalid,
     generating,
     generated,
     generateError,
@@ -153,6 +156,7 @@ export const GenerateStepContent: React.FC<GenerateStepContentProps> = (props) =
           selectedTemplate={selectedTemplate}
           templateSegments={templateSegments}
           onServerClipsChange={onServerClipsChange}
+          onTemplateInvalid={onTemplateInvalid}
         />
       )
     case 3:
@@ -189,7 +193,7 @@ export const GenerateStepContent: React.FC<GenerateStepContentProps> = (props) =
         />
       )
     case 5:
-      /* 确认生成页：批量=逐任务进度网格；单视频=进度状态卡（成片播放器在左侧大区域） */
+      /* 确认生成页：批量=逐任务进度网格；单视频=仅渲染进度/失败状态（完成后只显示成片播放器，播放器在按钮上方） */
       if (previewCount > 1) {
         return (
           <BatchGenerationGrid
@@ -199,10 +203,10 @@ export const GenerateStepContent: React.FC<GenerateStepContentProps> = (props) =
           />
         )
       }
-      /* 单视频：渲染进度 / 失败重试 / 完成提示（成片播放器在右侧栏） */
+      /* 单视频：生成中显示进度卡、失败显示重试卡；生成完成后不再渲染提示卡，页面只保留成片播放器+操作按钮 */
+      if (generated && !generating && !generateError) return null
       return (
         <div className="xx-form-section">
-          <h3>🎬 确认生成</h3>
           {generating && (
             <div className="xx-gen-progress-card">
               <div className="xx-gen-progress-header">
@@ -234,14 +238,6 @@ export const GenerateStepContent: React.FC<GenerateStepContentProps> = (props) =
               </button>
             </div>
           )}
-          {generated && !generating && (
-            <div className="xx-gen-success-card">
-              <div className="xx-gen-success-info">
-                <div className="xx-gen-success-title">✅ 视频生成完成！</div>
-                <div className="xx-gen-success-sub">右侧可预览成片，点击「下一步」选择封面</div>
-              </div>
-            </div>
-          )}
         </div>
       )
     case 6:

apps/web/src/pages/generate/components/Step2MaterialSelect.tsx

@@ -23,6 +23,8 @@ interface Step2MaterialSelectProps {
   templateSegments?: TemplateSegment[]
   /** 服务端 clips 创建成功后的回调 */
   onServerClipsChange?: (clips: EditPlanClip[]) => void
+  /** 当前模板创建片段返回 404/400/422（模板失效）时的自动回退回调（#1777） */
+  onTemplateInvalid?: () => boolean
 }
 
 const Step2MaterialSelect: React.FC<Step2MaterialSelectProps> = (props) => {
@@ -36,16 +38,25 @@ const Step2MaterialSelect: React.FC<Step2MaterialSelectProps> = (props) => {
 
       <div className="xx-form-field" style={{ marginTop: 12 }}>
         <label>选择视频库</label>
-        <select
-          value={m.selectedLibraryId}
-          onChange={(e) => m.setSelectedLibraryId(e.target.value)}
-        >
-          {m.libraries.map((lib) => (
-            <option key={lib.id} value={lib.id}>
-              {lib.name}
-            </option>
-          ))}
-        </select>
+        {m.libraries.length === 0 && !m.materialsLoading ? (
+          <div className="xx-empty-state">
+            <p>暂无视频素材库</p>
+            <p style={{ fontSize: 13, color: "var(--text-tertiary)" }}>
+              请先在「素材库」中创建视频素材库并上传视频
+            </p>
+          </div>
+        ) : (
+          <select
+            value={m.selectedLibraryId}
+            onChange={(e) => m.setSelectedLibraryId(e.target.value)}
+          >
+            {m.libraries.map((lib) => (
+              <option key={lib.id} value={lib.id}>
+                {lib.name}
+              </option>
+            ))}
+          </select>
+        )}
       </div>
 
       {m.materialMode === "manual" && (

apps/web/src/pages/generate/components/step7-confirm/GenerationStatus.tsx

@@ -1,117 +0,0 @@
-import React from "react"
-import { LoadingOutlined, CheckCircleFilled, CloseCircleOutlined } from "@ant-design/icons"
-import type { GeneratedVideo } from "@/api/template-editor"
-
-interface GenerationStatusProps {
-  generating: boolean
-  generated: boolean
-  generateError: string | null
-  progress: number
-  generatedVideos: GeneratedVideo[]
-  getGenerationPhase: (progress: number) => { icon: string; label: string }
-  onScrollToPreview: () => void
-  onRetry: () => void
-  onDismissError: () => void
-}
-
-const GenerationStatus: React.FC<GenerationStatusProps> = ({
-  generating,
-  generated,
-  generateError,
-  progress,
-  generatedVideos,
-  getGenerationPhase,
-  onScrollToPreview,
-  onRetry,
-  onDismissError,
-}) => {
-  return (
-    <div style={{ marginTop: 16 }}>
-      {!generating && !generated && !generateError && (
-        <div className="xx-gen-progress-card" style={{ opacity: 0.85 }}>
-          <div className="xx-gen-progress-header">
-            <div className="xx-gen-progress-icon">🎬</div>
-            <div className="xx-gen-progress-info">
-              <div className="xx-gen-progress-phase">尚未开始生成视频</div>
-              <div className="xx-gen-progress-sub">
-                请返回「选择标题」步骤，点击「确认生成视频」开始渲染最终视频
-              </div>
-            </div>
-          </div>
-        </div>
-      )}
-      {generating && (
-        <div className="xx-gen-progress-card">
-          <div className="xx-gen-progress-header">
-            <div className="xx-gen-progress-icon">
-              <LoadingOutlined />
-            </div>
-            <div className="xx-gen-progress-info">
-              <div className="xx-gen-progress-phase">
-                {getGenerationPhase(progress).icon} {getGenerationPhase(progress).label}
-              </div>
-              <div className="xx-gen-progress-sub">预计还需 1-2 分钟，请稍候…</div>
-            </div>
-            <div className="xx-gen-progress-percent">{Math.round(progress)}%</div>
-          </div>
-          <div className="xx-gen-progress-bar">
-            <div
-              className="xx-gen-progress-bar-fill"
-              style={{ width: `${Math.min(Math.round(progress), 100)}%` }}
-            />
-          </div>
-          <div className="xx-gen-progress-tip">
-            💡 生成过程中可以切换到其他页面操作，完成后会自动通知
-          </div>
-        </div>
-      )}
-      {generated && !generating && (
-        <div className="xx-gen-success-card">
-          <div className="xx-gen-success-icon">
-            <CheckCircleFilled style={{ fontSize: 32, color: "#52c41a" }} />
-          </div>
-          <div className="xx-gen-success-info">
-            <div className="xx-gen-success-title">视频生成完成！</div>
-            <div className="xx-gen-success-sub">
-              共生成 {generatedVideos.length} 条视频，可在右侧预览或前往成片库查看
-            </div>
-          </div>
-          <button
-            type="button"
-            className="xx-btn xx-btn-primary xx-btn-sm"
-            onClick={onScrollToPreview}
-          >
-            查看结果
-          </button>
-        </div>
-      )}
-      {generateError && !generating && (
-        <div className="xx-gen-error-card">
-          <div className="xx-gen-error-icon">
-            <CloseCircleOutlined style={{ fontSize: 28, color: "#ef4444" }} />
-          </div>
-          <div className="xx-gen-error-info">
-            <div className="xx-gen-error-title">生成失败</div>
-            <div className="xx-gen-error-msg">
-              {typeof generateError === "string" ? generateError : JSON.stringify(generateError)}
-            </div>
-          </div>
-          <div style={{ display: "flex", gap: 8 }}>
-            <button type="button" className="xx-btn xx-btn-primary xx-btn-sm" onClick={onRetry}>
-              🔄 重试
-            </button>
-            <button
-              type="button"
-              className="xx-btn xx-btn-ghost xx-btn-sm"
-              onClick={onDismissError}
-            >
-              知道了
-            </button>
-          </div>
-        </div>
-      )}
-    </div>
-  )
-}
-
-export default GenerationStatus

apps/web/src/pages/generate/generate.css

@@ -117,10 +117,6 @@
   grid-template-columns: 1fr;
 }
 
-.xx-generate-layout.full-width .xx-generate-right-col {
-  display: none;
-}
-
 /* ============================================================
    左侧表单区 generate-form
    ============================================================ */
@@ -2323,28 +2319,6 @@
    生成结果（右侧）
    ================================================================ */
 
-.xx-generate-right-col {
-  display: flex;
-  flex-direction: column;
-  align-items: center;
-  gap: 16px;
-}
-
-/* ── 内联视频播放器（右侧） ── */
-.xx-inline-video-player {
-  width: 100%;
-  max-width: 320px;
-  background: var(--bg-surface, #fff);
-  border: 1px solid var(--border-primary, #e2e8f0);
-  border-radius: 16px;
-  padding: 16px;
-  box-shadow: 0 2px 8px rgba(0, 0, 0, 0.04);
-}
-
-.xx-inline-video-player video {
-  background: #000;
-}
-
 .xx-preview-header {
   display: flex;
   align-items: center;
@@ -2710,11 +2684,12 @@
 
 /* ── 封面设置区域改造样式 ── */
 
-/* 封面操作按钮区 */
+/* 封面操作按钮区（单视频全宽页居中） */
 .xx-cover-actions {
   display: flex;
   gap: 12px;
   margin-bottom: 12px;
+  justify-content: center;
 }
 
 /* 已选模板文字 */
@@ -3202,10 +3177,11 @@
   margin: 0;
 }
 
-/* ── 批量封面网格 ── */
+/* ── 批量封面网格（单卡/少卡时居中排列，卡片限宽不拉伸） ── */
 .xx-cover-grid {
   display: grid;
-  grid-template-columns: repeat(auto-fill, minmax(180px, 1fr));
+  grid-template-columns: repeat(auto-fill, minmax(180px, 220px));
+  justify-content: center;
   gap: 16px;
 }
 

apps/web/src/pages/generate/hooks/step2-materials/useMaterialLibrary.ts

@@ -8,11 +8,22 @@ import type { AssetItem } from "@/api/assets"
  * 管理素材库列表、当前选中库、素材列表加载
  */
 export function useMaterialLibrary() {
-  /* ── 素材库数据 API ── */
-  const { data: libraries = [] } = useQuery({
-    queryKey: ["asset-libraries"],
-    queryFn: getAssetLibraries,
+  /* ── 素材库数据 API ──
+   * Step2 是视频选片，只拉取 kind=video 的素材库（#1777）：
+   * 后端按 kind 查询参数过滤，前端 getAssetLibraries("video") 再兜底过滤一次，
+   * 避免配音库(voice)/图片库(image) 混进「选择视频库」下拉。
+   * queryKey 带 kind，与素材管理页/配音页的 ["asset-libraries"] 全量缓存隔离。
+   */
+  const { data: allLibraries = [] } = useQuery({
+    queryKey: ["asset-libraries", "video"],
+    queryFn: () => getAssetLibraries("video"),
+    staleTime: 60_000,
   })
+  // 前端兜底过滤：仅保留 kind=video 的素材库（后端按 kind 查询参数过滤）
+  const libraries = useMemo(
+    () => allLibraries.filter((lib) => lib.kind === "video"),
+    [allLibraries],
+  )
   const [selectedLibraryId, setSelectedLibraryId] = useState<string>("")
 
   // 自动选中第一个视频库

apps/web/src/pages/generate/hooks/useGenerateFormState/index.ts

@@ -40,6 +40,8 @@ export interface GenerateFormState {
   selectedTemplate: string
   setSelectedTemplate: (id: string) => void
   userTemplates: EditingTemplate[]
+  /** 当前选中模板在创建片段时被判失效（404/400/422）后的运行时自动回退 */
+  handleInvalidTemplate: () => boolean
 
   /* 素材 */
   selectedMaterials: string[]
@@ -131,7 +133,8 @@ export const useGenerateFormState = (): GenerateFormState => {
   const [currentStep, setCurrentStep] = useState(1)
 
   /* ── 模板选择 ── */
-  const { selectedTemplate, setSelectedTemplate, userTemplates } = useTemplateSelection()
+  const { selectedTemplate, setSelectedTemplate, userTemplates, handleInvalidTemplate } =
+    useTemplateSelection()
 
   /* ── source_edit_plan_id：仅取 URL 参数，无则 null 让后端兜底 ── */
   // selectedTemplate 是模板 ID 而非 edit_plan_id，不能混淆；
@@ -228,6 +231,7 @@ export const useGenerateFormState = (): GenerateFormState => {
     selectedTemplate,
     setSelectedTemplate,
     userTemplates,
+    handleInvalidTemplate,
     selectedMaterials,
     setSelectedMaterials,
     materialMode,

apps/web/src/pages/generate/hooks/useGenerateFormState/templateFallback.ts

@@ -0,0 +1,81 @@
+/**
+ * 失效模板判定与自动回退工具（#1777）
+ *
+ * 背景：用户进入生成页后，之前选中的模板可能已被删除、或从未配置片段。
+ * 调用片段相关接口（PUT/POST /templates/{id}/editor/clips[...]/from-assets）时：
+ *   - 模板不存在 → 后端返回 404（并行工单 #1774 把「模板不存在」统一为该状态码）
+ *   - 模板无片段配置 → 当前部分场景返回 400（detail 含「片段配置」），
+ *     参数校验类错误返回 422
+ * 这三类响应都说明「当前选中的模板不可用于生成」，应清除失效选择并自动切换到
+ * 第一个有效模板，同时提示用户，而不是让页面卡死、无任何反馈。
+ */
+import type { EditingTemplate } from "@/api/editing-planner"
+
+/** 失效模板相关的 HTTP 状态码 */
+const INVALID_TEMPLATE_STATUSES = new Set([404, 400, 422])
+
+/**
+ * 从任意抛出值（axios 错误）提取 HTTP 状态码。
+ * 非 axios 错误 / 无响应时返回 null。
+ */
+export function getHttpStatus(err: unknown): number | null {
+  if (!err || typeof err !== "object") return null
+  const status = (err as { response?: { status?: number }; status?: number })?.response?.status
+  return typeof status === "number" ? status : null
+}
+
+/** 安全提取后端错误文本（detail/message/msg，422 数组也兜底拼一下） */
+function extractErrorText(err: unknown): string {
+  if (!err || typeof err !== "object") return ""
+  const data = (err as { response?: { data?: unknown } })?.response?.data
+  if (!data) return ""
+  try {
+    const text = JSON.stringify(data)
+    return typeof text === "string" ? text : ""
+  } catch {
+    return ""
+  }
+}
+
+/**
+ * 判断一次 clips/from-assets 请求失败是否因为「模板失效」。
+ *
+ * 严格判定，避免把无关的 400/422（例如素材参数问题）误判为模板失效：
+ *   - 404：模板/编辑计划不存在，一定是模板失效
+ *   - 400：仅当后端文本明确提到「片段配置」（无片段配置无法创建片段）才判定
+ *   - 422：参数校验类，from-assets 场景下命中「片段/segments」相关字段才判定
+ */
+export function isInvalidTemplateError(err: unknown): boolean {
+  const status = getHttpStatus(err)
+  if (status === null || !INVALID_TEMPLATE_STATUSES.has(status)) return false
+  if (status === 404) return true
+
+  const text = extractErrorText(err)
+  if (status === 400) {
+    // 后端当前返回：「模板没有片段配置，无法创建片段」
+    return /片段配置|没有片段|无片段|segments?|clip.*config/i.test(text)
+  }
+  // 422：FastAPI 校验错误，命中模板片段相关字段
+  return /segment|clip|片段|模板/i.test(text)
+}
+
+/**
+ * 判断模板是否可用于生成（有效模板）。
+ *
+ * 有效 = 处于激活态（is_active !== false，字段缺失视为 true 兼容旧后端）
+ *        且至少配置了一个片段。
+ * 与后端 valid_only 过滤口径保持一致（#1769/#1772），这里是前端双保险。
+ */
+export function isValidTemplate(template: EditingTemplate | null | undefined): boolean {
+  if (!template) return false
+  if (template.is_active === false) return false
+  return (template.segments?.length ?? 0) > 0
+}
+
+/** 从模板列表中取出第一个有效模板，没有则返回 null */
+export function findFirstValidTemplate(
+  templates: EditingTemplate[] | null | undefined,
+): EditingTemplate | null {
+  if (!Array.isArray(templates)) return null
+  return templates.find(isValidTemplate) ?? null
+}

apps/web/src/pages/generate/hooks/useGenerateFormState/useTemplateSelection.ts

@@ -1,22 +1,89 @@
-import { useState, useEffect } from "react"
+import { useState, useEffect, useRef, useCallback, useMemo } from "react"
 import { useQuery } from "@tanstack/react-query"
+import { message } from "antd"
 import { getEditingTemplates } from "@/api/editing-planner"
 import type { EditingTemplate } from "@/api/editing-planner"
+import { findFirstValidTemplate, isValidTemplate } from "./templateFallback"
+
+/** 失效模板自动切换的提示文案 */
+export const INVALID_TEMPLATE_FALLBACK_TOAST = "原模板已失效，已自动切换"
 
 export function useTemplateSelection() {
+  // selectedTemplate 纯内存状态，绝不写入 localStorage/sessionStorage/URL，
+  // 因此失效模板 ID 不会被持久化、刷新后也不会恢复（#1777 要求 4）
   const [selectedTemplate, setSelectedTemplate] = useState("")
-  const { data: userTemplates = [] } = useQuery<EditingTemplate[]>({
+
+  const { data: allTemplates = [] } = useQuery<EditingTemplate[]>({
     queryKey: ["generate-templates"],
-    queryFn: () => getEditingTemplates(),
+    // valid_only：后端过滤掉没有片段配置的无效模板（#1769/#1772）。
+    // 旧后端忽略该 query 参数时，下方 isValidTemplate 前端兜底再过滤一次。
+    queryFn: () => getEditingTemplates({ validOnly: true }),
     staleTime: 60_000,
   })
 
-  /* 模板加载完成后自动选中第一个 */
-  useEffect(() => {
-    if (userTemplates.length > 0 && !selectedTemplate) {
-      setSelectedTemplate(userTemplates[0].id)
-    }
-  }, [userTemplates, selectedTemplate])
+  // 双保险：后端 valid_only 已过滤，前端再按 is_active + segments 兜底，
+  // 保证下拉/自动选择只包含可用于生成的有效模板。
+  // 用 useMemo 缓存引用，避免每次渲染都 .filter 创建新数组，
+  // 导致下游 useTitleCoverSync effect 无限触发、覆盖用户手动修改（#1789）
+  const validTemplates = useMemo(() => allTemplates.filter(isValidTemplate), [allTemplates])
+  const userTemplates = validTemplates
 
-  return { selectedTemplate, setSelectedTemplate, userTemplates }
+  // 用 ref 持有最新值，供稳定回调 handleInvalidTemplate 使用（避免闭包拿到旧值）
+  const templatesRef = useRef(validTemplates)
+  templatesRef.current = validTemplates
+  const selectedRef = useRef(selectedTemplate)
+  selectedRef.current = selectedTemplate
+  // 已提示过失效的模板 ID，避免用户停留在失效模板上时 clips 防抖请求反复弹 toast；
+  // 用户手动切换/成功切换后重置，保证下一个失效模板仍能提示
+  const fallbackNotifiedRef = useRef<string>("")
+
+  /* 自动选择：模板加载完成且当前未选中时，自动选中第一个有效模板。
+   * 用户手动选择（setSelectedTemplate 被显式调用）后 selectedTemplate 非空，
+   * 本 effect 直接 return，绝不覆盖用户的手动选择（#1777 要求 4：手动优先）。 */
+  useEffect(() => {
+    if (selectedTemplate) return
+    const firstValid = validTemplates[0]
+    if (firstValid) {
+      setSelectedTemplate(firstValid.id)
+    }
+  }, [validTemplates, selectedTemplate])
+
+  /** 用户手动选择模板：优先级最高，重置失效提示标记 */
+  const handleSelectTemplate = useCallback((id: string) => {
+    fallbackNotifiedRef.current = ""
+    setSelectedTemplate(id)
+  }, [])
+
+  /**
+   * 运行时失效回退（#1777 要求 3）：
+   * 创建片段接口返回 404（模板不存在）/ 400/422（模板无片段配置）时调用。
+   *  - 清除失效选择，自动切换到第一个有效模板，并 toast 提示；
+   *  - 没有有效模板时清空选择，Step1 展示明确的「暂无可用模板」空状态引导，
+   *    不让用户卡在失效模板上。
+   * 返回 true 表示已按「模板失效」处理（调用方可据此静默原始错误提示）。
+   */
+  const handleInvalidTemplate = useCallback((): boolean => {
+    const current = selectedRef.current
+    // 同一个失效模板只提示一次（clips 防抖 effect 在素材/模板变化时会反复触发）
+    if (current && fallbackNotifiedRef.current === current) return true
+
+    const fallback = findFirstValidTemplate(templatesRef.current)
+    fallbackNotifiedRef.current = current || "__empty__"
+    if (fallback) {
+      setSelectedTemplate(fallback.id)
+      message.warning(INVALID_TEMPLATE_FALLBACK_TOAST)
+    } else {
+      // 没有任何有效模板：清空选择，交由 Step1 空状态引导用户去模板编辑器创建
+      setSelectedTemplate("")
+      message.warning("当前没有可用模板，请先在「模板编辑器」中创建并配置片段")
+    }
+    return true
+  }, [])
+
+  return {
+    selectedTemplate,
+    setSelectedTemplate: handleSelectTemplate,
+    userTemplates,
+    handleInvalidTemplate,
+  }
 }

apps/web/src/pages/generate/hooks/useGenerateFormState/useTitleCoverSync.ts

@@ -1,4 +1,4 @@
-import { useEffect } from "react"
+import { useEffect, useRef } from "react"
 import type { TitleSettings } from "../../types"
 import type { CoverConfig } from "../../types/cover"
 import type { EditingTemplate } from "@/api/editing-planner"
@@ -11,7 +11,11 @@ interface UseTitleCoverSyncOptions {
 }
 
 /**
- * 当选中模板变化时，自动同步标题和封面配置
+ * 当选中模板变化时，自动同步标题和封面配置。
+ *
+ * #1789 修复：userTemplates 用 ref 持有最新值，不放入依赖数组。
+ * 否则每次渲染 .filter() 创建的新数组引用都会触发 effect，
+ * 从模板 title_config 覆盖用户手动修改（如字号滑块拖动），导致回弹。
  */
 export function useTitleCoverSync({
   selectedTemplate,
@@ -19,8 +23,12 @@ export function useTitleCoverSync({
   setTitleSettings,
   setCoverSettings,
 }: UseTitleCoverSyncOptions) {
+  // 用 ref 持有最新 userTemplates，避免数组引用变化导致 effect 反复触发
+  const templatesRef = useRef(userTemplates)
+  templatesRef.current = userTemplates
+
   useEffect(() => {
-    const tpl = userTemplates.find((t) => t.id === selectedTemplate)
+    const tpl = templatesRef.current.find((t) => t.id === selectedTemplate)
     if (tpl?.title_config) {
       setTitleSettings((prev: TitleSettings) => ({
         ...prev,
@@ -43,5 +51,7 @@ export function useTitleCoverSync({
         thumbnail_url: tpl.cover_config!.thumbnail_url || prev.thumbnail_url,
       }))
     }
-  }, [selectedTemplate, userTemplates, setTitleSettings, setCoverSettings])
+    // eslint-disable-next-line react-hooks/exhaustive-deps
+  }, [selectedTemplate, setTitleSettings, setCoverSettings])
+  // ↑ 移除 userTemplates，只在 selectedTemplate 真正变化时触发
 }

apps/web/src/pages/generate/hooks/useStep2Materials.ts

@@ -10,6 +10,7 @@ import { updateEditPlanClips, createClipsFromAssets, getEditPlanClips } from "@/
 import { useMaterialLibrary } from "./step2-materials/useMaterialLibrary"
 import { useSmartMatch } from "./step2-materials/useSmartMatch"
 import { useDraftAutoSave } from "./useDraftAutoSave"
+import { isInvalidTemplateError } from "./useGenerateFormState/templateFallback"
 
 interface UseStep2MaterialsProps {
   materialMode: "manual" | "auto"
@@ -24,6 +25,8 @@ interface UseStep2MaterialsProps {
   templateSegments?: TemplateSegment[]
   /** 服务端 clips 创建成功后的回调，用于通知预览播放器 */
   onServerClipsChange?: (clips: EditPlanClip[]) => void
+  /** 当前模板创建片段返回 404/400/422（模板失效）时的自动回退回调（#1777） */
+  onTemplateInvalid?: () => boolean
 }
 
 export function useStep2Materials({
@@ -36,6 +39,7 @@ export function useStep2Materials({
   selectedTemplate,
   templateSegments,
   onServerClipsChange,
+  onTemplateInvalid,
 }: UseStep2MaterialsProps) {
   const {
     libraries,
@@ -102,6 +106,8 @@ export function useStep2Materials({
   selectedTemplateRef.current = selectedTemplate
   const onServerClipsChangeRef = useRef(onServerClipsChange)
   onServerClipsChangeRef.current = onServerClipsChange
+  const onTemplateInvalidRef = useRef(onTemplateInvalid)
+  onTemplateInvalidRef.current = onTemplateInvalid
 
   useEffect(() => {
     const tid = selectedTemplateRef.current
@@ -123,11 +129,12 @@ export function useStep2Materials({
       const requiredClipsCount = segs.length > 0 ? segs.length : undefined
 
       try {
-        // 1. 清空旧片段
-        await updateEditPlanClips(tid, [], controller.signal)
+        // 1. 清空旧片段（静默全局 toast：模板失效时由下方回退统一提示）
+        await updateEditPlanClips(tid, [], controller.signal, true)
         // 2. 调用后端 from-assets 接口创建片段（异步秒级返回，60s 超时仅为兜底）
         await createClipsFromAssets(tid, ids, "main", requiredClipsCount, {
           signal: controller.signal,
+          silentErrorToast: true,
         })
         // 3. 获取服务端生成的 clips（含 start_time/duration），供预览播放器使用
         const clipList = await getEditPlanClips(tid, { limit: 500 })
@@ -146,6 +153,14 @@ export function useStep2Materials({
           message.error("智能选片失败，请重试")
           return
         }
+        // 模板失效（404 模板不存在 / 400/422 无片段配置）：
+        // 清空失效选择并自动切到第一个有效模板 + toast，避免页面卡死无提示（#1777）
+        if (isInvalidTemplateError(err)) {
+          console.warn("[useStep2Materials] 当前模板已失效，触发自动回退:", err)
+          onServerClipsChangeRef.current?.([])
+          onTemplateInvalidRef.current?.()
+          return
+        }
         console.warn("[useStep2Materials] 写入 clips 失败:", err)
       }
     }, 800)

apps/web/src/pages/voice-materials/hooks/useVoiceMaterials/actions/useVoiceUpload.ts

@@ -41,7 +41,7 @@ export function useVoiceUpload({ voiceLibrary, createLibMutation }: UseVoiceUplo
           }
           const libs = await queryClient.fetchQuery({
             queryKey: ["asset-libraries"],
-            queryFn: getAssetLibraries,
+            queryFn: () => getAssetLibraries(),
           })
           lib = libs.find((l: AssetLibraryItem) => l.kind === "voice")
           if (!lib) throw new Error("无法创建配音库")

apps/web/src/pages/voice-materials/hooks/useVoiceMaterials/useVoiceMaterialData.ts

@@ -24,7 +24,7 @@ export function useVoiceMaterialData({ keyword, gender, tagIds }: UseVoiceMateri
   // ── 获取 voice 类型素材库 ─────────────────────────────────
   const { data: libraries = [] } = useQuery({
     queryKey: ["asset-libraries"],
-    queryFn: getAssetLibraries,
+    queryFn: () => getAssetLibraries(),
     staleTime: 60_000,
   })
 

apps/web/src/pages/voices/hooks/useVoiceUpload.ts

@@ -26,7 +26,7 @@ export function useVoiceUpload({ showToast }: UseVoiceUploadProps) {
         /* 获取或创建默认配音库 */
         const libs = await queryClient.fetchQuery({
           queryKey: ["asset-libraries"],
-          queryFn: getAssetLibraries,
+          queryFn: () => getAssetLibraries(),
         })
         const lib = libs.find((l) => l.kind === "voice")
         if (!lib) throw new Error("配音库不存在，请先在配音库页面创建")

apps/web/src/test/pages/generate/templateFallback.test.ts

@@ -0,0 +1,140 @@
+/**
+ * 失效模板判定/回退纯函数单测（#1777）
+ */
+import { describe, it, expect } from "vitest"
+import type { EditingTemplate } from "@/api/editing-planner"
+import {
+  getHttpStatus,
+  isInvalidTemplateError,
+  isValidTemplate,
+  findFirstValidTemplate,
+} from "@/pages/generate/hooks/useGenerateFormState/templateFallback"
+
+function makeTemplate(partial: Partial<EditingTemplate> & { id: string }): EditingTemplate {
+  return {
+    name: partial.id,
+    mode: "pip",
+    category: "默认",
+    tags: [],
+    title_config: {
+      ai_auto_select: false,
+      content: "",
+      font_preset: "",
+      font_color: "",
+      font_size: 28,
+      position: "top",
+    },
+    subtitle_config: {
+      enabled: true,
+      position: "bottom",
+      font: "",
+      color: "",
+      size: 20,
+      animation: "",
+    },
+    bgm_config: { enabled: false, music_id: "" },
+    segments: [{ segment_order: 0, material_type: null }],
+    is_active: true,
+    created_at: "",
+    updated_at: "",
+    ...partial,
+  } as EditingTemplate
+}
+
+function axiosError(status: number, data?: unknown) {
+  return { isAxiosError: true, response: { status, data } }
+}
+
+describe("getHttpStatus", () => {
+  it("提取 axios 错误的 HTTP 状态码", () => {
+    expect(getHttpStatus(axiosError(404))).toBe(404)
+    expect(getHttpStatus(axiosError(400))).toBe(400)
+  })
+  it("非 axios/无响应错误返回 null", () => {
+    expect(getHttpStatus(new Error("network"))).toBeNull()
+    expect(getHttpStatus(null)).toBeNull()
+    expect(getHttpStatus(undefined)).toBeNull()
+    expect(getHttpStatus({ isAxiosError: true })).toBeNull()
+  })
+})
+
+describe("isInvalidTemplateError", () => {
+  it("404 始终判定为模板失效（模板不存在）", () => {
+    expect(isInvalidTemplateError(axiosError(404))).toBe(true)
+    expect(isInvalidTemplateError(axiosError(404, { detail: "Not Found" }))).toBe(true)
+  })
+
+  it("400 且后端文案提到「片段配置」判定为模板无片段配置", () => {
+    expect(
+      isInvalidTemplateError(axiosError(400, { detail: "模板没有片段配置，无法创建片段" })),
+    ).toBe(true)
+  })
+
+  it("400 但文案与片段配置无关 → 不误判", () => {
+    expect(isInvalidTemplateError(axiosError(400, { detail: "素材参数错误" }))).toBe(false)
+  })
+
+  it("422 命中片段/模板字段判定为失效", () => {
+    expect(
+      isInvalidTemplateError(
+        axiosError(422, { detail: [{ loc: ["body", "segments"], msg: "field required" }] }),
+      ),
+    ).toBe(true)
+  })
+
+  it("其他状态码（401/403/500/超时/网络）不判定为模板失效", () => {
+    expect(isInvalidTemplateError(axiosError(401))).toBe(false)
+    expect(isInvalidTemplateError(axiosError(403))).toBe(false)
+    expect(isInvalidTemplateError(axiosError(500))).toBe(false)
+    expect(isInvalidTemplateError({ code: "ECONNABORTED", message: "timeout of 60000ms" })).toBe(
+      false,
+    )
+    expect(isInvalidTemplateError(new Error("Network Error"))).toBe(false)
+  })
+})
+
+describe("isValidTemplate", () => {
+  it("有片段且未被标记 inactive → 有效", () => {
+    expect(isValidTemplate(makeTemplate({ id: "t1" }))).toBe(true)
+  })
+  it("segments 为空 → 无效（无片段配置）", () => {
+    expect(isValidTemplate(makeTemplate({ id: "t2", segments: [] }))).toBe(false)
+  })
+  it("is_active=false → 无效（已停用/删除）", () => {
+    expect(isValidTemplate(makeTemplate({ id: "t3", is_active: false }))).toBe(false)
+  })
+  it("is_active 字段缺失时视为有效（兼容旧后端）", () => {
+    const t = makeTemplate({ id: "t4" })
+    delete (t as Partial<EditingTemplate>).is_active
+    expect(isValidTemplate(t)).toBe(true)
+  })
+  it("null/undefined → 无效", () => {
+    expect(isValidTemplate(null)).toBe(false)
+    expect(isValidTemplate(undefined)).toBe(false)
+  })
+})
+
+describe("findFirstValidTemplate", () => {
+  it("跳过无效模板，返回第一个有效模板", () => {
+    const list = [
+      makeTemplate({ id: "empty", segments: [] }),
+      makeTemplate({ id: "inactive", is_active: false }),
+      makeTemplate({ id: "valid1" }),
+      makeTemplate({ id: "valid2" }),
+    ]
+    expect(findFirstValidTemplate(list)?.id).toBe("valid1")
+  })
+  it("全部无效 → null（用于空状态引导）", () => {
+    expect(
+      findFirstValidTemplate([
+        makeTemplate({ id: "a", segments: [] }),
+        makeTemplate({ id: "b", is_active: false }),
+      ]),
+    ).toBeNull()
+  })
+  it("空数组/null → null", () => {
+    expect(findFirstValidTemplate([])).toBeNull()
+    expect(findFirstValidTemplate(null)).toBeNull()
+    expect(findFirstValidTemplate(undefined)).toBeNull()
+  })
+})

apps/web/src/test/pages/generate/useMaterialLibrary.test.tsx

@@ -0,0 +1,74 @@
+/**
+ * useMaterialLibrary Hook 单测（#1777）
+ * - Step2 视频库选择器只拉取 kind=video 的素材库，配音库(voice)/图片库(image) 不混入
+ * - 自动选中第一个视频库
+ */
+import { describe, it, expect, vi, beforeEach } from "vitest"
+import { renderHook, waitFor } from "@testing-library/react"
+import { QueryClient, QueryClientProvider } from "@tanstack/react-query"
+import type { ReactNode } from "react"
+import type { AssetItem, AssetLibraryItem } from "@/api/assets"
+
+vi.mock("@/api/assets", () => ({
+  getAssetLibraries: vi.fn(),
+  getAssets: vi.fn(),
+  isAssetUsable: vi.fn(() => true),
+}))
+
+import { getAssetLibraries, getAssets } from "@/api/assets"
+import { useMaterialLibrary } from "@/pages/generate/hooks/step2-materials/useMaterialLibrary"
+
+const mockGetLibraries = vi.mocked(getAssetLibraries)
+const mockGetAssets = vi.mocked(getAssets)
+
+function lib(id: string, kind: AssetLibraryItem["kind"], name = id): AssetLibraryItem {
+  return { id, name, kind }
+}
+
+function createWrapper() {
+  const queryClient = new QueryClient({
+    defaultOptions: { queries: { retry: false, gcTime: 0 } },
+  })
+  return ({ children }: { children: ReactNode }) =>
+    (<QueryClientProvider client={queryClient}>{children}</QueryClientProvider>) as ReactNode
+}
+
+beforeEach(() => {
+  vi.clearAllMocks()
+  mockGetAssets.mockResolvedValue({ items: [] as AssetItem[], total: 0 })
+})
+
+describe("useMaterialLibrary (#1777 kind=video 过滤)", () => {
+  it("按 kind=video 拉取素材库（后端参数过滤）", async () => {
+    mockGetLibraries.mockResolvedValueOnce([lib("v1", "video")])
+    renderHook(() => useMaterialLibrary(), { wrapper: createWrapper() })
+
+    await waitFor(() => expect(mockGetLibraries).toHaveBeenCalledTimes(1))
+    expect(mockGetLibraries).toHaveBeenCalledWith("video")
+  })
+
+  it("下拉库列表只包含视频库（自动选中第一个视频库）", async () => {
+    mockGetLibraries.mockResolvedValueOnce([
+      lib("voice-1", "voice"),
+      lib("img-1", "image"),
+      lib("video-1", "video"),
+      lib("video-2", "video"),
+    ])
+    const { result } = renderHook(() => useMaterialLibrary(), { wrapper: createWrapper() })
+
+    await waitFor(() => expect(result.current.libraries).toHaveLength(2))
+    expect(result.current.libraries.map((l) => l.id)).toEqual(["video-1", "video-2"])
+    expect(result.current.libraries.every((l) => l.kind === "video")).toBe(true)
+    // 自动选中第一个视频库
+    expect(result.current.selectedLibraryId).toBe("video-1")
+  })
+
+  it("没有视频库时库列表为空且不自动选中（UI 展示空状态）", async () => {
+    mockGetLibraries.mockResolvedValueOnce([lib("voice-1", "voice"), lib("img-1", "image")])
+    const { result } = renderHook(() => useMaterialLibrary(), { wrapper: createWrapper() })
+
+    await waitFor(() => expect(mockGetLibraries).toHaveBeenCalled())
+    expect(result.current.libraries).toEqual([])
+    expect(result.current.selectedLibraryId).toBe("")
+  })
+})

apps/web/src/test/pages/generate/useTemplateSelection.test.tsx

@@ -0,0 +1,174 @@
+/**
+ * useTemplateSelection Hook 单测（#1777）
+ * - 自动选择跳过无片段/inactive 模板，只选第一个有效模板
+ * - 传 validOnly=true 给后端
+ * - 用户手动选择优先，自动逻辑不覆盖
+ * - handleInvalidTemplate：失效时自动切到第一个有效模板 + toast；无有效模板时清空
+ * - selectedTemplate 仅内存态，不写入 localStorage/sessionStorage
+ */
+import { describe, it, expect, vi, beforeEach, afterEach } from "vitest"
+import { renderHook, waitFor, act } from "@testing-library/react"
+import { QueryClient, QueryClientProvider } from "@tanstack/react-query"
+import type { ReactNode } from "react"
+
+// antd message mock（拦截 toast）——vi.hoisted 保证 mock 工厂可引用
+const { messageMock } = vi.hoisted(() => ({
+  messageMock: {
+    warning: vi.fn(),
+    error: vi.fn(),
+    success: vi.fn(),
+    info: vi.fn(),
+    loading: vi.fn(() => vi.fn()),
+  },
+}))
+vi.mock("antd", () => ({ message: messageMock }))
+
+vi.mock("@/api/editing-planner", () => ({
+  getEditingTemplates: vi.fn(),
+}))
+
+import { getEditingTemplates } from "@/api/editing-planner"
+import type { EditingTemplate } from "@/api/editing-planner"
+import { useTemplateSelection } from "@/pages/generate/hooks/useGenerateFormState/useTemplateSelection"
+
+const mockGetTemplates = vi.mocked(getEditingTemplates)
+
+function tpl(id: string, partial: Partial<EditingTemplate> = {}): EditingTemplate {
+  return {
+    id,
+    name: id,
+    mode: "pip",
+    category: "默认",
+    tags: [],
+    title_config: {
+      ai_auto_select: false,
+      content: "",
+      font_preset: "",
+      font_color: "",
+      font_size: 28,
+      position: "top",
+    },
+    subtitle_config: {
+      enabled: true,
+      position: "bottom",
+      font: "",
+      color: "",
+      size: 20,
+      animation: "",
+    },
+    bgm_config: { enabled: false, music_id: "" },
+    segments: [{ segment_order: 0, material_type: null }],
+    is_active: true,
+    created_at: "",
+    updated_at: "",
+    ...partial,
+  } as EditingTemplate
+}
+
+function createWrapper() {
+  const queryClient = new QueryClient({
+    defaultOptions: { queries: { retry: false, gcTime: 0 } },
+  })
+  return ({ children }: { children: ReactNode }) =>
+    (<QueryClientProvider client={queryClient}>{children}</QueryClientProvider>) as ReactNode
+}
+
+beforeEach(() => {
+  vi.clearAllMocks()
+  localStorage.clear()
+  sessionStorage.clear()
+})
+
+afterEach(() => {
+  localStorage.clear()
+  sessionStorage.clear()
+})
+
+describe("useTemplateSelection (#1777)", () => {
+  it("请求模板时传 validOnly=true，并自动选中第一个有片段的有效模板", async () => {
+    mockGetTemplates.mockResolvedValueOnce([
+      tpl("empty", { segments: [] }),
+      tpl("inactive", { is_active: false }),
+      tpl("valid-a"),
+      tpl("valid-b"),
+    ])
+    const { result } = renderHook(() => useTemplateSelection(), { wrapper: createWrapper() })
+
+    await waitFor(() => expect(result.current.selectedTemplate).toBe("valid-a"))
+    expect(mockGetTemplates).toHaveBeenCalledWith({ validOnly: true })
+    // 暴露给 UI 的 userTemplates 已过滤掉无效模板
+    expect(result.current.userTemplates.map((t) => t.id)).toEqual(["valid-a", "valid-b"])
+  })
+
+  it("列表全部无效时 selectedTemplate 为空（交空状态引导），不选中失效模板", async () => {
+    mockGetTemplates.mockResolvedValueOnce([
+      tpl("empty", { segments: [] }),
+      tpl("inactive", { is_active: false }),
+    ])
+    const { result } = renderHook(() => useTemplateSelection(), { wrapper: createWrapper() })
+    await waitFor(() => expect(mockGetTemplates).toHaveBeenCalled())
+    // 给 effect 一个 tick
+    await waitFor(() => expect(result.current.selectedTemplate).toBe(""))
+    expect(result.current.userTemplates).toHaveLength(0)
+  })
+
+  it("用户手动选择优先：自动逻辑不会覆盖手动选择", async () => {
+    mockGetTemplates.mockResolvedValueOnce([tpl("a"), tpl("b")])
+    const { result } = renderHook(() => useTemplateSelection(), { wrapper: createWrapper() })
+    await waitFor(() => expect(result.current.selectedTemplate).toBe("a"))
+
+    act(() => result.current.setSelectedTemplate("b"))
+    expect(result.current.selectedTemplate).toBe("b")
+
+    // 重新渲染 / refetch 后仍保持用户的手动选择
+    await waitFor(() => expect(result.current.selectedTemplate).toBe("b"))
+  })
+
+  it("handleInvalidTemplate：当前模板失效时自动切到第一个有效模板并 toast", async () => {
+    mockGetTemplates.mockResolvedValueOnce([tpl("bad", { segments: [] }), tpl("good")])
+    const { result } = renderHook(() => useTemplateSelection(), { wrapper: createWrapper() })
+    // 自动选中有效模板 good（bad 无片段不会被自动选中）
+    await waitFor(() => expect(result.current.selectedTemplate).toBe("good"))
+    messageMock.warning.mockClear()
+
+    // 模拟运行时用户停留在一个已失效的模板 id（外部/草稿态），触发回退
+    act(() => result.current.setSelectedTemplate("stale-id"))
+    expect(result.current.selectedTemplate).toBe("stale-id")
+
+    act(() => {
+      const handled = result.current.handleInvalidTemplate()
+      expect(handled).toBe(true)
+    })
+    await waitFor(() => expect(result.current.selectedTemplate).toBe("good"))
+    expect(messageMock.warning).toHaveBeenCalledWith("原模板已失效，已自动切换")
+  })
+
+  it("handleInvalidTemplate：无有效模板时清空选择并提示去创建", async () => {
+    mockGetTemplates.mockResolvedValueOnce([tpl("bad", { segments: [] })])
+    const { result } = renderHook(() => useTemplateSelection(), { wrapper: createWrapper() })
+    await waitFor(() => expect(result.current.userTemplates).toHaveLength(0))
+
+    act(() => result.current.setSelectedTemplate("stale-id"))
+    act(() => {
+      result.current.handleInvalidTemplate()
+    })
+    await waitFor(() => expect(result.current.selectedTemplate).toBe(""))
+    expect(messageMock.warning).toHaveBeenCalledWith(expect.stringContaining("没有可用模板"))
+  })
+
+  it("失效模板 ID 不写入任何持久化存储", async () => {
+    mockGetTemplates.mockResolvedValueOnce([tpl("good")])
+    const { result } = renderHook(() => useTemplateSelection(), { wrapper: createWrapper() })
+    await waitFor(() => expect(result.current.selectedTemplate).toBe("good"))
+
+    act(() => result.current.setSelectedTemplate("stale-invalid-id"))
+    act(() => result.current.handleInvalidTemplate())
+
+    const ls = JSON.stringify(localStorage)
+    const ss = JSON.stringify(sessionStorage)
+    expect(ls).not.toContain("stale-invalid-id")
+    expect(ss).not.toContain("stale-invalid-id")
+    // URL 也不含
+    expect(window.location.href).not.toContain("stale-invalid-id")
+  })
+})

apps/worker/video_processing/bgm_mixer.py

@@ -39,6 +39,8 @@ class BGMConfig:
     sidechain_attack: float = 0.02  # 攻击时间
     sidechain_release: float = 0.5  # 释放时间
     sidechain_threshold: float = -25.0  # 触发阈值（dB）
+    audio_offset: float = 0.0  # BGM 段落起始偏移（秒），#1767 策略二
+    volume_adjust_db: float = 0.0  # 音量微调 dB（-3~+3），#1767 策略三
 
     @classmethod
     def from_config_dict(cls, bgm_path: str, config: dict) -> "BGMConfig":
@@ -54,6 +56,8 @@ class BGMConfig:
             sidechain_attack=float(config.get("sidechain_attack", 0.02)),
             sidechain_release=float(config.get("sidechain_release", 0.5)),
             sidechain_threshold=float(config.get("sidechain_threshold", -25.0)),
+            audio_offset=float(config.get("audio_offset", 0.0)),
+            volume_adjust_db=float(config.get("volume_adjust_db", 0.0)),
         )
 
 
@@ -84,14 +88,18 @@ def prepare_bgm_track(
         target_duration = 5.0  # 兜底
 
     bgm_dur = probe_duration(bgm.bgm_path)
-    needs_loop = bgm.loop_enabled and bgm_dur > 0 and bgm_dur < target_duration * 0.9
+    # #1767：seek 后有效时长 = 总时长 - 偏移
+    effective_dur = (
+        max(1.0, bgm_dur - bgm.audio_offset) if bgm.audio_offset > 0 and bgm_dur > bgm.audio_offset else bgm_dur
+    )
+    needs_loop = bgm.loop_enabled and effective_dur > 0 and effective_dur < target_duration * 0.9
 
     # 构建滤镜链
     filter_parts: list[str] = []
 
     if needs_loop:
-        # 计算需要循环多少次才能铺满
-        loop_count = max(1, int(target_duration / bgm_dur) + 2)
+        # 计算需要循环多少次才能铺满（基于 seek 后有效时长）
+        loop_count = max(1, int(target_duration / effective_dur) + 2)
         # aloop 滤镜：循环指定次数
         filter_parts.append(f"aloop=loop={loop_count}:size=0")
 
@@ -113,11 +121,30 @@ def prepare_bgm_track(
     filter_parts.append(f"atrim=0:{target_duration:.3f}")
     filter_parts.append("asetpts=N/SR/TB")  # 重置时间戳
 
-    filter_str = ",".join(filter_parts)
+    # #1767：BGM 段落差异化 — 使用 -ss 从偏移位置开始（seek 效率高，不读跳过部分）
+    seek_args: list[str] = []
+    if bgm.audio_offset > 0 and bgm_dur > bgm.audio_offset:
+        seek_args = ["-ss", f"{bgm.audio_offset:.3f}"]
+        logger.info("[bgm] #1767 audio_offset=%.1fs（段落差异化）", bgm.audio_offset)
+
+    # #1767：音量微调 — dB 转线性系数（10^(dB/20)）
+    db_adjust_filter = ""
+    if abs(bgm.volume_adjust_db) > 0.01:
+        linear_factor = 10.0 ** (bgm.volume_adjust_db / 20.0)
+        db_adjust_filter = f",volume={linear_factor:.4f}"
+        logger.info("[bgm] #1767 volume_adjust=%.0fdB → linear=%.4f", bgm.volume_adjust_db, linear_factor)
+
+    # #1767：追加 dB 微调到滤镜链末尾
+    if db_adjust_filter:
+        filter_str_base = ",".join(filter_parts)
+        filter_str = filter_str_base + db_adjust_filter
+    else:
+        filter_str = ",".join(filter_parts)
 
     command = [
         FFMPEG_BIN,
         "-y",
+        *seek_args,
         "-i",
         bgm.bgm_path,
         "-filter:a",

apps/worker/video_processing/ffmpeg_utils.py

@@ -59,13 +59,18 @@ def build_xfade_filter_chain(
     transitions: list[str],
     *,
     transition_duration: float = DEFAULT_TRANSITION_DURATION,
+    transition_durations: list[float] | None = None,
+    jitters: list[float] | None = None,
     output_label: str = "outv",
 ) -> tuple[str, float]:
+    """构建 xfade 转场滤镜链（re-export，#1766 增加逐转场时长与 jitter 支持）."""
     return _build_xfade_filter_chain_base(
         clip_durations,
         clip_video_labels,
         transitions,
         transition_duration=transition_duration,
+        transition_durations=transition_durations,
+        jitters=jitters,
         output_label=output_label,
     )
 

apps/worker/video_processing/transition_engine.py

@@ -105,17 +105,24 @@ class TransitionEngine:
         transitions: list[str],
         *,
         transition_duration: float | None = None,
+        transition_durations: list[float] | None = None,
+        jitters: list[float] | None = None,
         output_label: str = "outv",
     ) -> tuple[str, float]:
         """构建 xfade 转场滤镜链.
 
         对每步转场应用验证和降级，然后调用底层 ffmpeg_utils 构建。
 
+        #1766 增强：支持逐转场独立时长（transition_durations）和位置微调（jitters）。
+
         Args:
             clip_durations: 每个片段的时长
             clip_video_labels: 每个片段的视频流标签
             transitions: 每个片段对应的转场效果
-            transition_duration: 统一转场时长，None 则使用引擎默认值
+            transition_duration: 全局默认转场时长，None 则使用引擎默认值
+            transition_durations: #1766 逐转场时长列表，与 transitions 等长；
+                None 时使用各 resolved config 的 duration
+            jitters: #1766 逐转场位置偏移列表（秒）
             output_label: 最终输出标签
 
         Returns:
@@ -127,6 +134,8 @@ class TransitionEngine:
                 clip_video_labels=clip_video_labels,
                 transitions=transitions,
                 transition_duration=transition_duration or self._default_duration,
+                transition_durations=transition_durations,
+                jitters=jitters,
                 output_label=output_label,
             )
 
@@ -134,17 +143,20 @@ class TransitionEngine:
         resolved = self.resolve_clip_transitions(transitions, clip_durations)
         resolved_effects = [c.effect for c in resolved]
 
-        # 使用统一的时长（取各转场中最大的时长作为基准，底层会做每步钳制）
-        dur = transition_duration or self._default_duration
-        if not dur:
-            dur = max(c.duration for c in resolved) if resolved else DEFAULT_TRANSITION_DURATION
+        # #1766: 逐转场时长（优先使用传入的 transition_durations，否则用 resolved config）
+        if transition_durations is not None:
+            resolved_durations = list(transition_durations)
+        else:
+            resolved_durations = [c.duration for c in resolved]
 
         # 调用底层构建
         return build_xfade_filter_chain(
             clip_durations=clip_durations,
             clip_video_labels=clip_video_labels,
             transitions=resolved_effects,
-            transition_duration=dur,
+            transition_duration=transition_duration or self._default_duration,
+            transition_durations=resolved_durations,
+            jitters=jitters,
             output_label=output_label,
         )
 

apps/worker/video_processing/unified_render_service.py

@@ -1870,16 +1870,27 @@ class UnifiedRenderService:
                     )
                 else:
                     # 有转场效果：用 TransitionEngine 构建 xfade 链
-                    layer_dur = 0.0
-                    for d in layer_transition_durations:
-                        if d > 0:
-                            layer_dur = d
-                            break
+                    # #1766: 提取逐转场时长（每个 clip 的 transition_duration，跳过第一个）
+                    # layer_transition_durations[i] 对应 clip i 的转场，第 0 个忽略
+                    per_transition_durations = [d for idx, d in enumerate(layer_transition_durations) if idx > 0]
+                    # #1766: 提取每个 clip 的 jitter（存在 config 中），跳过第一个
+                    layer_jitters = [
+                        (
+                            all_clips[layer_clip_indices[idx]].config.get("transition_jitter", 0.0)
+                            if isinstance(all_clips[layer_clip_indices[idx]].config, dict)
+                            else 0.0
+                        )
+                        for idx in range(len(layer_clip_indices))
+                    ]
+                    per_transition_jitters = [j for idx, j in enumerate(layer_jitters) if idx > 0]
                     xfade_filter, xfade_estimated_dur = self._transition_engine.build_xfade_chain(
                         clip_durations=layer_durations,
                         clip_video_labels=layer_labels,
                         transitions=layer_transitions,
-                        transition_duration=layer_dur if layer_dur > 0 else None,
+                        transition_durations=(
+                            per_transition_durations if any(d > 0 for d in per_transition_durations) else None
+                        ),
+                        jitters=per_transition_jitters if any(j != 0.0 for j in per_transition_jitters) else None,
                         output_label=out_label,
                     )
                     if xfade_filter:

docs/schema-metadata-snapshot.json

@@ -2126,6 +2126,14 @@
           "type": "JSON",
           "unique": false
         },
+        {
+          "index": false,
+          "name": "is_default",
+          "nullable": false,
+          "primary_key": false,
+          "type": "BOOLEAN",
+          "unique": false
+        },
         {
           "index": false,
           "name": "created_at",
@@ -2142,6 +2150,13 @@
           ],
           "name": "ix_projects_owner_user_id",
           "unique": false
+        },
+        {
+          "columns": [
+            "owner_user_id"
+          ],
+          "name": "uq_projects_owner_default",
+          "unique": true
         }
       ],
       "primary_key": [
@@ -3570,4 +3585,4 @@
       ]
     }
   }
-}
+}

packages/adapters/in_memory/asset_library_repository.py

@@ -35,14 +35,37 @@ class InMemoryAssetLibraryRepository:
             return True
         return False
 
-    def increment_asset_count(self, library_id: str, size_delta: int) -> None:
+    def increment_asset_count(self, library_id: str, count_delta: int = 1, size_delta: int = 0) -> None:
         library = self._libraries.get(library_id)
         if library:
-            library.asset_count += 1
+            library.asset_count += count_delta
             library.total_size += size_delta
 
-    def decrement_asset_count(self, library_id: str, size_delta: int) -> None:
+    def decrement_asset_count(self, library_id: str, count_delta: int = 1, size_delta: int = 0) -> None:
         library = self._libraries.get(library_id)
         if library:
-            library.asset_count = max(0, library.asset_count - 1)
+            library.asset_count = max(0, library.asset_count - count_delta)
             library.total_size = max(0, library.total_size - size_delta)
+
+    def recount_assets(self, library_id: str) -> int:
+        """InMemory 实现无法真正重算（没有 asset 数据源），返回当前计数。"""
+        library = self._libraries.get(library_id)
+        return library.asset_count if library else 0
+
+    def get_or_create_default_library(
+        self,
+        project_id: str,
+        kind: AssetLibraryKind,
+        *,
+        name: str | None = None,
+    ) -> AssetLibrary:
+        """幂等获取/创建默认素材库（Issue #1775，内存实现，模拟唯一约束语义）。"""
+        for lib in self._libraries.values():
+            if lib.project_id == project_id and lib.kind == kind:
+                return lib
+        # 回退到 find_by_project
+        for lib in self.find_by_project(project_id, kind):
+            return lib
+        library_name = name or f"{kind.value}素材库"
+        library = AssetLibrary.create(project_id=project_id, name=library_name, kind=kind)
+        return self.create(library)

packages/adapters/in_memory/project_repository.py

@@ -29,3 +29,30 @@ class InMemoryProjectRepository:
             del self._items[project_id]
             return True
         return False
+
+    def find_default_by_owner(self, owner_user_id: str) -> Project | None:
+        """查找用户的默认项目（Issue #1775 幂等接口，内存实现）。"""
+        for p in self._items.values():
+            if p.owner_user_id == owner_user_id and getattr(p, "is_default", False):
+                return p
+        return None
+
+    def get_or_create_default_project(
+        self,
+        owner_user_id: str,
+        *,
+        name: str = "默认项目",
+        description: str = "小程序自动创建的默认项目",
+    ) -> Project:
+        """幂等获取/创建默认项目（内存实现，模拟 DB 部分唯一索引语义）。"""
+        existing = self.find_default_by_owner(owner_user_id)
+        if existing is not None:
+            return existing
+        project = Project.create(
+            owner_user_id=owner_user_id,
+            name=name,
+            description=description,
+            is_default=True,
+        )
+        self._items[project.id] = project
+        return project

packages/adapters/sqlalchemy_impl/asset_library_repository.py

@@ -77,16 +77,148 @@ class SQLAlchemyAssetLibraryRepository:
             return True
         return False
 
-    async def increment_asset_count(self, library_id: str, size_delta: int) -> None:
-        model = self.session.query(AssetLibraryModel).filter(AssetLibraryModel.id == library_id).first()
-        if model:
-            model.asset_count = (model.asset_count or 0) + 1
-            model.total_size = (model.total_size or 0) + size_delta
-            self.session.commit()
+    def increment_asset_count(self, library_id: str, count_delta: int = 1, size_delta: int = 0) -> None:
+        """原子递增素材计数（Issue #1776）。
 
-    async def decrement_asset_count(self, library_id: str, size_delta: int) -> None:
-        model = self.session.query(AssetLibraryModel).filter(AssetLibraryModel.id == library_id).first()
-        if model:
-            model.asset_count = max(0, (model.asset_count or 0) - 1)
-            model.total_size = max(0, (model.total_size or 0) - size_delta)
+        使用 SQL 级 UPDATE 保证并发安全，不单独 commit（由调用方统一事务提交）。
+        """
+        from sqlalchemy import func
+
+        self.session.query(AssetLibraryModel).filter(AssetLibraryModel.id == library_id).update(
+            {
+                AssetLibraryModel.asset_count: func.coalesce(AssetLibraryModel.asset_count, 0) + count_delta,
+                AssetLibraryModel.total_size: func.coalesce(AssetLibraryModel.total_size, 0) + size_delta,
+            }
+        )
+
+    def decrement_asset_count(self, library_id: str, count_delta: int = 1, size_delta: int = 0) -> None:
+        """原子递减素材计数（Issue #1776），下限为 0 防止负数。
+
+        使用 SQL 级 UPDATE 保证并发安全，不单独 commit（由调用方统一事务提交）。
+        使用 CASE WHEN 兼容 SQLite（测试）和 PostgreSQL（生产）。
+        """
+        from sqlalchemy import case, func
+
+        self.session.query(AssetLibraryModel).filter(AssetLibraryModel.id == library_id).update(
+            {
+                AssetLibraryModel.asset_count: case(
+                    (func.coalesce(AssetLibraryModel.asset_count, 0) - count_delta < 0, 0),
+                    else_=func.coalesce(AssetLibraryModel.asset_count, 0) - count_delta,
+                ),
+                AssetLibraryModel.total_size: case(
+                    (func.coalesce(AssetLibraryModel.total_size, 0) - size_delta < 0, 0),
+                    else_=func.coalesce(AssetLibraryModel.total_size, 0) - size_delta,
+                ),
+            }
+        )
+
+    def recount_assets(self, library_id: str) -> int:
+        """重算素材库计数（Issue #1776）。
+
+        直接查询实际素材数量（排除已删除），更新 asset_count 和 total_size。
+        返回重算后的实际计数。
+        """
+        from sqlalchemy import func
+
+        from packages.adapters.sqlalchemy_impl.models import AssetModel
+
+        # 查询实际计数（排除 deleted）
+        actual_count = (
+            self.session.query(func.count(AssetModel.id))
+            .filter(
+                AssetModel.asset_library_id == library_id,
+                AssetModel.status != "deleted",
+            )
+            .scalar()
+            or 0
+        )
+        # 查询实际总大小
+        actual_size = (
+            self.session.query(func.coalesce(func.sum(AssetModel.file_size), 0))
+            .filter(
+                AssetModel.asset_library_id == library_id,
+                AssetModel.status != "deleted",
+            )
+            .scalar()
+            or 0
+        )
+        # 更新素材库记录
+        self.session.query(AssetLibraryModel).filter(AssetLibraryModel.id == library_id).update(
+            {
+                AssetLibraryModel.asset_count: actual_count,
+                AssetLibraryModel.total_size: actual_size,
+            }
+        )
+        return actual_count
+
+    def get_or_create_default_library(
+        self,
+        project_id: str,
+        kind: AssetLibraryKind,
+        *,
+        name: str | None = None,
+    ) -> AssetLibrary:
+        """幂等获取/创建项目下指定 kind 的默认素材库（Issue #1775）。
+
+        依赖唯一约束 uq_asset_libraries_project_kind(project_id, kind)：
+        并发创建只有一个成功，其余 IntegrityError 后回滚重查，
+        保证同一项目同 kind 永远只有一个素材库。
+        """
+        from sqlalchemy.exc import IntegrityError
+
+        default_names = {
+            AssetLibraryKind.VIDEO: "视频素材库",
+            AssetLibraryKind.VOICE: "配音素材库",
+            AssetLibraryKind.IMAGE: "图片素材库",
+        }
+        library_name = name or default_names.get(kind, f"{kind.value}素材库")
+
+        # 快速路径
+        existing = (
+            self.session.query(AssetLibraryModel)
+            .filter(AssetLibraryModel.project_id == project_id, AssetLibraryModel.kind == kind.value)
+            .first()
+        )
+        if existing:
+            return self._to_entity(existing)
+
+        library = AssetLibrary.create(project_id=project_id, name=library_name, kind=kind)
+        model = AssetLibraryModel(
+            id=library.id,
+            project_id=library.project_id,
+            name=library.name,
+            kind=library.kind.value,
+            asset_count=0,
+            total_size=0,
+            created_at=library.created_at,
+            updated_at=library.updated_at,
+        )
+        try:
+            self.session.add(model)
             self.session.commit()
+            return library
+        except IntegrityError:
+            self.session.rollback()
+            existing = (
+                self.session.query(AssetLibraryModel)
+                .filter(
+                    AssetLibraryModel.project_id == project_id,
+                    AssetLibraryModel.kind == kind.value,
+                )
+                .first()
+            )
+            if existing:
+                return self._to_entity(existing)
+            raise
+
+    def _to_entity(self, model: AssetLibraryModel) -> AssetLibrary:
+        return AssetLibrary(
+            id=model.id,
+            project_id=model.project_id,
+            name=model.name,
+            kind=AssetLibraryKind(model.kind),
+            asset_count=int(model.asset_count or 0),
+            total_size=int(model.total_size or 0),
+            created_at=model.created_at,
+            updated_at=model.updated_at,
+        )

packages/adapters/sqlalchemy_impl/asset_repository.py

@@ -3,7 +3,7 @@ from datetime import datetime, timezone
 
 from sqlalchemy.orm import Session
 
-from packages.adapters.sqlalchemy_impl.models import AssetModel, AssetTagModel
+from packages.adapters.sqlalchemy_impl.models import AssetLibraryModel, AssetModel, AssetTagModel
 from packages.domain import Asset, AssetStatus, ClassificationStatus
 
 
@@ -141,6 +141,16 @@ class SQLAlchemyAssetRepository:
         self.session.add(model)
         self.session.flush()
         self._sync_asset_tags(asset.id, asset.tag_ids)
+        # Issue #1776: 自动维护素材库计数（同事务内原子更新）
+        if asset.library_id and asset.status.value != "deleted":
+            from sqlalchemy import func
+
+            self.session.query(AssetLibraryModel).filter(AssetLibraryModel.id == asset.library_id).update(
+                {
+                    AssetLibraryModel.asset_count: func.coalesce(AssetLibraryModel.asset_count, 0) + 1,
+                    AssetLibraryModel.total_size: func.coalesce(AssetLibraryModel.total_size, 0) + asset.file_size,
+                }
+            )
         self.session.commit()
         return asset
 
@@ -175,7 +185,27 @@ class SQLAlchemyAssetRepository:
     def delete(self, asset_id: str) -> bool:
         model = self.session.query(AssetModel).filter(AssetModel.id == asset_id).first()
         if model:
+            library_id = model.asset_library_id
+            file_size = model.file_size or 0
+            # 只统计非 deleted 状态的素材
+            was_counted = model.status != "deleted"
             self.session.delete(model)
+            # Issue #1776: 自动维护素材库计数
+            if library_id and was_counted:
+                from sqlalchemy import case, func
+
+                self.session.query(AssetLibraryModel).filter(AssetLibraryModel.id == library_id).update(
+                    {
+                        AssetLibraryModel.asset_count: case(
+                            (func.coalesce(AssetLibraryModel.asset_count, 0) - 1 < 0, 0),
+                            else_=func.coalesce(AssetLibraryModel.asset_count, 0) - 1,
+                        ),
+                        AssetLibraryModel.total_size: case(
+                            (func.coalesce(AssetLibraryModel.total_size, 0) - file_size < 0, 0),
+                            else_=func.coalesce(AssetLibraryModel.total_size, 0) - file_size,
+                        ),
+                    }
+                )
             self.session.commit()
             return True
         return False
@@ -187,11 +217,44 @@ class SQLAlchemyAssetRepository:
         from datetime import datetime, timezone
 
         now = datetime.now(timezone.utc)
+        # 先查询待删除素材的库分布（用于更新计数）
+        to_delete = (
+            self.session.query(AssetModel.asset_library_id, AssetModel.file_size)
+            .filter(AssetModel.id.in_(asset_ids), AssetModel.status != "deleted")
+            .all()
+        )
+        if not to_delete:
+            return 0
+        # 按库分组统计
+        library_deltas: dict[str, tuple[int, int]] = {}  # library_id -> (count_delta, size_delta)
+        for lib_id, size in to_delete:
+            if lib_id not in library_deltas:
+                library_deltas[lib_id] = (0, 0)
+            c, s = library_deltas[lib_id]
+            library_deltas[lib_id] = (c + 1, s + (size or 0))
+        # 执行软删除
         count = (
             self.session.query(AssetModel)
             .filter(AssetModel.id.in_(asset_ids), AssetModel.status != "deleted")
             .update({AssetModel.status: "deleted", AssetModel.updated_at: now}, synchronize_session=False)
         )
+        # Issue #1776: 自动维护各素材库计数
+        if library_deltas:
+            from sqlalchemy import case, func
+
+            for lib_id, (count_delta, size_delta) in library_deltas.items():
+                self.session.query(AssetLibraryModel).filter(AssetLibraryModel.id == lib_id).update(
+                    {
+                        AssetLibraryModel.asset_count: case(
+                            (func.coalesce(AssetLibraryModel.asset_count, 0) - count_delta < 0, 0),
+                            else_=func.coalesce(AssetLibraryModel.asset_count, 0) - count_delta,
+                        ),
+                        AssetLibraryModel.total_size: case(
+                            (func.coalesce(AssetLibraryModel.total_size, 0) - size_delta < 0, 0),
+                            else_=func.coalesce(AssetLibraryModel.total_size, 0) - size_delta,
+                        ),
+                    }
+                )
         self.session.commit()
         return count
 

packages/adapters/sqlalchemy_impl/models.py

@@ -1,7 +1,7 @@
 from datetime import datetime, timezone
 from typing import Any
 
-from sqlalchemy import JSON, Boolean, Column, DateTime, Float, Integer, String, Text, UniqueConstraint
+from sqlalchemy import JSON, Boolean, Column, DateTime, Float, Index, Integer, String, Text, UniqueConstraint, text
 from sqlalchemy.orm import declarative_base
 
 Base: Any = declarative_base()
@@ -44,12 +44,18 @@ class UserModel(Base):
 
 class ProjectModel(Base):
     __tablename__ = "projects"
+    __table_args__ = (
+        # Issue #1775: 每个用户至多一个默认项目（部分唯一索引，只约束 is_default=true 的行）。
+        # 注意：不加 UniqueConstraint（那会要求全表唯一），用部分索引表达"每用户一个默认项目"。
+        Index("uq_projects_owner_default", "owner_user_id", unique=True, postgresql_where=text("is_default = true")),
+    )
 
     id = Column(String(36), primary_key=True)
     owner_user_id = Column(String(36), nullable=False, index=True)
     name = Column(String(100), nullable=False)
     description = Column(Text, nullable=False, default="")
     shared_users = Column(JSON, nullable=False, default=list)  # 被共享的用户 ID 列表
+    is_default = Column(Boolean, nullable=False, default=False, server_default="false")
     extra_meta = Column("metadata", JSON, nullable=False, default=dict)
     created_at = Column(DateTime, nullable=False, default=lambda: datetime.now(timezone.utc))
 

packages/adapters/sqlalchemy_impl/project_repository.py

@@ -15,6 +15,7 @@ class SQLAlchemyProjectRepository:
             name=model.name,
             description=model.description,
             shared_users=model.shared_users or [],
+            is_default=bool(getattr(model, "is_default", False)),
             created_at=model.created_at,
         )
 
@@ -33,9 +34,12 @@ class SQLAlchemyProjectRepository:
                 name=project.name,
                 description=project.description,
                 shared_users=project.shared_users,
+                is_default=project.is_default,
                 created_at=project.created_at,
             )
             self.session.add(model)
+        if existing:
+            existing.is_default = project.is_default
         self.session.commit()
         return project
 
@@ -76,3 +80,59 @@ class SQLAlchemyProjectRepository:
         self.session.delete(model)
         self.session.commit()
         return True
+
+    def find_default_by_owner(self, owner_user_id: str) -> Project | None:
+        """查找用户的默认项目（is_default=true）。"""
+        model = (
+            self.session.query(ProjectModel)
+            .filter(ProjectModel.owner_user_id == owner_user_id, ProjectModel.is_default.is_(True))
+            .first()
+        )
+        return self._to_entity(model) if model else None
+
+    def get_or_create_default_project(
+        self,
+        owner_user_id: str,
+        *,
+        name: str = "默认项目",
+        description: str = "小程序自动创建的默认项目",
+    ) -> Project:
+        """幂等获取/创建用户的默认项目（Issue #1775）。
+
+        依赖部分唯一索引 uq_projects_owner_default（每用户至多一条 is_default=true）：
+        并发创建时只有一个 INSERT 成功，其余触发 IntegrityError 后回滚重查，
+        保证同一用户永远只有一个默认项目。
+        """
+        from sqlalchemy.exc import IntegrityError
+
+        # 快速路径：已有默认项目
+        existing = self.find_default_by_owner(owner_user_id)
+        if existing is not None:
+            return existing
+
+        project = Project.create(
+            owner_user_id=owner_user_id,
+            name=name,
+            description=description,
+            is_default=True,
+        )
+        model = ProjectModel(
+            id=project.id,
+            owner_user_id=project.owner_user_id,
+            name=project.name,
+            description=project.description,
+            shared_users=project.shared_users,
+            is_default=True,
+            created_at=project.created_at,
+        )
+        try:
+            self.session.add(model)
+            self.session.commit()
+            return project
+        except IntegrityError:
+            # 并发：另一个请求已插入默认项目，回滚后重查
+            self.session.rollback()
+            existing = self.find_default_by_owner(owner_user_id)
+            if existing is not None:
+                return existing
+            raise

packages/adapters/sqlalchemy_impl/template_clip_config_repository.py

@@ -6,7 +6,10 @@ from typing import List, Optional
 
 from sqlalchemy.orm import Session
 
-from packages.adapters.sqlalchemy_impl.models import TemplateClipConfigModel
+from packages.adapters.sqlalchemy_impl.models import (
+    TemplateClipConfigModel,
+    TemplateModel,
+)
 from packages.domain.template_clip_config import (
     ClipType,
     TemplateClipConfig,
@@ -38,6 +41,24 @@ class SQLAlchemyTemplateClipConfigRepository:
         models = query.offset(skip).limit(limit).all()
         return [self._model_to_entity(m) for m in models]
 
+    def template_owned_by(self, template_id: str, user_id: str) -> bool:
+        """校验旧模板主表 ``templates`` 中模板归属当前用户且未删除（is_active=True）.
+
+        片段配置主表 ``template_clip_configs`` 本身没有 user_id 列，
+        归属关系通过模板主表 ``templates.user_id`` 确定。
+        新表 ``edit_templates`` 为全局模板库（无 user_id 列），不走此校验。
+        """
+        return (
+            self.session.query(TemplateModel.id)
+            .filter(
+                TemplateModel.id == template_id,
+                TemplateModel.user_id == user_id,
+                TemplateModel.is_active.is_(True),
+            )
+            .first()
+            is not None
+        )
+
     def get(self, config_id: str) -> Optional[TemplateClipConfig]:
         """根据 ID 获取配置"""
         model = self.session.query(TemplateClipConfigModel).filter(TemplateClipConfigModel.id == config_id).first()

packages/adapters/sqlalchemy_impl/template_repository.py

@@ -10,6 +10,7 @@ from __future__ import annotations
 import uuid
 from typing import List, Optional
 
+from sqlalchemy import or_
 from sqlalchemy.orm import Session
 
 from packages.adapters.sqlalchemy_impl.models import (
@@ -28,6 +29,20 @@ class SQLAlchemyTemplateRepository:
     def __init__(self, session: Session) -> None:
         self.session = session
 
+    def _filter_with_segment_configs(self, query):
+        """只保留在 template_clip_configs 或 template_segments 中存在片段配置的模板。
+
+        两张表都没有记录的模板无法用于生成（from-assets 会 400），
+        剪辑页选模板时应排除；模板编辑器不传 valid_only，仍可见全部模板。
+        """
+        has_clip_config = self.session.query(TemplateClipConfigModel.id).filter(
+            TemplateClipConfigModel.template_id == TemplateModel.id,
+        )
+        has_segment = self.session.query(TemplateSegmentModel.id).filter(
+            TemplateSegmentModel.template_id == TemplateModel.id,
+        )
+        return query.filter(or_(has_clip_config.exists(), has_segment.exists()))
+
     # ── Template CRUD ──
 
     def list_by_user(
@@ -40,11 +55,14 @@ class SQLAlchemyTemplateRepository:
         tag: Optional[str] = None,
         keyword: Optional[str] = None,
         mode: Optional[str] = None,
+        valid_only: bool = False,
     ) -> List[Template]:
         query = self.session.query(TemplateModel).filter(
             TemplateModel.user_id == user_id,
             TemplateModel.is_active.is_(True),
         )
+        if valid_only:
+            query = self._filter_with_segment_configs(query)
         if category:
             query = query.filter(TemplateModel.category == category)
         if mode:
@@ -102,6 +120,27 @@ class SQLAlchemyTemplateRepository:
         template.segments = self.list_segments(template.id)
         return template
 
+    def get_active(self, template_id: str, user_id: str) -> Optional[Template]:
+        """获取归属当前用户且未删除（is_active=True）的模板，否则返回 None.
+
+        用于编辑器访问门禁：模板不存在、已软删除或不属于当前用户时返回 None，
+        由调用方映射为 404。与 :meth:`get` 的区别是额外过滤 is_active。
+        """
+        model = (
+            self.session.query(TemplateModel)
+            .filter(
+                TemplateModel.id == template_id,
+                TemplateModel.user_id == user_id,
+                TemplateModel.is_active.is_(True),
+            )
+            .first()
+        )
+        if model is None:
+            return None
+        template = self._model_to_entity(model)
+        template.segments = self.list_segments(template.id)
+        return template
+
     def create(self, template: Template) -> Template:
         model = TemplateModel(
             id=template.id,
@@ -173,11 +212,14 @@ class SQLAlchemyTemplateRepository:
         tag: Optional[str] = None,
         keyword: Optional[str] = None,
         mode: Optional[str] = None,
+        valid_only: bool = False,
     ) -> int:
         query = self.session.query(TemplateModel).filter(
             TemplateModel.user_id == user_id,
             TemplateModel.is_active.is_(True),
         )
+        if valid_only:
+            query = self._filter_with_segment_configs(query)
         if category:
             query = query.filter(TemplateModel.category == category)
         if mode:

packages/application/template/commands.py

@@ -62,6 +62,7 @@ class ListTemplatesFilter:
     tag: Optional[str] = None
     keyword: Optional[str] = None
     mode: Optional[str] = None
+    valid_only: bool = False
 
 
 @dataclass

packages/application/template/use_cases.py

@@ -106,6 +106,7 @@ class ListTemplatesUseCase:
             tag=filter.tag,
             keyword=filter.keyword,
             mode=filter.mode,
+            valid_only=filter.valid_only,
         )
 
 
@@ -127,6 +128,7 @@ class CountTemplatesUseCase:
             tag=filter.tag,
             keyword=filter.keyword,
             mode=filter.mode,
+            valid_only=filter.valid_only,
         )
 
 

packages/domain/bgm_pool.py

@@ -0,0 +1,314 @@
+"""BGM 池差异化分配 — 打破变体间音频指纹一致性 (Issue #1767).
+
+三层递进策略：
+1. **BGM 池分配（核心）**：维护风格匹配的 BGM 池，每个变体基于 variant_seed
+   随机分配一首不同 BGM，保证变体间音频指纹不同。
+2. **段落差异化（池不够时的补充）**：同一首 BGM 做差异化裁剪，不同变体使用
+   不同起始点/段落，进一步降低音频相似度。
+3. **音量微调**：不同变体 BGM 音量 ±3dB 微调，混音比例有微小差异。
+
+约束：
+- 不破坏现有单视频 BGM 选择逻辑（单视频不走池分配）
+- BGM 情绪/风格与视频内容匹配（基于源 plan 的 BGM style 做风格筛选）
+- 分配可复现（同 seed 同结果）
+"""
+
+from __future__ import annotations
+
+import logging
+import random
+from dataclasses import dataclass, field
+
+logger = logging.getLogger(__name__)
+
+
+# ── BGM 池条目 ──────────────────────────────────────────────────────────────
+
+
+@dataclass(frozen=True)
+class BGMPoolEntry:
+    """BGM 池条目"""
+
+    id: str
+    preset_id: str  # 关联 PRESET_BGM_LIBRARY 中的 ID（用于渲染侧解析音频路径）
+    mood: str  # 情绪/风格：upbeat / relax / tech / commerce / emotional / cinematic
+    duration: float  # 时长（秒）
+    audio_url: str = ""  # CDN/OSS 直链（优先级高于 preset_id）
+    tags: list[str] = field(default_factory=list)
+
+
+# ── BGM 池（10 首，覆盖 6 种风格） ──────────────────────────────────────────
+
+BGM_POOL: list[BGMPoolEntry] = [
+    # upbeat (轻快)
+    BGMPoolEntry(
+        id="pool_upbeat_001", preset_id="bgm_upbeat_001", mood="upbeat", duration=120.0, tags=["轻快", "阳光", "vlog"]
+    ),
+    BGMPoolEntry(
+        id="pool_upbeat_002", preset_id="bgm_upbeat_002", mood="upbeat", duration=95.0, tags=["轻快", "电子", "运动"]
+    ),
+    BGMPoolEntry(
+        id="pool_upbeat_003", preset_id="bgm_upbeat_003", mood="upbeat", duration=110.0, tags=["轻快", "夏日", "旅行"]
+    ),
+    # relax (治愈)
+    BGMPoolEntry(
+        id="pool_relax_001", preset_id="bgm_relax_001", mood="relax", duration=180.0, tags=["治愈", "钢琴", "冥想"]
+    ),
+    BGMPoolEntry(
+        id="pool_relax_002", preset_id="bgm_relax_002", mood="relax", duration=150.0, tags=["治愈", "自然", "放松"]
+    ),
+    BGMPoolEntry(
+        id="pool_relax_003", preset_id="bgm_relax_003", mood="relax", duration=200.0, tags=["治愈", "古典", "钢琴"]
+    ),
+    # tech (科技)
+    BGMPoolEntry(
+        id="pool_tech_001", preset_id="bgm_tech_001", mood="tech", duration=85.0, tags=["科技", "电子", "数码"]
+    ),
+    BGMPoolEntry(
+        id="pool_tech_002", preset_id="bgm_tech_002", mood="tech", duration=100.0, tags=["科技", "极简", "AI"]
+    ),
+    # commerce (电商)
+    BGMPoolEntry(
+        id="pool_commerce_001",
+        preset_id="bgm_commerce_001",
+        mood="commerce",
+        duration=75.0,
+        tags=["电商", "时尚", "带货"],
+    ),
+    BGMPoolEntry(
+        id="pool_commerce_002",
+        preset_id="bgm_commerce_002",
+        mood="commerce",
+        duration=90.0,
+        tags=["电商", "品牌", "品质"],
+    ),
+]
+
+
+# ── 风格 → 情绪映射 ────────────────────────────────────────────────────────
+# preset_bgm.py 中 style 字段 → bgm_pool.py 中 mood 字段
+
+STYLE_TO_MOOD: dict[str, str] = {
+    "upbeat": "upbeat",
+    "relax": "relax",
+    "tech": "tech",
+    "commerce": "commerce",
+    "emotional": "emotional",
+    "cinematic": "cinematic",
+}
+
+
+# ── 策略一：BGM 池分配 ─────────────────────────────────────────────────────
+
+
+def get_bgm_pool_candidates(source_mood: str | None = None) -> list[BGMPoolEntry]:
+    """获取 BGM 池候选列表。
+
+    如果指定了 source_mood，优先返回同 mood 的条目；
+    如果同 mood 条目不足 2 个，降级返回全池（保证有足够候选）。
+
+    Args:
+        source_mood: 源 BGM 的情绪/风格（来自 preset_bgm.py 的 style 字段）
+
+    Returns:
+        候选 BGM 列表（至少 2 个条目）
+    """
+    if not source_mood:
+        return list(BGM_POOL)
+
+    mood = STYLE_TO_MOOD.get(source_mood, source_mood)
+    matched = [e for e in BGM_POOL if e.mood == mood]
+
+    # 同 mood 至少要有 2 首，否则无法"差异化"，降级全池
+    if len(matched) >= 2:
+        return matched
+    return list(BGM_POOL)
+
+
+def select_bgm_from_pool(
+    variant_seed: int,
+    candidates: list[BGMPoolEntry] | None = None,
+) -> BGMPoolEntry:
+    """基于 variant_seed 从候选池中选一首 BGM（可复现）。
+
+    Args:
+        variant_seed: 变体随机种子
+        candidates: 候选池（None 时使用全池）
+
+    Returns:
+        选中的 BGM 条目
+    """
+    pool = candidates if candidates is not None else list(BGM_POOL)
+    if not pool:
+        pool = list(BGM_POOL)
+    rng = random.Random(variant_seed)
+    return rng.choice(pool)
+
+
+# ── 策略二：段落差异化 ─────────────────────────────────────────────────────
+
+
+def generate_bgm_segment_offset(variant_seed: int, bgm_duration: float) -> float:
+    """为变体生成 BGM 段落起始偏移（策略二）。
+
+    不同变体从同一首 BGM 的不同位置开始播放，进一步降低音频指纹相似度。
+
+    偏移范围 [0, max_offset]，max_offset = min(30s, bgm_duration * 0.3)。
+    量化到 5 秒整数倍，便于复现和调试。
+
+    Args:
+        variant_seed: 变体随机种子
+        bgm_duration: BGM 总时长（秒）
+
+    Returns:
+        起始偏移（秒），0 ~ max_offset 之间，5s 步长
+    """
+    rng = random.Random(variant_seed + 7919)  # 加素数偏移，避免与 BGM 选择 seed 序列重合
+    max_offset = min(30.0, bgm_duration * 0.3)
+    steps = int(max_offset // 5.0)
+    if steps <= 0:
+        return 0.0
+    return float(rng.randint(0, steps) * 5)
+
+
+# ── 策略三：音量微调 ─────────────────────────────────────────────────────
+
+
+def generate_bgm_volume_adjust(variant_seed: int) -> float:
+    """为变体生成 BGM 音量微调值（策略三）。
+
+    ±3dB 微调，让不同变体的 BGM/配音混音比例有微小差异。
+    离散步长：-3, -2, -1, 0, 1, 2, 3 dB。
+
+    Args:
+        variant_seed: 变体随机种子
+
+    Returns:
+        音量调整值（dB），-3.0 ~ 3.0
+    """
+    rng = random.Random(variant_seed + 104729)  # 另一个素数偏移
+    return float(rng.choice([-3, -2, -1, 0, 1, 2, 3]))
+
+
+# ── 批量分配入口 ──────────────────────────────────────────────────────────
+
+
+def allocate_bgm_pool_for_variants(
+    source_bgm_config: dict,
+    variant_seeds: list[int],
+) -> list[dict]:
+    """为批量变体分配不同的 BGM 池配置。
+
+    整合三层策略：池分配 + 段落偏移 + 音量微调。
+    每个变体得到一个 dict，可直接合并到 plan.config["bgm"] 中。
+
+    Args:
+        source_bgm_config: 源 plan 的 BGM 配置（用于风格匹配）
+        variant_seeds: 每个变体的随机种子列表
+
+    Returns:
+        每个变体的 BGM 池配置 dict 列表（与 variant_seeds 等长），
+        每项包含 preset_id / audio_url / audio_offset / volume_adjust_db。
+        如果源 BGM 未启用，返回空列表。
+    """
+    if not source_bgm_config or not source_bgm_config.get("enabled", False):
+        return []
+    if not variant_seeds:
+        return []
+
+    # 从源 BGM 配置中推断风格
+    source_mood = _infer_source_mood(source_bgm_config)
+
+    # 策略一：获取候选池
+    candidates = get_bgm_pool_candidates(source_mood)
+
+    # 为每个变体分配不同的 BGM（尽量不重复）
+    assignments = _assign_unique_bgm(candidates, variant_seeds)
+
+    results = []
+    for i, (entry, seed) in enumerate(zip(assignments, variant_seeds, strict=False)):
+        # 策略二：段落偏移
+        offset = generate_bgm_segment_offset(seed, entry.duration)
+
+        # 策略三：音量微调
+        volume_adj = generate_bgm_volume_adjust(seed)
+
+        result = {
+            "preset_id": entry.preset_id,
+            "audio_url": entry.audio_url,
+            "audio_offset": offset,
+            "volume_adjust_db": volume_adj,
+            "bgm_pool_entry_id": entry.id,
+            "bgm_pool_mood": entry.mood,
+        }
+        results.append(result)
+        logger.info(
+            "变体 %d BGM 池分配: seed=%d bgm=%s mood=%s offset=%.1fs vol_adj=%+.0fdB",
+            i,
+            seed,
+            entry.id,
+            entry.mood,
+            offset,
+            volume_adj,
+        )
+
+    return results
+
+
+def _infer_source_mood(source_bgm_config: dict) -> str | None:
+    """从源 BGM 配置推断风格/情绪。
+
+    优先级：
+    1. preset_id → 查 preset_bgm 库获取 style
+    2. bgm_pool_mood → 上游已设置过（二次分配场景）
+    3. 无法推断 → None（返回全池候选）
+    """
+    preset_id = source_bgm_config.get("preset_id", "")
+    if preset_id:
+        from packages.domain.preset_bgm import get_preset_bgm
+
+        preset = get_preset_bgm(preset_id)
+        if preset:
+            return STYLE_TO_MOOD.get(preset.style, preset.style)
+
+    # 如果之前已经分配过 BGM 池，直接用 mood
+    pool_mood = source_bgm_config.get("bgm_pool_mood", "")
+    if pool_mood:
+        return pool_mood
+
+    return None
+
+
+def _assign_unique_bgm(
+    candidates: list[BGMPoolEntry],
+    variant_seeds: list[int],
+) -> list[BGMPoolEntry]:
+    """尽量让每个变体选到不同的 BGM。
+
+    策略：用 seed 选 BGM，如果与前面变体重复，用递增 seed 重试。
+    如果候选池大小 < 变体数，允许重复但不连续。
+    """
+    if not candidates or not variant_seeds:
+        return []
+
+    assignments: list[BGMPoolEntry] = []
+    used_ids: set[str] = set()
+
+    for i, seed in enumerate(variant_seeds):
+        rng = random.Random(seed)
+        # 先尝试选一个没用过的
+        chosen = None
+        for _attempt in range(len(candidates)):
+            candidate = rng.choice(candidates)
+            if candidate.id not in used_ids:
+                chosen = candidate
+                break
+        if chosen is None:
+            # 候选池已用完，允许重复但取下一个（循环）
+            idx = i % len(candidates)
+            chosen = candidates[idx]
+
+        assignments.append(chosen)
+        used_ids.add(chosen.id)
+
+    return assignments

packages/domain/entities.py

@@ -70,10 +70,11 @@ class Project:
     name: str
     description: str = ""
     shared_users: list[str] = field(default_factory=list)  # 被共享的用户 ID 列表
+    is_default: bool = False  # 是否为用户的默认项目（小程序自动创建），DB 部分唯一索引保证每人至多一个
     created_at: datetime = field(default_factory=lambda: datetime.now(timezone.utc))
 
     @classmethod
-    def create(cls, owner_user_id: str, name: str, description: str = "") -> "Project":
+    def create(cls, owner_user_id: str, name: str, description: str = "", is_default: bool = False) -> "Project":
         clean_name = name.strip()
         if not clean_name:
             raise ValueError("项目名称不能为空")
@@ -83,6 +84,7 @@ class Project:
             name=clean_name,
             description=description.strip(),
             shared_users=[],
+            is_default=is_default,
         )
 
     def is_owner(self, user_id: str) -> bool:

packages/domain/transition_randomizer.py

@@ -0,0 +1,177 @@
+"""转场位置与类型随机化（Issue #1766）.
+
+为不同变体生成不同的转场序列（类型 + 时长 + 位置微调），
+打破"所有变体转场节奏完全一致"的结构相似性，
+降低平台查重识别为结构相似视频的风险。
+
+设计要点：
+1. 转场类型池：5 种效果（dissolve / zoom / slideleft / wipeleft / fade）
+2. 硬切概率：保证 30%-50% 的转场是硬切（保持节奏感）
+3. 转场时长随机：0.3s ~ 0.8s
+4. 转场位置微调：±0.5s 偏移（通过 xfade jitter 实现）
+5. 与 #1764 节奏模板协同：长片段之间的转场倾向于更长时长
+"""
+
+from __future__ import annotations
+
+import logging
+import random
+
+logger = logging.getLogger(__name__)
+
+# ── 转场类型池 ──────────────────────────────────────────────────────────────
+
+#: 非硬切转场类型池（5 种效果，均为 FFmpeg xfade 支持的 transition 名称）
+#: 选择标准：视觉效果差异大、FFmpeg 渲染稳定、肉眼可区分
+TRANSITION_POOL: list[str] = [
+    "dissolve",  # 溶解
+    "zoomin",  # 缩放（放大进入）
+    "slideleft",  # 左滑入
+    "wipeleft",  # 左擦除
+    "fade",  # 淡入淡出
+]
+
+#: 硬切（无转场效果），由 build_xfade_filter_chain 特殊处理（concat filter）
+HARD_CUT = "cut"
+
+# ── 时长约束 ────────────────────────────────────────────────────────────────
+
+#: 随机转场时长下限（秒）
+TRANSITION_DURATION_MIN = 0.3
+
+#: 随机转场时长上限（秒）
+TRANSITION_DURATION_MAX = 0.8
+
+# ── 硬切比例 ────────────────────────────────────────────────────────────────
+
+#: 硬切概率下限（至少 30% 硬切，保持节奏感）
+CUT_RATIO_MIN = 0.3
+
+#: 硬切概率上限（最多 50% 硬切，保证足够视觉变化）
+CUT_RATIO_MAX = 0.5
+
+# ── 位置微调 ────────────────────────────────────────────────────────────────
+
+#: 转场位置最大偏移（秒），实际偏移在 [-MAX, +MAX] 均匀分布
+#: 正 = 转场推迟（多留一点前一片段），负 = 转场提前
+TIMING_JITTER_MAX = 0.5
+
+
+# ── 内部辅助 ────────────────────────────────────────────────────────────────
+
+
+def _cut_probability_for_pair(
+    prev_duration: float,
+    next_duration: float,
+) -> float:
+    """根据相邻片段时长计算硬切概率。
+
+    与 #1764 节奏模板协同：
+    - 两片段都较短（<3s，快节奏）→ 硬切概率更高（节奏更紧凑）
+    - 两片段都较长（>6s，慢节奏）→ 硬切概率稍低（留出过渡空间）
+    - 混合场景 → 基准概率（CUT_RATIO_MIN + CUT_RATIO_MAX）/ 2
+
+    返回概率始终在 [CUT_RATIO_MIN, CUT_RATIO_MAX] 范围内。
+    """
+    base = (CUT_RATIO_MIN + CUT_RATIO_MAX) / 2  # 0.4
+    avg_dur = (prev_duration + next_duration) / 2
+
+    if avg_dur < 3.0:
+        # 快节奏：硬切概率偏高
+        return min(CUT_RATIO_MAX, base + 0.1)
+    elif avg_dur > 6.0:
+        # 慢节奏：硬切概率偏低（更多视觉过渡）
+        return max(CUT_RATIO_MIN, base - 0.1)
+    return base
+
+
+def _apply_jitter(
+    base_duration: float,
+    jitter: float,
+    clip_duration: float,
+) -> float:
+    """给转场时长应用微调偏移，钳制到安全范围。
+
+    Args:
+        base_duration: 基础转场时长
+        jitter: 偏移量（可正可负）
+        clip_duration: 较短的相邻片段时长（转场不能超过此值）
+
+    Returns:
+        钳制后的实际转场时长
+    """
+    effective = base_duration + jitter
+    # 上界：不超过相邻片段时长的 40%（留足内容时间），也不超过 MAX
+    upper = min(TRANSITION_DURATION_MAX, clip_duration * 0.4)
+    lower = TRANSITION_DURATION_MIN if jitter < 0 else max(TRANSITION_DURATION_MIN, base_duration)
+    return max(lower, min(upper, effective))
+
+
+# ── 核心函数 ────────────────────────────────────────────────────────────────
+
+
+def generate_transition_plan(
+    num_transitions: int,
+    *,
+    clip_durations: list[float] | None = None,
+    rng: random.Random | None = None,
+) -> list[dict]:
+    """为变体生成一组随机化的转场计划。
+
+    每个转场点独立随机选择类型和时长，硬切比例保持在 30%-50%。
+
+    Args:
+        num_transitions: 转场点数量（= 主片段数 - 1）
+        clip_durations: 各片段时长（用于协同节奏：长片段间转场更长），
+            长度应 >= num_transitions + 1；不足时用默认值
+        rng: 可选随机数生成器（测试可注入固定种子）
+
+    Returns:
+        转场计划列表，每项：
+        - effect: str — "cut" 或 TRANSITION_POOL 中某一效果
+        - duration: float — 转场时长（cut 为 0.0）
+        - jitter: float — 位置偏移量（秒，-0.5 ~ +0.5）
+    """
+    if num_transitions <= 0:
+        return []
+    if rng is None:
+        rng = random.Random()
+
+    if clip_durations is None:
+        clip_durations = [5.0] * (num_transitions + 1)
+
+    plan: list[dict] = []
+    for i in range(num_transitions):
+        prev_dur = clip_durations[i] if i < len(clip_durations) else 5.0
+        next_dur = clip_durations[i + 1] if (i + 1) < len(clip_durations) else 5.0
+
+        # 计算硬切概率（协同节奏）
+        cut_prob = _cut_probability_for_pair(prev_dur, next_dur)
+
+        # 随机决定是否硬切
+        if rng.random() < cut_prob:
+            effect = HARD_CUT
+            duration = 0.0
+        else:
+            effect = rng.choice(TRANSITION_POOL)
+            base_dur = rng.uniform(TRANSITION_DURATION_MIN, TRANSITION_DURATION_MAX)
+            # 协同节奏：长片段间转场基础时长更长
+            avg_dur = (prev_dur + next_dur) / 2
+            if avg_dur > 6.0:
+                base_dur = min(TRANSITION_DURATION_MAX, base_dur * 1.15)
+            # 应用位置微调
+            jitter = rng.uniform(-TIMING_JITTER_MAX, TIMING_JITTER_MAX)
+            shorter_clip = min(prev_dur, next_dur)
+            duration = _apply_jitter(base_dur, jitter, shorter_clip)
+
+        jitter_val = rng.uniform(-TIMING_JITTER_MAX, TIMING_JITTER_MAX) if effect != HARD_CUT else 0.0
+
+        plan.append(
+            {
+                "effect": effect,
+                "duration": round(duration, 3),
+                "jitter": round(jitter_val, 3),
+            }
+        )
+
+    return plan

packages/domain/variant_plan_selector.py

@@ -29,6 +29,7 @@ import logging
 import random
 
 from packages.domain.plan_generator_utils import _resolve_start_time
+from packages.domain.transition_randomizer import generate_transition_plan
 
 logger = logging.getLogger(__name__)
 
@@ -129,7 +130,10 @@ def reselect_clips_for_variant(
     Raises:
         ValueError: 源片段为空 / 素材池为空 / 素材时长全为 0（无法差异化选片）。
     """
-    rng = rng or random.Random()
+    if rng is None:
+        rng = random.Random()
+    elif isinstance(rng, int):
+        rng = random.Random(rng)
     if not source_clips:
         raise ValueError("源 plan 无片段，无法为变体重新选片")
     if not candidate_asset_ids:
@@ -222,6 +226,9 @@ def reselect_clips_for_variant(
             batch_segments.setdefault(aid, []).append(interval)
         result[idx] = _base_clip_data(src, asset_id=aid, start=start, duration=target_dur)
 
+    # ── 4. #1766 转场随机化：为相邻 main 片段对生成随机转场序列 ────────
+    _apply_transition_randomization(result, rng)
+
     return [c for c in result if c is not None]
 
 
@@ -312,7 +319,10 @@ def generate_visual_perturbation(rng: random.Random | None = None) -> dict:
     - speed_factor: 0.95~1.05 速度微调（±5%，肉眼不太敏感但时间轴不同）
     - brightness_shift: -10~+10 亮度偏移（eq=brightness，画面明暗差异）
     """
-    rng = rng or random.Random()
+    if rng is None:
+        rng = random.Random()
+    elif isinstance(rng, int):
+        rng = random.Random(rng)
     return {
         "hflip": rng.random() < 0.3,
         "zoom_ratio": round(1.0 + rng.uniform(0, 0.08), 4),
@@ -321,7 +331,7 @@ def generate_visual_perturbation(rng: random.Random | None = None) -> dict:
     }
 
 
-def generate_pixel_perturbation(rng: random.Random | None = None) -> dict:
+def generate_pixel_perturbation(rng: random.Random | int | None = None) -> dict:
     """为一个变体生成像素级扰动滤镜参数（Issue #1765）。
 
     在现有视觉扰动（hflip/zoom/brightness）基础上，额外叠加 2-3 种
@@ -336,7 +346,10 @@ def generate_pixel_perturbation(rng: random.Random | None = None) -> dict:
     返回 dict，可直接存入 plan.config["pixel_perturbation"]。
     渲染侧读取后追加到 ffmpeg filter chain。
     """
-    rng = rng or random.Random()
+    if rng is None:
+        rng = random.Random()
+    elif isinstance(rng, int):
+        rng = random.Random(rng)
 
     # 可用滤镜池
     filter_options = ["noise", "unsharp", "curves", "color_balance"]
@@ -369,6 +382,52 @@ def generate_pixel_perturbation(rng: random.Random | None = None) -> dict:
     return result
 
 
+def _apply_transition_randomization(
+    result: list[dict | None],
+    rng: random.Random,
+) -> None:
+    """#1766 对 result 中相邻 main 片段应用转场随机化（就地修改）。
+
+    为每对相邻 main 片段独立选择：
+    - 转场类型（TRANSITION_POOL 中随机，或硬切）
+    - 转场时长（0.3s ~ 0.8s，协同片段时长）
+    - 位置微调 jitter（±0.5s，存入 config["transition_jitter"]）
+
+    硬切比例保证在 30%-50%。intro/outro 等非 main 片段的转场保持源值不变。
+    """
+    # 收集 main 片段的索引（按 order 排序）
+    main_indices = [i for i, c in enumerate(result) if c is not None and c.get("clip_type", "main") == "main"]
+
+    if len(main_indices) < 2:
+        # 不足 2 个 main 片段，无转场点可随机化
+        return
+
+    num_transitions = len(main_indices) - 1
+    # 用 main 片段的 duration 作为协同节奏的输入
+    clip_durations = [result[i]["duration"] for i in main_indices]
+
+    plan = generate_transition_plan(
+        num_transitions,
+        clip_durations=clip_durations,
+        rng=rng,
+    )
+
+    # 将转场计划应用到每对相邻 main 片段
+    # plan[k] 是 main_indices[k] → main_indices[k+1] 之间的转场
+    # 转场信息存储在"目标 clip"（即每对的第二个）的 transition_effect/duration
+    for k, transition_info in enumerate(plan):
+        target_idx = main_indices[k + 1]
+        if result[target_idx] is None:
+            continue
+        clip = result[target_idx]
+        clip["transition_effect"] = transition_info["effect"]
+        clip["transition_duration"] = transition_info["duration"]
+        # jitter 存入 config，供渲染侧 xfade_builder 读取
+        cfg = clip.get("config") or {}
+        cfg["transition_jitter"] = transition_info["jitter"]
+        clip["config"] = cfg
+
+
 def _base_clip_data(src: dict, *, asset_id: str, start: float, duration: float | None = None) -> dict:
     """从源片段构造落库 dict（保留骨架/转场/文案/速度，替换素材与起点）。"""
     return {

packages/domain/video_filter_builder.py

@@ -14,7 +14,7 @@
 from __future__ import annotations
 
 from dataclasses import dataclass
-from typing import TYPE_CHECKING
+from typing import TYPE_CHECKING, Any
 
 from packages.domain.template_clip_config import TransitionEffect
 
@@ -373,3 +373,193 @@ def _append_audio_concat(parts: list[str], clip_chains: list[ClipFilterChain]) -
     # concat 滤镜（使用 audio_label 作为输入）
     audio_inputs = "".join(f"[{c.audio_label}]" for c in audio_chains)
     parts.append(f"{audio_inputs}concat=n={len(audio_chains)}:v=0:a=1[outa]")
+
+
+# ── 标题 drawtext 滤镜构建（#1789）─────────────────────────────────────────────
+
+# drawtext 字体搜索路径：按优先级列出常见安装位置
+# 服务器使用 Noto Sans SC（思源黑体）作为默认字体
+DRAWTEXT_FONT_SEARCH_PATHS: list[str] = [
+    "/usr/share/fonts/opentype/noto/NotoSansCJK-Regular.ttc",
+    "/usr/share/fonts/noto-cjk/NotoSansCJK-Regular.ttc",
+    "/usr/share/fonts/google-noto-cjk/NotoSansCJK-Regular.ttc",
+    "/usr/share/fonts/truetype/noto/NotoSansSC-Regular.ttf",
+    "/usr/share/fonts/noto/NotoSansSC-Regular.ttf",
+    "/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf",
+]
+
+# 前端字体名 → drawtext 字体搜索关键字
+DRAWTEXT_FONT_MAP: dict[str, str] = {
+    "思源黑体": "NotoSansCJK",
+    "思源宋体": "NotoSerifCJK",
+    "苹方": "NotoSansCJK",
+    "PingFang": "NotoSansCJK",
+    "微软雅黑": "NotoSansCJK",
+    "楷体": "NotoSerifCJK",
+    "华康俪金黑": "NotoSansCJK",
+}
+
+
+def _escape_drawtext_text(text: str) -> str:
+    """转义 drawtext 特殊字符。
+
+    FFmpeg drawtext 要求转义：
+    - \\  →  \\\\
+    - '   →  \\\\'
+    - :   →  \\\\:
+    - %   →  %%（drawtext 中 % 是时间码特殊字符）
+    """
+    result = text.replace("\\", "\\\\\\\\")
+    result = result.replace("'", "\\\\'")
+    result = result.replace(":", "\\\\:")
+    result = result.replace("%", "%%")
+    return result
+
+
+def _resolve_font_path(font_name: str) -> str:
+    """解析字体名到服务器实际字体文件路径。
+
+    查找策略：
+    1. 通过 DRAWTEXT_FONT_MAP 映射前端字体名到服务器关键字
+    2. 在 DRAWTEXT_FONT_SEARCH_PATHS 中查找匹配路径
+    3. 未找到则返回空字符串（drawtext 使用内置默认字体）
+    """
+    keyword = DRAWTEXT_FONT_MAP.get(font_name, font_name)
+    import os
+
+    for path in DRAWTEXT_FONT_SEARCH_PATHS:
+        if keyword.lower() in path.lower() and os.path.isfile(path):
+            return path
+    # fallback：遍历搜索任意可用字体
+    for path in DRAWTEXT_FONT_SEARCH_PATHS:
+        if os.path.isfile(path):
+            return path
+    return ""
+
+
+def build_title_drawtext_filter(
+    title_config: dict[str, Any],
+    output_width: int = DEFAULT_OUTPUT_WIDTH,
+    output_height: int = DEFAULT_OUTPUT_HEIGHT,
+) -> str | None:
+    """从 title_config 生成 FFmpeg drawtext 滤镜字符串。
+
+    支持前端 TitleSettings 的全部参数：
+    - text / 标题文字
+    - font / 字体名
+    - font_size / 字号
+    - font_color / 颜色（#RRGGBB）
+    - position / 位置（top / center / bottom / custom）
+    - bold / 粗体
+    - stroke / 描边
+    - shadow / 阴影
+    - pos_x, pos_y / 自由位置坐标
+
+    Args:
+        title_config: 标题配置 dict（来自 plan.config["title"]）
+        output_width: 输出视频宽度
+        output_height: 输出视频高度
+
+    Returns:
+        drawtext 滤镜字符串；标题为空或 disabled 时返回 None
+    """
+    if not title_config or not isinstance(title_config, dict):
+        return None
+
+    # 字段名归一化：兼容 content/text、font_preset/font 两套命名
+    text = (title_config.get("text") or title_config.get("content") or "").strip()
+    if not text:
+        return None
+
+    enabled = title_config.get("enabled", True)
+    if not enabled:
+        return None
+
+    # ── 样式参数 ──
+    font_name = title_config.get("font") or title_config.get("font_preset") or "思源黑体"
+    font_size = int(title_config.get("font_size") or title_config.get("size") or 36)
+    font_color = title_config.get("font_color") or title_config.get("color") or "#ffffff"
+    # 去掉 # 前缀（drawtext 用纯 hex 或颜色名）
+    if font_color.startswith("#"):
+        font_color = font_color[1:]
+
+    position = title_config.get("position", "top")
+    bold = bool(title_config.get("bold", True))
+    stroke = title_config.get("stroke")
+    shadow = title_config.get("shadow")
+
+    # ── 构建 drawtext 参数 ──
+    params: list[str] = []
+
+    # 字体文件
+    font_path = _resolve_font_path(font_name)
+    if font_path:
+        escaped_path = font_path.replace("\\", "\\\\").replace(":", "\\\\:").replace("'", "\\\\'")
+        params.append(f"fontfile='{escaped_path}'")
+
+    # 文字内容
+    params.append(f"text='{_escape_drawtext_text(text)}'")
+
+    # 字号 & 颜色
+    params.append(f"fontsize={font_size}")
+    params.append(f"fontcolor={font_color}")
+
+    # 粗体：bold 在 drawtext 中通过 font 的 Bold 变体实现
+    # 若字体有 Bold 变体可用 fontfont=bold；否则通过 borderw 模拟
+    if bold:
+        # 使用 font 参数尝试加载 Bold 变体（Noto Sans SC 有 Bold 变体文件）
+        params.append("font=bold")
+
+    # 描边（borderw 需要 libfreetype 支持）
+    if stroke:
+        if isinstance(stroke, bool):
+            border_width = 2
+            border_color = "black"
+        elif isinstance(stroke, dict):
+            border_width = int(stroke.get("width", 2)) if stroke.get("enabled", True) else 0
+            border_color = (stroke.get("color") or "#000000").lstrip("#")
+        else:
+            border_width = 0
+            border_color = "black"
+        if border_width > 0:
+            params.append(f"borderw={border_width}")
+            params.append(f"bordercolor={border_color}")
+
+    # 阴影（shadowcolor + shadowx/y）
+    if shadow:
+        if isinstance(shadow, bool):
+            params.append("shadowcolor=black")
+            params.append("shadowx=2")
+            params.append("shadowy=2")
+        elif isinstance(shadow, dict):
+            if shadow.get("enabled", True):
+                params.append(f"shadowcolor={(shadow.get('color') or '#000000').lstrip('#')}")
+                params.append(f"shadowx={int(shadow.get('offset_x', 2))}")
+                params.append(f"shadowy={int(shadow.get('offset_y', 2))}")
+
+    # ── 位置计算 ──
+    # 优先使用自定义坐标 pos_x / pos_y
+    pos_x = title_config.get("pos_x")
+    pos_y = title_config.get("pos_y")
+    if (
+        position == "custom"
+        and isinstance(pos_x, (int, float))
+        and isinstance(pos_y, (int, float))
+        and not isinstance(pos_x, bool)
+        and not isinstance(pos_y, bool)
+    ):
+        params.append(f"x={int(pos_x)}")
+        params.append(f"y={int(pos_y)}")
+    else:
+        # 三档预设位置：top / center / bottom
+        # x 始终水平居中：(w-text_w)/2
+        params.append("x=(w-text_w)/2")
+        if position == "center":
+            params.append("y=(h-text_h)/2")
+        elif position == "bottom":
+            params.append("y=h-text_h-50")
+        else:
+            # top（默认）
+            params.append("y=50")
+
+    return "drawtext=" + ":".join(params)

packages/domain/voice_duration_planner.py

@@ -8,11 +8,13 @@
    禁止慢放、禁止截断配音；
 4. 任何情况下不得因素材时长/数量报错打断用户。
 
-#1764 节奏模板：
-- 预设 6 种权重序列，不同变体用不同节奏模板
+#1764 节奏模板 + #1768 多样化增强：
+- 预设 8 种权重序列，不同变体用不同节奏模板
 - 片段时长 = 配音总时长 × 该片段权重 / 权重总和
 - 平均分配作为权重全 1 的特例保留
 - 每个片段 >= MIN_CLIP_DURATION（2秒）
+- #1768：最大片段时长 <= 素材可用时长 × 90%
+- #1768：成片总时长与配音时长误差 <= TOTAL_DURATION_TOLERANCE（0.5s）
 
 本模块为纯函数：输入片段骨架（每段转场效果/时长）与配音总时长，
 输出每段目标时长（target duration）与成片总时长。不碰 DB、不碰素材。
@@ -42,6 +44,8 @@ RHYTHM_TEMPLATES: list[list[int]] = [
     [3, 1, 1, 1, 3],  # 两端长，中间短
     [1, 1, 3, 2, 1],  # 后段渐长
     [2, 1, 1, 3, 1],  # 前段较长 + 第4段最长
+    [2, 2, 1, 1, 2],  # #1768 前重后轻
+    [3, 2, 1, 2, 1],  # #1768 渐弱节奏
 ]
 
 
@@ -87,13 +91,6 @@ def adapt_template_length(template: list[int], clip_count: int) -> list[int]:
     return result
 
 
-#: 单段最小时长（秒）：低于此值播放器/渲染链路易出问题
-MIN_CLIP_DURATION = 1.0
-
-#: 成片总时长与配音时长的可接受误差（秒）
-TOTAL_DURATION_TOLERANCE = 0.5
-
-
 def transition_overlap_seconds(transition_effect: Optional[str], transition_duration: float) -> float:
     """转场导致的相邻片段重叠时长。
 
@@ -112,6 +109,7 @@ def plan_clip_durations(
     transition_effects: Optional[list[Optional[str]]] = None,
     transition_durations: Optional[list[float]] = None,
     rhythm_template: Optional[list[int]] = None,
+    asset_durations: Optional[list[float]] = None,
 ) -> list[float]:
     """把配音总时长分配到 clip_count 段，返回每段目标时长（秒）。
 
@@ -129,6 +127,9 @@ def plan_clip_durations(
         transition_effects: 每段转场效果（长度 clip_count，index 0 的转场无效）。
         transition_durations: 每段转场时长（长度 clip_count）。
 
+        asset_durations: #1768 每段可用素材时长（秒），用于钳制最大片段时长
+            <= 素材可用时长 × 90%。长度 clip_count；None 或空则不钳制上限。
+
     Returns:
         每段目标时长列表（长度 clip_count）；无配音/非法输入返回 []。
     """
@@ -187,12 +188,58 @@ def plan_clip_durations(
         result[min_idx] = MIN_CLIP_DURATION
         result[max_idx] = round(result[max_idx] - deficit, 3)
 
+    # #1768：最大片段时长钳制（<= 素材可用时长 × 90%）
+    if asset_durations and len(asset_durations) == clip_count:
+        for _ in range(3):  # 迭代收敛
+            clamped = False
+            for i in range(len(result)):
+                try:
+                    max_dur = float(asset_durations[i]) * 0.9
+                except (TypeError, ValueError, IndexError):
+                    continue
+                if result[i] > max_dur and max_dur >= MIN_CLIP_DURATION:
+                    excess = result[i] - max_dur
+                    result[i] = round(max_dur, 3)
+                    # 将多余时长分配给最短的未超限片段
+                    candidates = [
+                        j
+                        for j in range(len(result))
+                        if j != i
+                        and (
+                            not asset_durations
+                            or j >= len(asset_durations)
+                            or result[j] < float(asset_durations[j]) * 0.9
+                        )
+                    ]
+                    if candidates:
+                        shortest = min(candidates, key=lambda j: result[j])
+                        result[shortest] = round(result[shortest] + excess, 3)
+                    clamped = True
+            if not clamped:
+                break
+
     # 末段吸收舍入误差
     total_assigned = sum(result[:-1])
     result[-1] = round(gross - total_assigned, 3)
     if result[-1] < MIN_CLIP_DURATION:
         result[-1] = MIN_CLIP_DURATION
 
+    # #1768：时长总和误差校验（成片净时长 ≈ 配音时长）
+    net_total = total_output_duration(result, transition_effects, transition_durations)
+    deviation = abs(net_total - voice)
+    if deviation > TOTAL_DURATION_TOLERANCE:
+        logger.warning(
+            "#1768 时长总和误差 %.3fs 超过阈值 %.1fs（voice=%.2fs, net=%.2fs），末段补偿修正",
+            deviation,
+            TOTAL_DURATION_TOLERANCE,
+            voice,
+            net_total,
+        )
+        # 修正末段使净时长回归配音时长
+        result[-1] = round(result[-1] + (voice - net_total), 3)
+        if result[-1] < MIN_CLIP_DURATION:
+            result[-1] = MIN_CLIP_DURATION
+
     return result
 
 

packages/domain/xfade_builder.py

@@ -109,6 +109,8 @@ def build_xfade_filter_chain(
     transitions: list[str],
     *,
     transition_duration: float = DEFAULT_TRANSITION_DURATION,
+    transition_durations: list[float] | None = None,
+    jitters: list[float] | None = None,
     output_label: str = "outv",
 ) -> tuple[str, float]:
     """构建 xfade 转场滤镜链.
@@ -116,11 +118,19 @@ def build_xfade_filter_chain(
     对每步 xfade 自动钳制 transition duration，确保
     ``offset + td ≤ first_input_duration``，避免 FFmpeg exit 234。
 
+    #1766 增强：支持逐转场独立时长（transition_durations）和位置微调（jitters）。
+    传入 transition_durations 时，每个转场点使用各自的时长，而非全局统一值。
+    jitters 用于在 offset 上做 ±N 秒微调，实现转场位置随机化。
+
     Args:
         clip_durations: 每个片段的时长（必须与 trim 后的实际时长一致）
         clip_video_labels: 每个片段的视频流标签（如 "v0", "v1"）
         transitions: 每个片段对应的转场效果（第一个片段的转场被忽略）
-        transition_duration: 转场时长（秒）
+        transition_duration: 全局默认转场时长（秒），transition_durations 缺失时 fallback
+        transition_durations: #1766 逐转场时长列表（与 transitions 等长），
+            第 i 项对应 transitions[i] 的时长；None 时使用 transition_duration
+        jitters: #1766 逐转场位置偏移列表（秒），与 transitions 等长，
+            正值推迟转场、负值提前转场；None 时不做微调
         output_label: 最终输出标签
 
     Returns:
@@ -150,15 +160,28 @@ def build_xfade_filter_chain(
         else:
             first_input_dur = cumulative - total_transition
 
-        # 正确的 offset 计算：offset 应相对于累积输出时长
-        # offset = 累积输出中，转场开始的时间点
-        # = first_input_dur - transition_duration
-        # 这样每个转场之间的"纯内容"时长等于原始 clip 时长
-        offset = max(0.0, first_input_dur - transition_duration)
+        # #1766: 逐转场时长（优先）或全局默认
+        step_duration = (
+            transition_durations[i - 1]
+            if transition_durations and (i - 1) < len(transition_durations)
+            else transition_duration
+        )
 
-        # 安全钳制：offset + td 不能超过第一个输入的时长
+        # #1766: 位置微调 jitter
+        jitter = jitters[i - 1] if jitters and (i - 1) < len(jitters) else 0.0
+
+        # offset = 转场开始点（相对于累积输出起点）
+        # 基础 offset = first_input_dur - step_duration
+        # jitter > 0 推迟转场（offset 增大）；jitter < 0 提前转场（offset 减小）
+        offset = max(0.0, first_input_dur - step_duration) + jitter
+
+        # 安全钳制：offset 不能超出可用范围
+        max_offset = max(0.0, first_input_dur - 0.001)
+        offset = max(0.0, min(offset, max_offset))
+
+        # 安全钳制 td：offset + td 不能超过第一个输入的时长
         available = max(0.0, first_input_dur - offset)
-        safe_td = min(transition_duration, available)
+        safe_td = min(step_duration, available)
 
         # 同时不能超过剩余总时长
         remaining = max(0.0, sum(clip_durations) - cumulative)

packages/ports/asset_library_repository.py

@@ -30,9 +30,16 @@ class AssetLibraryRepository(ABC):
         pass
 
     @abstractmethod
-    async def increment_asset_count(self, library_id: str, size_delta: int) -> None:
+    def increment_asset_count(self, library_id: str, count_delta: int = 1, size_delta: int = 0) -> None:
+        """原子递增素材计数（Issue #1776）。"""
         pass
 
     @abstractmethod
-    async def decrement_asset_count(self, library_id: str, size_delta: int) -> None:
+    def decrement_asset_count(self, library_id: str, count_delta: int = 1, size_delta: int = 0) -> None:
+        """原子递减素材计数（Issue #1776）。"""
+        pass
+
+    @abstractmethod
+    def recount_assets(self, library_id: str) -> int:
+        """重算素材库计数（Issue #1776）。"""
         pass

packages/ports/template_repository.py

@@ -18,6 +18,7 @@ class TemplateRepositoryPort(Protocol):
         tag: Optional[str] = None,
         keyword: Optional[str] = None,
         mode: Optional[str] = None,
+        valid_only: bool = False,
     ) -> List[Template]: ...
     def get(self, template_id: str, user_id: str) -> Optional[Template]: ...
     def create(self, template: Template) -> Template: ...
@@ -31,6 +32,7 @@ class TemplateRepositoryPort(Protocol):
         tag: Optional[str] = None,
         keyword: Optional[str] = None,
         mode: Optional[str] = None,
+        valid_only: bool = False,
     ) -> int: ...
     def copy_template(self, template_id: str, user_id: str, new_name: str) -> Template: ...
     def list_segments(self, template_id: str) -> List[TemplateSegment]: ...

scripts/recount_asset_counts.py

@@ -0,0 +1,159 @@
+#!/usr/bin/env python3
+"""素材库计数重算脚本（Issue #1776）。
+
+用法:
+    # Dry-run: 输出差异清单，不执行修改
+    python scripts/recount_asset_counts.py --dry-run
+
+    # 执行修正
+    python scripts/recount_asset_counts.py
+
+    # 只处理指定项目
+    python scripts/recount_asset_counts.py --project-id <project_id>
+
+    # 只处理指定素材库
+    python scripts/recount_asset_counts.py --library-id <library_id>
+"""
+
+import argparse
+import sys
+from pathlib import Path
+
+# 添加项目根目录到 path
+sys.path.insert(0, str(Path(__file__).parent.parent))
+
+from sqlalchemy import create_engine, func
+from sqlalchemy.orm import Session
+
+from packages.adapters.sqlalchemy_impl.models import AssetLibraryModel, AssetModel
+
+
+def get_db_session() -> Session:
+    """创建数据库 session。"""
+    import os
+
+    database_url = os.getenv("DATABASE_URL")
+    if not database_url:
+        print("ERROR: DATABASE_URL environment variable not set")
+        sys.exit(1)
+    engine = create_engine(database_url)
+    return Session(engine)
+
+
+def check_discrepancies(session: Session, project_id: str | None = None, library_id: str | None = None) -> list[dict]:
+    """检查素材库计数差异。
+
+    返回列表，每项包含:
+    - library_id: 素材库 ID
+    - library_name: 素材库名称
+    - recorded_count: 记录的计数
+    - actual_count: 实际计数
+    - delta: 差异 (actual - recorded)
+    """
+    query = session.query(AssetLibraryModel)
+    if project_id:
+        query = query.filter(AssetLibraryModel.project_id == project_id)
+    if library_id:
+        query = query.filter(AssetLibraryModel.id == library_id)
+
+    libraries = query.all()
+    discrepancies = []
+
+    for lib in libraries:
+        # 查询实际计数（排除 deleted）
+        actual_count = (
+            session.query(func.count(AssetModel.id))
+            .filter(
+                AssetModel.asset_library_id == lib.id,
+                AssetModel.status != "deleted",
+            )
+            .scalar()
+            or 0
+        )
+        actual_size = (
+            session.query(func.coalesce(func.sum(AssetModel.file_size), 0))
+            .filter(
+                AssetModel.asset_library_id == lib.id,
+                AssetModel.status != "deleted",
+            )
+            .scalar()
+            or 0
+        )
+        recorded_count = int(lib.asset_count or 0)
+        recorded_size = int(lib.total_size or 0)
+
+        if actual_count != recorded_count or actual_size != recorded_size:
+            discrepancies.append(
+                {
+                    "library_id": lib.id,
+                    "library_name": lib.name,
+                    "project_id": lib.project_id,
+                    "kind": lib.kind,
+                    "recorded_count": recorded_count,
+                    "actual_count": actual_count,
+                    "count_delta": actual_count - recorded_count,
+                    "recorded_size": recorded_size,
+                    "actual_size": actual_size,
+                    "size_delta": actual_size - recorded_size,
+                }
+            )
+
+    return discrepancies
+
+
+def fix_discrepancies(session: Session, discrepancies: list[dict]) -> int:
+    """修正素材库计数。返回修正数量。"""
+    fixed = 0
+    for d in discrepancies:
+        session.query(AssetLibraryModel).filter(AssetLibraryModel.id == d["library_id"]).update(
+            {
+                AssetLibraryModel.asset_count: d["actual_count"],
+                AssetLibraryModel.total_size: d["actual_size"],
+            }
+        )
+        fixed += 1
+    session.commit()
+    return fixed
+
+
+def main():
+    parser = argparse.ArgumentParser(description="素材库计数重算脚本（Issue #1776）")
+    parser.add_argument("--dry-run", action="store_true", help="只输出差异清单，不执行修正")
+    parser.add_argument("--project-id", type=str, help="只处理指定项目")
+    parser.add_argument("--library-id", type=str, help="只处理指定素材库")
+    args = parser.parse_args()
+
+    session = get_db_session()
+
+    try:
+        discrepancies = check_discrepancies(session, args.project_id, args.library_id)
+
+        if not discrepancies:
+            print("✅ 所有素材库计数一致，无需修正")
+            return
+
+        # 输出差异清单
+        print(f"发现 {len(discrepancies)} 个素材库计数不一致:\n")
+        print(f"{'Library ID':<40} {'Name':<20} {'Recorded':<10} {'Actual':<10} {'Delta':<10}")
+        print("-" * 90)
+        for d in discrepancies:
+            print(
+                f"{d['library_id']:<40} {d['library_name'][:20]:<20} {d['recorded_count']:<10} {d['actual_count']:<10} {d['count_delta']:+<10}"
+            )
+
+        total_delta = sum(d["count_delta"] for d in discrepancies)
+        print(f"\n总计差异: {total_delta:+d}")
+
+        if args.dry_run:
+            print("\n[DRY-RUN] 未执行修正。移除 --dry-run 参数以执行修正。")
+        else:
+            print("\n正在执行修正...")
+            fixed = fix_discrepancies(session, discrepancies)
+            print(f"✅ 已修正 {fixed} 个素材库计数")
+
+    finally:
+        session.close()
+
+
+if __name__ == "__main__":
+    main()

tests/unit/domain/test_rhythm_templates_v2.py

@@ -0,0 +1,317 @@
+"""#1768 节奏模板多样化增强 — 单元测试。
+
+覆盖：
+- 8 种预设模板完整性
+- MIN_CLIP_DURATION = 2.0（修复旧 1.0 覆盖 bug）
+- 最大片段时长钳制（<= 素材可用时长 × 90%）
+- 时长总和误差校验（<= 0.5s）
+- asset_durations 参数向后兼容（None/空 = 不钳制）
+"""
+
+from __future__ import annotations
+
+import pytest
+
+from packages.domain.voice_duration_planner import (
+    MIN_CLIP_DURATION,
+    RHYTHM_TEMPLATES,
+    TOTAL_DURATION_TOLERANCE,
+    adapt_template_length,
+    get_rhythm_template,
+    plan_clip_durations,
+    total_output_duration,
+)
+
+# ── 模板池 ──────────────────────────────────────────────────────────────────
+
+
+class TestRhythmTemplatesPool:
+    """#1768 模板池扩展到 8 种。"""
+
+    def test_template_count_is_8(self):
+        assert len(RHYTHM_TEMPLATES) == 8
+
+    def test_all_templates_have_5_segments(self):
+        for tpl in RHYTHM_TEMPLATES:
+            assert len(tpl) == 5
+
+    def test_new_template_22112_exists(self):
+        assert [2, 2, 1, 1, 2] in RHYTHM_TEMPLATES
+
+    def test_new_template_32121_exists(self):
+        assert [3, 2, 1, 2, 1] in RHYTHM_TEMPLATES
+
+    def test_original_6_templates_preserved(self):
+        originals = [
+            [1, 1, 1, 1, 1],
+            [2, 1, 3, 1, 2],
+            [1, 2, 1, 2, 1],
+            [3, 1, 1, 1, 3],
+            [1, 1, 3, 2, 1],
+            [2, 1, 1, 3, 1],
+        ]
+        for orig in originals:
+            assert orig in RHYTHM_TEMPLATES
+
+    def test_all_weights_positive(self):
+        for tpl in RHYTHM_TEMPLATES:
+            assert all(w > 0 for w in tpl)
+
+    def test_weight_sum_variety(self):
+        """不同模板权重和应不完全相同，确保节奏有差异。"""
+        sums = {sum(t) for t in RHYTHM_TEMPLATES}
+        assert len(sums) >= 3  # 至少有 3 种不同的权重和
+
+
+# ── 常量修复 ─────────────────────────────────────────────────────────────────
+
+
+class TestConstantsFixed:
+    """#1768 修复 MIN_CLIP_DURATION 从 1.0 回到 2.0。"""
+
+    def test_min_clip_duration_is_2(self):
+        assert MIN_CLIP_DURATION == 2.0
+
+    def test_total_duration_tolerance_is_05(self):
+        assert TOTAL_DURATION_TOLERANCE == 0.5
+
+
+# ── get_rhythm_template ─────────────────────────────────────────────────────
+
+
+class TestGetRhythmTemplate:
+    def test_none_seed_returns_average(self):
+        assert get_rhythm_template(None) == [1, 1, 1, 1, 1]
+
+    def test_same_seed_returns_same_template(self):
+        for seed in [0, 42, 999, 123456]:
+            t1 = get_rhythm_template(seed)
+            t2 = get_rhythm_template(seed)
+            assert t1 == t2
+
+    def test_different_seeds_can_yield_different_templates(self):
+        """大量 seed 应能命中多个不同模板。"""
+        results = {tuple(get_rhythm_template(s)) for s in range(200)}
+        assert len(results) >= 5  # 200 个 seed 至少命中 5 种模板
+
+
+# ── adapt_template_length ────────────────────────────────────────────────────
+
+
+class TestAdaptTemplateLength:
+    def test_exact_match(self):
+        tpl = [2, 2, 1, 1, 2]
+        assert adapt_template_length(tpl, 5) == tpl
+
+    def test_truncate(self):
+        tpl = [2, 2, 1, 1, 2]
+        assert adapt_template_length(tpl, 3) == [2, 2, 1]
+
+    def test_extend_cycles(self):
+        tpl = [2, 2, 1, 1, 2]
+        result = adapt_template_length(tpl, 8)
+        assert len(result) == 8
+        assert result == [2, 2, 1, 1, 2, 2, 2, 1]
+
+    def test_zero_clips(self):
+        assert adapt_template_length([1, 1, 1], 0) == []
+
+    def test_negative_clips(self):
+        assert adapt_template_length([1, 1, 1], -1) == []
+
+
+# ── plan_clip_durations 基础行为 ────────────────────────────────────────────
+
+
+class TestPlanClipDurationsBasic:
+    def test_invalid_inputs(self):
+        assert plan_clip_durations(0, 30.0) == []
+        assert plan_clip_durations(-1, 30.0) == []
+        assert plan_clip_durations(5, 0.0) == []
+        assert plan_clip_durations(5, -10.0) == []
+        assert plan_clip_durations(5, "abc") == []
+
+    def test_average_distribution_no_transitions(self):
+        result = plan_clip_durations(5, 30.0)
+        assert len(result) == 5
+        assert abs(sum(result) - 30.0) < 0.01
+
+    def test_all_segments_above_min(self):
+        result = plan_clip_durations(5, 30.0, rhythm_template=[3, 1, 1, 1, 3])
+        for dur in result:
+            assert dur >= MIN_CLIP_DURATION
+
+    def test_with_rhythm_template(self):
+        tpl = [2, 2, 1, 1, 2]
+        result = plan_clip_durations(5, 30.0, rhythm_template=tpl)
+        assert len(result) == 5
+        # 权重和 = 8，每段应大致为 7.5, 7.5, 3.75, 3.75, 7.5
+        assert result[0] > result[2]  # 权重 2 > 权重 1
+        assert abs(sum(result) - 30.0) < 0.5
+
+    def test_total_duration_matches_voice(self):
+        """成片净时长 ≈ 配音时长（无转场时完全等于）。"""
+        for voice in [15.0, 30.0, 60.0, 120.0]:
+            result = plan_clip_durations(5, voice)
+            net = total_output_duration(result)
+            assert abs(net - voice) <= TOTAL_DURATION_TOLERANCE
+
+    def test_with_transitions(self):
+        """有转场时成片净时长也应 ≈ 配音时长。"""
+        effects = [None, "xfade", "xfade", "xfade", "xfade"]
+        durations = [0.0, 1.0, 1.0, 1.0, 1.0]
+        result = plan_clip_durations(5, 30.0, transition_effects=effects, transition_durations=durations)
+        net = total_output_duration(result, effects, durations)
+        assert abs(net - 30.0) <= TOTAL_DURATION_TOLERANCE
+
+
+# ── #1768 最小片段时长钳制 ────────────────────────────────────────────────────
+
+
+class TestMinClipDurationClamp:
+    def test_min_duration_2s_enforced(self):
+        """极端权重下，所有片段仍 >= 2.0s。"""
+        tpl = [10, 1, 1, 1, 1]
+        result = plan_clip_durations(5, 20.0, rhythm_template=tpl)
+        for dur in result:
+            assert dur >= 2.0, f"片段时长 {dur} < MIN_CLIP_DURATION(2.0)"
+
+    def test_short_voice_still_meets_minimum(self):
+        """配音极短时保底每段 MIN_CLIP_DURATION。"""
+        result = plan_clip_durations(5, 3.0)
+        for dur in result:
+            assert dur >= MIN_CLIP_DURATION
+
+
+# ── #1768 最大片段时长钳制 ────────────────────────────────────────────────────
+
+
+class TestMaxClipDurationClamp:
+    def test_no_clamp_without_asset_durations(self):
+        """不传 asset_durations 时不做上限钳制（向后兼容）。"""
+        tpl = [5, 1, 1, 1, 1]
+        result = plan_clip_durations(5, 30.0, rhythm_template=tpl)
+        # 第一段权重 5/9 * 30 = 16.67，不应被钳制
+        assert result[0] > 10.0
+
+    def test_no_clamp_with_empty_asset_durations(self):
+        """asset_durations 为空列表时不做上限钳制。"""
+        tpl = [5, 1, 1, 1, 1]
+        result = plan_clip_durations(5, 30.0, rhythm_template=tpl, asset_durations=[])
+        assert result[0] > 10.0
+
+    def test_clamp_respects_90_percent(self):
+        """有素材时长时，片段时长 <= 素材可用时长 × 90%。"""
+        tpl = [5, 1, 1, 1, 1]
+        # 素材只有第一段短（12s），90% = 10.8s
+        asset_durs = [12.0, 60.0, 60.0, 60.0, 60.0]
+        result = plan_clip_durations(5, 30.0, rhythm_template=tpl, asset_durations=asset_durs)
+        max_allowed = 12.0 * 0.9
+        assert result[0] <= max_allowed + 0.01, f"第一段 {result[0]} 超过 90% 上限 {max_allowed}"
+
+    def test_clamp_does_not_violate_min(self):
+        """素材极短时钳制不违反 MIN_CLIP_DURATION。"""
+        # 素材 2.0s，90% = 1.8s < MIN(2.0)，不应钳制到 1.8
+        asset_durs = [2.0, 60.0, 60.0, 60.0, 60.0]
+        result = plan_clip_durations(5, 30.0, asset_durations=asset_durs)
+        for dur in result:
+            assert dur >= MIN_CLIP_DURATION
+
+    def test_clamp_preserves_total(self):
+        """钳制后总时长仍应接近配音时长。"""
+        asset_durs = [10.0, 60.0, 60.0, 60.0, 60.0]
+        voice = 30.0
+        result = plan_clip_durations(5, voice, asset_durations=asset_durs)
+        net = total_output_duration(result)
+        assert abs(net - voice) <= TOTAL_DURATION_TOLERANCE + 0.5  # 允许略多误差
+
+    def test_all_assets_short(self):
+        """所有素材都短时，钳制全部生效但不违反最小值。"""
+        asset_durs = [8.0, 8.0, 8.0, 8.0, 8.0]
+        result = plan_clip_durations(5, 30.0, asset_durations=asset_durs)
+        for dur in result:
+            assert dur >= MIN_CLIP_DURATION
+            max_allowed = 8.0 * 0.9
+            # 如果 max_allowed >= MIN_CLIP_DURATION 才钳制
+            if max_allowed >= MIN_CLIP_DURATION:
+                assert dur <= max_allowed + 0.1
+
+
+# ── #1768 时长总和误差校验 ────────────────────────────────────────────────────
+
+
+class TestTotalDurationTolerance:
+    def test_no_transition_exact_match(self):
+        """无转场时总时长精确等于配音。"""
+        result = plan_clip_durations(5, 25.0)
+        assert abs(sum(result) - 25.0) < 0.01
+
+    def test_with_transition_within_tolerance(self):
+        """有转场时净时长在 0.5s 以内。"""
+        effects = [None, "xfade", "fade", "xfade", "fade"]
+        durations = [0.0, 0.8, 1.2, 0.5, 1.0]
+        result = plan_clip_durations(5, 45.0, transition_effects=effects, transition_durations=durations)
+        net = total_output_duration(result, effects, durations)
+        assert abs(net - 45.0) <= TOTAL_DURATION_TOLERANCE
+
+    @pytest.mark.parametrize("voice", [10.0, 20.0, 30.0, 60.0, 120.0])
+    def test_various_voice_durations(self, voice):
+        result = plan_clip_durations(5, voice)
+        net = total_output_duration(result)
+        assert abs(net - voice) <= TOTAL_DURATION_TOLERANCE
+
+    @pytest.mark.parametrize("tpl", RHYTHM_TEMPLATES)
+    def test_each_template_within_tolerance(self, tpl):
+        """每种模板分配的总时长都应在误差范围内。"""
+        adapted = adapt_template_length(tpl, 5)
+        result = plan_clip_durations(5, 30.0, rhythm_template=adapted)
+        net = total_output_duration(result)
+        assert (
+            abs(net - 30.0) <= TOTAL_DURATION_TOLERANCE
+        ), f"模板 {tpl} 总时长误差 {abs(net - 30.0):.3f}s > {TOTAL_DURATION_TOLERANCE}s"
+
+
+# ── #1768 组合场景 ──────────────────────────────────────────────────────────
+
+
+class TestCombinedScenarios:
+    def test_rhythm_plus_clamp_plus_tolerance(self):
+        """节奏模板 + 素材钳制 + 误差校验 同时生效。"""
+        tpl = [3, 2, 1, 2, 1]
+        effects = [None, "xfade", None, "xfade", None]
+        tdurs = [0.0, 1.0, 0.0, 1.0, 0.0]
+        asset_durs = [15.0, 60.0, 60.0, 60.0, 60.0]
+        voice = 30.0
+
+        adapted = adapt_template_length(tpl, 5)
+        result = plan_clip_durations(
+            5,
+            voice,
+            transition_effects=effects,
+            transition_durations=tdurs,
+            rhythm_template=adapted,
+            asset_durations=asset_durs,
+        )
+
+        # 最小值保证
+        for dur in result:
+            assert dur >= MIN_CLIP_DURATION
+
+        # 最大值钳制（第一段 90% = 13.5）
+        assert result[0] <= 15.0 * 0.9 + 0.1
+
+        # 总时长误差
+        net = total_output_duration(result, effects, tdurs)
+        assert abs(net - voice) <= TOTAL_DURATION_TOLERANCE + 0.5
+
+    def test_many_clips_with_cycling_template(self):
+        """片段数 > 模板长度时循环填充 + 钳制。"""
+        tpl = [2, 2, 1, 1, 2]
+        adapted = adapt_template_length(tpl, 8)
+        assert len(adapted) == 8
+
+        asset_durs = [20.0] * 8
+        result = plan_clip_durations(8, 40.0, rhythm_template=adapted, asset_durations=asset_durs)
+        assert len(result) == 8
+        for dur in result:
+            assert dur >= MIN_CLIP_DURATION

tests/unit/test_asset_count_sync_1776.py

@@ -0,0 +1,305 @@
+"""Issue #1776: asset_libraries.asset_count 同步维护测试。
+
+覆盖场景:
+1. 素材创建 → count +1
+2. 素材硬删除 → count -1
+3. 素材软删除（batch_delete）→ count -N
+4. 幂等上传（prepare 占位 + complete 复用）→ 不重复计数
+5. 重试场景（complete 重试）→ 不重复计数
+6. recount 方法修正计数
+"""
+
+import uuid
+from datetime import datetime, timezone
+
+import pytest
+from sqlalchemy import create_engine
+from sqlalchemy.orm import Session
+
+from packages.adapters.sqlalchemy_impl.asset_library_repository import SQLAlchemyAssetLibraryRepository
+from packages.adapters.sqlalchemy_impl.asset_repository import SQLAlchemyAssetRepository
+from packages.adapters.sqlalchemy_impl.models import AssetLibraryModel, AssetModel, Base
+from packages.domain import Asset, AssetLibrary, AssetLibraryKind, AssetStatus
+
+
+@pytest.fixture
+def db_session():
+    """创建测试用 SQLite 内存数据库。"""
+    engine = create_engine("sqlite:///:memory:")
+    Base.metadata.create_all(engine)
+    session = Session(engine)
+    yield session
+    session.close()
+
+
+@pytest.fixture
+def asset_repo(db_session):
+    return SQLAlchemyAssetRepository(db_session)
+
+
+@pytest.fixture
+def library_repo(db_session):
+    return SQLAlchemyAssetLibraryRepository(db_session)
+
+
+def _make_library(project_id: str, kind: AssetLibraryKind = AssetLibraryKind.VIDEO) -> AssetLibrary:
+    return AssetLibrary.create(project_id=project_id, name=f"测试{kind.value}库", kind=kind)
+
+
+def _make_asset(
+    library_id: str,
+    project_id: str,
+    *,
+    status: AssetStatus = AssetStatus.PROCESSING,
+    file_size: int = 1024,
+    file_hash: str = "",
+    client_upload_id: str = "",
+) -> Asset:
+    return Asset.create(
+        project_id=project_id,
+        library_id=library_id,
+        name=f"test_{uuid.uuid4().hex[:8]}.mp4",
+        storage_key=f"uploads/{uuid.uuid4().hex[:8]}/test.mp4",
+        mime_type="video/mp4",
+        status=status,
+        uploaded_by_user_id="test-user",
+        file_hash=file_hash,
+        client_upload_id=client_upload_id,
+        file_size=file_size,
+    )
+
+
+class TestAssetCountOnCreate:
+    """素材创建时计数递增。"""
+
+    def test_create_asset_increments_count(self, library_repo, asset_repo, db_session):
+        """创建一个素材 → count 从 0 变 1。"""
+        library = library_repo.create(_make_library("project-1"))
+        assert library.asset_count == 0
+
+        asset = _make_asset(library.id, "project-1")
+        asset_repo.create(asset)
+
+        # 重新查询验证计数
+        updated_library = library_repo.get(library.id)
+        assert updated_library.asset_count == 1
+        assert updated_library.total_size == 1024
+
+    def test_create_multiple_assets_increments_count(self, library_repo, asset_repo, db_session):
+        """创建多个素材 → count 累加。"""
+        library = library_repo.create(_make_library("project-1"))
+
+        for _i in range(3):
+            asset = _make_asset(library.id, "project-1", file_size=100 * (_i + 1))
+            asset_repo.create(asset)
+
+        updated_library = library_repo.get(library.id)
+        assert updated_library.asset_count == 3
+        assert updated_library.total_size == 100 + 200 + 300
+
+
+class TestAssetCountOnDelete:
+    """素材删除时计数递减。"""
+
+    def test_hard_delete_decrements_count(self, library_repo, asset_repo, db_session):
+        """硬删除素材 → count -1。"""
+        library = library_repo.create(_make_library("project-1"))
+        asset = asset_repo.create(_make_asset(library.id, "project-1"))
+        assert library_repo.get(library.id).asset_count == 1
+
+        asset_repo.delete(asset.id)
+
+        assert library_repo.get(library.id).asset_count == 0
+
+    def test_hard_delete_already_deleted_no_change(self, library_repo, asset_repo, db_session):
+        """删除已删除的素材 → count 不变。"""
+        library = library_repo.create(_make_library("project-1"))
+        asset = asset_repo.create(_make_asset(library.id, "project-1"))
+        # 先软删除（count 已经 -1）
+        asset_repo.batch_delete([asset.id])
+        assert library_repo.get(library.id).asset_count == 0
+
+        # 再硬删除（不应再 -1）
+        asset_repo.delete(asset.id)
+        assert library_repo.get(library.id).asset_count == 0
+
+    def test_batch_delete_decrements_count(self, library_repo, asset_repo, db_session):
+        """批量软删除 → count -N。"""
+        library = library_repo.create(_make_library("project-1"))
+        asset_ids = []
+        for _i in range(5):
+            asset = asset_repo.create(_make_asset(library.id, "project-1", file_size=200))
+            asset_ids.append(asset.id)
+        assert library_repo.get(library.id).asset_count == 5
+
+        # 删除 3 个
+        deleted_count = asset_repo.batch_delete(asset_ids[:3])
+        assert deleted_count == 3
+        assert library_repo.get(library.id).asset_count == 2
+        assert library_repo.get(library.id).total_size == 200 * 2
+
+    def test_batch_delete_skips_already_deleted(self, library_repo, asset_repo, db_session):
+        """批量删除已删除的素材 → count 不变。"""
+        library = library_repo.create(_make_library("project-1"))
+        asset_ids = []
+        for _i in range(3):
+            asset = asset_repo.create(_make_asset(library.id, "project-1"))
+            asset_ids.append(asset.id)
+        assert library_repo.get(library.id).asset_count == 3
+
+        # 先删除 2 个
+        asset_repo.batch_delete(asset_ids[:2])
+        assert library_repo.get(library.id).asset_count == 1
+
+        # 再删除同样的 2 个（应被跳过）
+        deleted_count = asset_repo.batch_delete(asset_ids[:2])
+        assert deleted_count == 0
+        assert library_repo.get(library.id).asset_count == 1
+
+    def test_count_never_negative(self, library_repo, asset_repo, db_session):
+        """计数下限为 0，不会出现负数。"""
+        library = library_repo.create(_make_library("project-1"))
+        # 手动设置计数为 0
+        library.asset_count = 0
+        library_repo.update(library)
+
+        # 尝试递减（通过直接调用 decrement）
+        library_repo.decrement_asset_count(library.id, count_delta=5)
+        db_session.commit()
+
+        updated = library_repo.get(library.id)
+        assert updated.asset_count == 0
+
+
+class TestIdempotentUpload:
+    """幂等上传场景：不重复计数。"""
+
+    def test_prepare_then_complete_no_double_count(self, library_repo, asset_repo, db_session):
+        """prepare 创建占位 + complete 复用占位 → count 只 +1。"""
+        library = library_repo.create(_make_library("project-1"))
+
+        # prepare 阶段：创建占位
+        placeholder = _make_asset(
+            library.id,
+            "project-1",
+            status=AssetStatus.PROCESSING,
+            client_upload_id="upload-token-123",
+        )
+        asset_repo.create(placeholder)
+        assert library_repo.get(library.id).asset_count == 1
+
+        # complete 阶段：查找已有占位并复用（通过 client_upload_id）
+        existing = asset_repo.find_by_library_and_client_upload_id(
+            library_id=library.id,
+            client_upload_id="upload-token-123",
+        )
+        assert existing is not None
+        # 复用占位，不创建新记录 → count 不变
+        assert library_repo.get(library.id).asset_count == 1
+
+    def test_complete_retry_no_double_count(self, library_repo, asset_repo, db_session):
+        """complete 重试（通过 file_hash 去重）→ count 只 +1。"""
+        library = library_repo.create(_make_library("project-1"))
+
+        # 第一次 complete：创建素材
+        asset1 = _make_asset(
+            library.id,
+            "project-1",
+            file_hash="hash-abc-123",
+        )
+        asset_repo.create(asset1)
+        assert library_repo.get(library.id).asset_count == 1
+
+        # 重试 complete：通过 file_hash 查找已有
+        existing = asset_repo.find_by_library_and_file_hash(
+            library_id=library.id,
+            file_hash="hash-abc-123",
+        )
+        assert existing is not None
+        assert existing.id == asset1.id
+        # 不创建新记录 → count 不变
+        assert library_repo.get(library.id).asset_count == 1
+
+
+class TestRecountAssets:
+    """recount_assets 方法修正计数。"""
+
+    def test_recount_fixes_drift(self, library_repo, asset_repo, db_session):
+        """计数漂移后，recount 能修正。"""
+        library = library_repo.create(_make_library("project-1"))
+        # 创建 3 个素材
+        for _ in range(3):
+            asset_repo.create(_make_asset(library.id, "project-1"))
+        assert library_repo.get(library.id).asset_count == 3
+
+        # 手动破坏计数（模拟历史数据问题）
+        db_session.query(AssetLibraryModel).filter(AssetLibraryModel.id == library.id).update(
+            {AssetLibraryModel.asset_count: 999, AssetLibraryModel.total_size: 999999}
+        )
+        db_session.commit()
+        assert library_repo.get(library.id).asset_count == 999
+
+        # recount 修正
+        actual = library_repo.recount_assets(library.id)
+        db_session.commit()
+
+        assert actual == 3
+        assert library_repo.get(library.id).asset_count == 3
+        assert library_repo.get(library.id).total_size == 1024 * 3
+
+    def test_recount_excludes_deleted(self, library_repo, asset_repo, db_session):
+        """recount 排除已删除素材。"""
+        library = library_repo.create(_make_library("project-1"))
+        assets = []
+        for _ in range(5):
+            asset = asset_repo.create(_make_asset(library.id, "project-1"))
+            assets.append(asset)
+        assert library_repo.get(library.id).asset_count == 5
+
+        # 软删除 2 个
+        asset_repo.batch_delete([assets[0].id, assets[1].id])
+        assert library_repo.get(library.id).asset_count == 3
+
+        # 破坏计数
+        db_session.query(AssetLibraryModel).filter(AssetLibraryModel.id == library.id).update(
+            {AssetLibraryModel.asset_count: 100}
+        )
+        db_session.commit()
+
+        # recount 应排除 deleted
+        actual = library_repo.recount_assets(library.id)
+        db_session.commit()
+
+        assert actual == 3
+        assert library_repo.get(library.id).asset_count == 3
+
+
+class TestConcurrentSafety:
+    """并发安全测试（SQLite 模拟有限并发）。"""
+
+    def test_increment_is_atomic(self, library_repo, db_session):
+        """increment_asset_count 使用 SQL 级 UPDATE，并发安全。"""
+        library = library_repo.create(_make_library("project-1"))
+
+        # 多次递增
+        for _ in range(10):
+            library_repo.increment_asset_count(library.id, count_delta=1, size_delta=100)
+        db_session.commit()
+
+        updated = library_repo.get(library.id)
+        assert updated.asset_count == 10
+        assert updated.total_size == 1000
+
+    def test_decrement_with_floor_zero(self, library_repo, db_session):
+        """decrement_asset_count 下限为 0。"""
+        library = library_repo.create(_make_library("project-1"))
+        library.asset_count = 3
+        library_repo.update(library)
+
+        # 尝试递减 10 次
+        for _ in range(10):
+            library_repo.decrement_asset_count(library.id, count_delta=1)
+        db_session.commit()
+
+        updated = library_repo.get(library.id)
+        assert updated.asset_count == 0

tests/unit/test_bgm_pool.py

@@ -0,0 +1,281 @@
+"""BGM 池差异化分配单元测试 (Issue #1767).
+
+覆盖：
+- BGM 池定义（10 首，覆盖 4 种风格）
+- 风格匹配：get_bgm_pool_candidates 按 mood 筛选
+- 变体分配：select_bgm_from_pool 基于 seed 可复现选择
+- 批量分配：allocate_bgm_pool_for_variants 确保不同变体不同 BGM
+- 段落差异化：generate_bgm_segment_offset 生成不同偏移
+- 音量微调：generate_bgm_volume_adjust ±3dB
+- 边界条件：空配置/未启用 BGM/未知 mood
+"""
+
+from __future__ import annotations
+
+import pytest
+
+from packages.domain.bgm_pool import (
+    BGM_POOL,
+    STYLE_TO_MOOD,
+    BGMPoolEntry,
+    allocate_bgm_pool_for_variants,
+    generate_bgm_segment_offset,
+    generate_bgm_volume_adjust,
+    get_bgm_pool_candidates,
+    select_bgm_from_pool,
+)
+
+
+class TestBGMPoolDefinition:
+    """BGM 池定义测试。"""
+
+    def test_pool_has_at_least_10_entries(self):
+        """BGM 池至少 10 首（满足 5-10 首需求）。"""
+        assert len(BGM_POOL) >= 10
+
+    def test_all_entries_have_required_fields(self):
+        """每条 BGM 池条目都有 id/preset_id/mood/duration。"""
+        for entry in BGM_POOL:
+            assert entry.id, "Missing id"
+            assert entry.preset_id, f"Missing preset_id in {entry.id}"
+            assert entry.mood, f"Missing mood in {entry.id}"
+            assert entry.duration > 0, f"Duration must be > 0 in {entry.id}"
+
+    def test_pool_covers_multiple_moods(self):
+        """池覆盖至少 3 种不同 mood。"""
+        moods = {e.mood for e in BGM_POOL}
+        assert len(moods) >= 3, f"Expected >= 3 moods, got {moods}"
+
+    def test_each_mood_has_at_least_2_entries(self):
+        """每种 mood 至少有 2 首（保证差异化有意义）。"""
+        mood_counts: dict[str, int] = {}
+        for entry in BGM_POOL:
+            mood_counts[entry.mood] = mood_counts.get(entry.mood, 0) + 1
+        for mood, count in mood_counts.items():
+            assert count >= 2, f"Mood '{mood}' only has {count} entries (need >= 2)"
+
+
+class TestGetBGMPoolCandidates:
+    """风格匹配测试。"""
+
+    def test_no_mood_returns_full_pool(self):
+        """不指定 mood 时返回全池。"""
+        candidates = get_bgm_pool_candidates(None)
+        assert len(candidates) == len(BGM_POOL)
+
+    def test_matching_mood_filters(self):
+        """指定已知 mood 时返回同 mood 条目。"""
+        candidates = get_bgm_pool_candidates("upbeat")
+        assert all(c.mood == "upbeat" for c in candidates)
+        assert len(candidates) >= 2
+
+    def test_unknown_mood_returns_full_pool(self):
+        """未知 mood 降级返回全池。"""
+        candidates = get_bgm_pool_candidates("nonexistent_style")
+        # 如果没有匹配到同 mood 的（>=2），降级全池
+        assert len(candidates) >= 2
+
+    def test_style_to_mood_mapping(self):
+        """STYLE_TO_MOOD 覆盖所有 preset_bgm 的 style。"""
+        assert "upbeat" in STYLE_TO_MOOD
+        assert "relax" in STYLE_TO_MOOD
+        assert "tech" in STYLE_TO_MOOD
+        assert "commerce" in STYLE_TO_MOOD
+
+
+class TestSelectBGMPool:
+    """变体 BGM 选择测试。"""
+
+    def test_same_seed_same_result(self):
+        """相同 seed 返回相同 BGM（可复现）。"""
+        result1 = select_bgm_from_pool(42)
+        result2 = select_bgm_from_pool(42)
+        assert result1.id == result2.id
+
+    def test_different_seeds_may_differ(self):
+        """不同 seed 可能返回不同 BGM。"""
+        results = set()
+        for seed in range(50):
+            entry = select_bgm_from_pool(seed)
+            results.add(entry.id)
+        assert len(results) >= 3, "Expected >= 3 different BGMs from 50 seeds"
+
+    def test_respects_candidates_filter(self):
+        """传入候选池时只从中选择。"""
+        candidates = [e for e in BGM_POOL if e.mood == "tech"]
+        for seed in range(20):
+            entry = select_bgm_from_pool(seed, candidates)
+            assert entry.mood == "tech"
+
+
+class TestBGMSegmentOffset:
+    """段落差异化测试（策略二）。"""
+
+    def test_offset_non_negative(self):
+        """偏移 >= 0。"""
+        for seed in range(50):
+            offset = generate_bgm_segment_offset(seed, 120.0)
+            assert offset >= 0.0
+
+    def test_offset_bounded(self):
+        """偏移 <= min(30s, duration * 0.3)。"""
+        for seed in range(50):
+            duration = 100.0
+            max_expected = min(30.0, duration * 0.3)
+            offset = generate_bgm_segment_offset(seed, duration)
+            assert offset <= max_expected + 0.1  # 容差
+
+    def test_different_seeds_different_offsets(self):
+        """不同 seed 产生不同偏移（统计验证）。"""
+        offsets = set()
+        for seed in range(30):
+            offsets.add(generate_bgm_segment_offset(seed, 120.0))
+        assert len(offsets) >= 3, "Expected >= 3 distinct offsets"
+
+    def test_offset_is_quantized(self):
+        """偏移是 5s 的整数倍。"""
+        for seed in range(20):
+            offset = generate_bgm_segment_offset(seed, 120.0)
+            assert offset % 5.0 == 0.0
+
+    def test_short_bgm_zero_offset(self):
+        """极短 BGM 偏移为 0。"""
+        offset = generate_bgm_segment_offset(42, 5.0)
+        # max_offset = min(30, 5*0.3) = 1.5, steps = int(1.5//5) = 0 → return 0
+        assert offset == 0.0
+
+    def test_offset_different_from_bgm_selection(self):
+        """偏移的 seed 序列与 BGM 选择的 seed 序列不同（加素数偏移）。"""
+        # 同一 seed，偏移和 BGM 选择应该独立
+        bgm = select_bgm_from_pool(42)
+        offset = generate_bgm_segment_offset(42, 120.0)
+        # 只是验证能正常运行，不直接断言独立性（统计测试需要大样本）
+        assert isinstance(offset, float)
+
+
+class TestBGMVolumeAdjust:
+    """音量微调测试（策略三）。"""
+
+    def test_volume_adjust_in_range(self):
+        """音量调整在 -3 ~ +3 dB 范围内。"""
+        for seed in range(50):
+            adj = generate_bgm_volume_adjust(seed)
+            assert -3.0 <= adj <= 3.0
+            assert adj == int(adj)  # 整数 dB 步进
+
+    def test_different_seeds_different_volumes(self):
+        """不同 seed 产生不同音量调整值。"""
+        values = set()
+        for seed in range(50):
+            values.add(generate_bgm_volume_adjust(seed))
+        assert len(values) >= 3, "Expected >= 3 distinct volume values"
+
+    def test_includes_zero(self):
+        """音量调整值集合包含 0（不变）。"""
+        values = {generate_bgm_volume_adjust(seed) for seed in range(100)}
+        assert 0.0 in values
+
+
+class TestAllocateBGMPoolForVariants:
+    """批量分配入口测试。"""
+
+    def test_disabled_bgm_returns_empty(self):
+        """源 BGM 未启用时返回空列表。"""
+        result = allocate_bgm_pool_for_variants({"enabled": False}, [1, 2, 3])
+        assert result == []
+
+    def test_empty_config_returns_empty(self):
+        """空配置返回空列表。"""
+        result = allocate_bgm_pool_for_variants({}, [1, 2, 3])
+        assert result == []
+
+    def test_empty_seeds_returns_empty(self):
+        """无变体时返回空列表。"""
+        result = allocate_bgm_pool_for_variants({"enabled": True, "preset_id": "bgm_upbeat_001"}, [])
+        assert result == []
+
+    def test_returns_correct_count(self):
+        """返回与 variant_seeds 等长的列表。"""
+        config = {"enabled": True, "preset_id": "bgm_upbeat_001"}
+        seeds = [100, 200, 300, 400]
+        result = allocate_bgm_pool_for_variants(config, seeds)
+        assert len(result) == 4
+
+    def test_each_entry_has_required_keys(self):
+        """每项都包含必要字段。"""
+        config = {"enabled": True, "preset_id": "bgm_upbeat_001"}
+        seeds = [100, 200, 300]
+        result = allocate_bgm_pool_for_variants(config, seeds)
+        for entry in result:
+            assert "preset_id" in entry
+            assert "audio_offset" in entry
+            assert "volume_adjust_db" in entry
+            assert "bgm_pool_entry_id" in entry
+            assert "bgm_pool_mood" in entry
+
+    def test_batch_3_variants_at_least_2_different_bgm(self):
+        """批量 3 个变体，至少 2 个不同 BGM。"""
+        config = {"enabled": True, "preset_id": "bgm_upbeat_001"}
+        seeds = [100, 200, 300]
+        result = allocate_bgm_pool_for_variants(config, seeds)
+        bgm_ids = {r["bgm_pool_entry_id"] for r in result}
+        assert len(bgm_ids) >= 2, f"Expected >= 2 different BGMs, got {bgm_ids}"
+
+    def test_style_matching_with_preset_id(self):
+        """源 BGM 有 preset_id 时按风格筛选。"""
+        config = {"enabled": True, "preset_id": "bgm_tech_001"}  # tech 风格
+        seeds = [100, 200, 300]
+        result = allocate_bgm_pool_for_variants(config, seeds)
+        # tech mood 至少有 2 首，所以应该筛选到 tech
+        for entry in result:
+            assert entry["bgm_pool_mood"] == "tech"
+
+    def test_reproducible_with_same_seeds(self):
+        """相同 seeds 产生相同分配（可复现）。"""
+        config = {"enabled": True, "preset_id": "bgm_upbeat_001"}
+        seeds = [42, 100, 200]
+        result1 = allocate_bgm_pool_for_variants(config, seeds)
+        result2 = allocate_bgm_pool_for_variants(config, seeds)
+        for r1, r2 in zip(result1, result2, strict=False):
+            assert r1["bgm_pool_entry_id"] == r2["bgm_pool_entry_id"]
+            assert r1["audio_offset"] == r2["audio_offset"]
+            assert r1["volume_adjust_db"] == r2["volume_adjust_db"]
+
+
+class TestIssue1767Acceptance:
+    """Issue #1767 验收测试。"""
+
+    def test_batch_3_videos_bgm_different_or_segment_different(self):
+        """批量生成 3 个视频，BGM 不同或起始段落不同。"""
+        config = {"enabled": True, "preset_id": "bgm_upbeat_001"}
+        seeds = [100, 200, 300]
+        result = allocate_bgm_pool_for_variants(config, seeds)
+
+        # 检查：BGM 不同 或 段落偏移不同
+        unique_combos = set()
+        for r in result:
+            combo = (r["bgm_pool_entry_id"], r["audio_offset"])
+            unique_combos.add(combo)
+
+        assert len(unique_combos) >= 2, f"Expected >= 2 unique (bgm, offset) combos, got {unique_combos}"
+
+    def test_bgm_volume_micro_adjust_doesnt_affect_voice_clarity(self):
+        """BGM 音量微调在 ±3dB 内，不影响配音清晰度。"""
+        config = {"enabled": True, "preset_id": "bgm_upbeat_001"}
+        seeds = [100, 200, 300]
+        result = allocate_bgm_pool_for_variants(config, seeds)
+
+        for r in result:
+            # ±3dB 是安全的微调范围，不会让 BGM 盖过配音
+            assert abs(r["volume_adjust_db"]) <= 3.0
+
+    def test_single_video_mode_unaffected(self):
+        """单视频模式不受影响（不走池分配）。"""
+        # 单视频不调用 allocate_bgm_pool_for_variants
+        # 只要不主动调用，就不会改变行为
+        # 这个测试验证函数签名和行为不会意外影响单视频
+        config = {"enabled": True, "preset_id": "bgm_upbeat_001"}
+        result = allocate_bgm_pool_for_variants(config, [42])  # 单变体
+        assert len(result) == 1
+        # 单项分配仍然有完整配置（不影响功能，只是差异化）
+        assert "preset_id" in result[0]

tests/unit/test_default_project_idempotent_1775.py

@@ -0,0 +1,177 @@
+"""默认项目/默认素材库幂等化测试（Issue #1775）。
+
+覆盖：
+- get_or_create_default_project：同用户幂等、不同用户独立、重复调用返回同一个
+- get_or_create_default_library：同项目同 kind 幂等、IntegrityError 后重查
+- Project.is_default 字段传递
+- ensure-default-context 组合逻辑（用内存仓储）
+"""
+
+from __future__ import annotations
+
+from unittest.mock import MagicMock
+
+import pytest
+
+from packages.adapters.in_memory.asset_library_repository import InMemoryAssetLibraryRepository
+from packages.adapters.in_memory.project_repository import InMemoryProjectRepository
+from packages.adapters.sqlalchemy_impl.asset_library_repository import (
+    SQLAlchemyAssetLibraryRepository,
+)
+from packages.adapters.sqlalchemy_impl.project_repository import SQLAlchemyProjectRepository
+from packages.domain import AssetLibrary, AssetLibraryKind, Project
+
+
+class TestDefaultProjectIdempotent:
+    """默认项目幂等。"""
+
+    def test_first_call_creates_default(self):
+        repo = InMemoryProjectRepository()
+        project = repo.get_or_create_default_project("user-1")
+        assert project.owner_user_id == "user-1"
+        assert project.is_default is True
+        assert project.name == "默认项目"
+
+    def test_second_call_returns_same_project(self):
+        repo = InMemoryProjectRepository()
+        p1 = repo.get_or_create_default_project("user-1")
+        p2 = repo.get_or_create_default_project("user-1")
+        assert p1.id == p2.id
+
+    def test_different_users_independent(self):
+        repo = InMemoryProjectRepository()
+        p1 = repo.get_or_create_default_project("user-1")
+        p2 = repo.get_or_create_default_project("user-2")
+        assert p1.id != p2.id
+        assert p1.owner_user_id == "user-1"
+        assert p2.owner_user_id == "user-2"
+
+    def test_find_default_by_owner(self):
+        repo = InMemoryProjectRepository()
+        created = repo.get_or_create_default_project("user-1")
+        found = repo.find_default_by_owner("user-1")
+        assert found is not None
+        assert found.id == created.id
+
+    def test_find_default_returns_none_when_no_default(self):
+        repo = InMemoryProjectRepository()
+        # 手动建一个非默认项目
+        normal = Project.create(owner_user_id="user-1", name="普通项目")
+        normal.is_default = False
+        repo.save(normal)
+        assert repo.find_default_by_owner("user-1") is None
+
+    def test_repeated_calls_after_creation_return_same(self):
+        """先创建默认项目后，后续多次调用均返回已有项目（测试 find 路径）。
+        注：真正的并发保护依赖 PostgreSQL partial unique index，
+        InMemory 仓储不做并发测试（无 DB 约束），并发场景由
+        TestSqlRepoIntegrityErrorRecovery 通过 SQLAlchemy + SQLite 验证。
+        """
+        repo = InMemoryProjectRepository()
+        first = repo.get_or_create_default_project("user-repeat")
+        for _ in range(9):
+            again = repo.get_or_create_default_project("user-repeat")
+            assert again.id == first.id
+        defaults = [p for p in repo.find_by_owner_user_id("user-repeat") if p.is_default]
+        assert len(defaults) == 1
+
+
+class TestDefaultLibraryIdempotent:
+    """默认素材库幂等。"""
+
+    def test_first_call_creates(self):
+        repo = InMemoryAssetLibraryRepository()
+        lib = repo.get_or_create_default_library("proj-1", AssetLibraryKind.VIDEO)
+        assert lib.project_id == "proj-1"
+        assert lib.kind == AssetLibraryKind.VIDEO
+
+    def test_second_call_returns_same(self):
+        repo = InMemoryAssetLibraryRepository()
+        l1 = repo.get_or_create_default_library("proj-1", AssetLibraryKind.VIDEO)
+        l2 = repo.get_or_create_default_library("proj-1", AssetLibraryKind.VIDEO)
+        assert l1.id == l2.id
+
+    def test_different_kinds_independent(self):
+        repo = InMemoryAssetLibraryRepository()
+        v = repo.get_or_create_default_library("proj-1", AssetLibraryKind.VIDEO)
+        a = repo.get_or_create_default_library("proj-1", AssetLibraryKind.VOICE)
+        i = repo.get_or_create_default_library("proj-1", AssetLibraryKind.IMAGE)
+        assert len({v.id, a.id, i.id}) == 3
+
+    def test_different_projects_independent(self):
+        repo = InMemoryAssetLibraryRepository()
+        l1 = repo.get_or_create_default_library("proj-1", AssetLibraryKind.VIDEO)
+        l2 = repo.get_or_create_default_library("proj-2", AssetLibraryKind.VIDEO)
+        assert l1.id != l2.id
+
+    def test_repeated_calls_after_creation_return_same(self):
+        """先创建后多次调用均返回同一素材库（测试 find 路径）。"""
+        repo = InMemoryAssetLibraryRepository()
+        first = repo.get_or_create_default_library("proj-repeat", AssetLibraryKind.VOICE)
+        for _ in range(9):
+            again = repo.get_or_create_default_library("proj-repeat", AssetLibraryKind.VOICE)
+            assert again.id == first.id
+
+
+class TestProjectIsDefaultField:
+    """Project.is_default 字段语义。"""
+
+    def test_create_non_default_by_default(self):
+        p = Project.create(owner_user_id="u", name="普通项目")
+        assert p.is_default is False
+
+    def test_create_default(self):
+        p = Project.create(owner_user_id="u", name="默认项目", is_default=True)
+        assert p.is_default is True
+
+
+class TestSqlRepoIntegrityErrorRecovery:
+    """SQLAlchemy 仓储：唯一约束冲突时回滚重查，返回已有记录（不报 500）。"""
+
+    def test_project_integrity_error_returns_existing(self):
+        from sqlalchemy.exc import IntegrityError
+
+        session = MagicMock()
+        # commit 第一次抛 IntegrityError（并发冲突），回滚后查询返回已有项目
+        existing_model = MagicMock()
+        existing_model.id = "existing-id"
+        existing_model.owner_user_id = "user-1"
+        existing_model.name = "默认项目"
+        existing_model.description = ""
+        existing_model.shared_users = []
+        existing_model.is_default = True
+        existing_model.created_at = None
+
+        session.commit.side_effect = [IntegrityError("stmt", {}, Exception("dup")), None]
+        # 第一次 query（快速路径 find_default）返回 None；rollback 后第二次返回 existing
+        session.query.return_value.filter.return_value.first.side_effect = [None, existing_model]
+
+        repo = SQLAlchemyProjectRepository(session)
+        result = repo.get_or_create_default_project("user-1")
+
+        assert result.id == "existing-id"
+        session.rollback.assert_called_once()
+
+    def test_library_integrity_error_returns_existing(self):
+        from sqlalchemy.exc import IntegrityError
+
+        session = MagicMock()
+        existing_model = MagicMock()
+        existing_model.id = "lib-existing"
+        existing_model.project_id = "proj-1"
+        existing_model.name = "视频素材库"
+        existing_model.kind = "video"
+        existing_model.asset_count = 0
+        existing_model.total_size = 0
+        existing_model.created_at = None
+        existing_model.updated_at = None
+
+        session.commit.side_effect = [IntegrityError("stmt", {}, Exception("dup")), None]
+        session.query.return_value.filter.return_value.first.side_effect = [None, existing_model]
+
+        repo = SQLAlchemyAssetLibraryRepository(session)
+        result = repo.get_or_create_default_library("proj-1", AssetLibraryKind.VIDEO)
+
+        assert result.id == "lib-existing"
+        assert result.kind == AssetLibraryKind.VIDEO
+        session.rollback.assert_called_once()

tests/unit/test_editor_clips_random_start.py

@@ -239,8 +239,8 @@ class TestEditorClipsBySegments:
         assert not hasattr(mock_plan_svc, "create_clip") or not mock_plan_svc.create_clip.called
 
     @patch("app.api.routes.templates_editor.clips.get_storage_service")
-    def test_no_segments_raises_400(self, mock_storage):
-        """模板没有 segment 配置时返回 400。"""
+    def test_no_segments_raises_422(self, mock_storage):
+        """模板存在但未配置片段时返回 422（配置错误，与 404 区分）。"""
         from app.api.routes.templates_editor.clips import (
             create_clips_from_assets_editor,
         )
@@ -264,11 +264,44 @@ class TestEditorClipsBySegments:
                     current_user=_make_auth_user(),
                 )
 
-        assert exc_info.value.status_code == 400
-        assert "片段配置" in exc_info.value.detail
+        assert exc_info.value.status_code == 422
+        assert "片段" in exc_info.value.detail
         # 不应调用替换方法
         mock_plan_svc.replace_all_clips_transactional.assert_not_called()
 
+    @patch("app.api.routes.templates_editor.clips.get_storage_service")
+    def test_template_not_found_raises_404(self, mock_storage):
+        """模板不存在/已删除/无权限（服务层抛 TemplateNotFoundError）时返回 404。"""
+        from app.api.routes.templates_editor.clips import (
+            create_clips_from_assets_editor,
+        )
+        from app.api.routes.templates_editor.schemas import ClipsFromAssetsRequest
+        from app.services.edit_template_service import TemplateNotFoundError
+
+        mock_plan_svc = _make_plan_svc()
+        mock_asset_repo = MagicMock()
+
+        body = ClipsFromAssetsRequest(asset_ids=["a1"])
+
+        with patch(
+            "app.api.routes.templates_editor.clips._get_template_segments",
+            side_effect=TemplateNotFoundError("tpl-missing"),
+        ):
+            with pytest.raises(HTTPException) as exc_info:
+                create_clips_from_assets_editor(
+                    template_id="tpl-missing",
+                    body=body,
+                    background_tasks=MagicMock(),
+                    plan_id=TEST_PLAN_ID,
+                    services=(MagicMock(), mock_plan_svc),
+                    asset_repo=mock_asset_repo,
+                    db=MagicMock(),
+                    current_user=_make_auth_user(),
+                )
+
+        assert exc_info.value.status_code == 404
+        mock_plan_svc.replace_all_clips_transactional.assert_not_called()
+
 
 class TestEditorClipsDurationAndStartTime:
     """测试素材时长获取、clip duration 缩短、start_time 传入。"""

tests/unit/test_get_template_segments_fallback.py

@@ -1,13 +1,17 @@
-"""_get_template_segments 回退路径测试.
+"""模板片段配置读取路径测试（#1774）.
 
-验证三级回退链：
-1. 新模板系统（tpl_svc.list_clip_configs）正常 → 直接返回
-2. 新模板系统主表不存在（ValueError）→ 直接查 template_clip_configs 表兜底
-3. 直接查表也失败 → 回退旧模板系统（template_segments）
-4. 全部失败 → 返回空列表
+收敛后模板读取走单一数据源，不再有"新表抛异常→降级查旧表→再降级查 segments"
+的异常控制流：
 
-覆盖 P0 修复：自建模板在 edit_templates 主表不存在但在 template_clip_configs 有记录时，
-from-assets 流程不再 400。
+- ``EditTemplateService.list_clip_configs_for_editor`` 显式判定模板归属/存在性：
+  1. 用户自建模板在旧表 ``templates``（归属 user_id，is_active=True）→ 直接读
+     ``template_clip_configs``；
+  2. 全局模板在新表 ``edit_templates``（无 user_id）→ 直接读 ``template_clip_configs``；
+  3. 两表都没有 → 抛 ``TemplateNotFoundError``（路由层映射 404）。
+- ``_get_template_segments`` 仅做配置→(order, min, max) 的映射与排序，
+  模板存在但无配置返回空列表（路由层映射 422）。
+
+使用真实 SQLite 内存库 + 真实仓储，验证端到端读路径不抛 ``ValueError: 模板不存在``。
 """
 
 from __future__ import annotations
@@ -15,21 +19,198 @@ from __future__ import annotations
 import os
 import sys
 from pathlib import Path
-from unittest.mock import MagicMock, PropertyMock
+from unittest.mock import MagicMock
 
 os.environ.setdefault("JWT_SECRET_KEY", "unit-test-secret-key-for-testing")
 os.environ.setdefault("DATABASE_URL", "sqlite:///test.db")
 
 sys.path.insert(0, str(Path(__file__).resolve().parents[2] / "apps" / "api"))
 
-from app.api.routes.templates_editor.clips import _get_template_segments
+import pytest  # noqa: E402
+from sqlalchemy import create_engine  # noqa: E402
+from sqlalchemy.orm import sessionmaker  # noqa: E402
+
+from packages.adapters.sqlalchemy_impl.models import (  # noqa: E402
+    Base,
+    EditTemplateModel,
+    TemplateClipConfigModel,
+    TemplateModel,
+)
 
-TEST_TEMPLATE_ID = "tmpl-orphan-001"
 DEFAULT_DUR = 5.0  # _DEFAULT_EDITOR_CLIP_DURATION
+USER_ID = "user-001"
+OTHER_USER_ID = "user-002"
+
+
+# ---------------------------------------------------------------------------
+# 真实内存 DB fixture
+# ---------------------------------------------------------------------------
+
+
+def _make_session():
+    engine = create_engine("sqlite:///:memory:")
+    Base.metadata.create_all(engine)
+    return sessionmaker(bind=engine)()
+
+
+def _seed_legacy_template(session, template_id: str, user_id: str, *, active: bool = True, clip_count: int = 3):
+    """创建旧表 templates 模板（+ template_clip_configs 片段配置）。"""
+    session.add(
+        TemplateModel(
+            id=template_id,
+            user_id=user_id,
+            name=f"模板-{template_id}",
+            mode="one_take",
+            is_active=active,
+        )
+    )
+    for order in range(clip_count):
+        session.add(
+            TemplateClipConfigModel(
+                id=f"cc-{template_id}-{order}",
+                template_id=template_id,
+                clip_type="main",
+                order=order,
+                min_duration=5.0,
+                max_duration=8.0,
+            )
+        )
+    session.commit()
+
+
+def _seed_global_template(session, template_id: str, *, status: str = "active", clip_count: int = 2):
+    """创建新表 edit_templates 全局模板（+ template_clip_configs 片段配置）。"""
+    session.add(
+        EditTemplateModel(
+            id=template_id,
+            name=f"全局模板-{template_id}",
+            template_type="default",
+            editing_mode="one_take",
+            status=status,
+        )
+    )
+    for order in range(clip_count):
+        session.add(
+            TemplateClipConfigModel(
+                id=f"gcc-{template_id}-{order}",
+                template_id=template_id,
+                clip_type="main",
+                order=order,
+                min_duration=3.0,
+                max_duration=6.0,
+            )
+        )
+    session.commit()
+
+
+# ---------------------------------------------------------------------------
+# Service 读路径测试
+# ---------------------------------------------------------------------------
+
+
+class TestListClipConfigsForEditor:
+    """list_clip_configs_for_editor 单一数据源 + 归属/存在性判定。"""
+
+    def test_legacy_user_template_returns_configs(self):
+        """用户自建模板（templates 表 + 3 条 clip_configs）→ 正常返回，不抛异常。"""
+        from app.services.edit_template_service import EditTemplateService
+
+        session = _make_session()
+        _seed_legacy_template(session, "tmpl-legacy", USER_ID, clip_count=3)
+
+        svc = EditTemplateService(session)
+        configs = svc.list_clip_configs_for_editor("tmpl-legacy", USER_ID)
+
+        assert len(configs) == 3
+        assert [c.order for c in configs] == [0, 1, 2]
+        assert all(c.min_duration == 5.0 for c in configs)
+
+    def test_missing_template_raises_not_found(self):
+        """模板不存在（两表都没有）→ TemplateNotFoundError。"""
+        from app.services.edit_template_service import EditTemplateService, TemplateNotFoundError
+
+        session = _make_session()
+        svc = EditTemplateService(session)
+
+        with pytest.raises(TemplateNotFoundError):
+            svc.list_clip_configs_for_editor("tmpl-not-exist", USER_ID)
+
+    def test_other_users_template_raises_not_found(self):
+        """他人模板（user_id 不匹配）→ TemplateNotFoundError（归属校验）。"""
+        from app.services.edit_template_service import EditTemplateService, TemplateNotFoundError
+
+        session = _make_session()
+        _seed_legacy_template(session, "tmpl-owner", OTHER_USER_ID, clip_count=3)
+
+        svc = EditTemplateService(session)
+        with pytest.raises(TemplateNotFoundError):
+            svc.list_clip_configs_for_editor("tmpl-owner", USER_ID)
+
+    def test_deleted_legacy_template_raises_not_found(self):
+        """已软删除（is_active=False）的旧表模板 → TemplateNotFoundError。"""
+        from app.services.edit_template_service import EditTemplateService, TemplateNotFoundError
+
+        session = _make_session()
+        _seed_legacy_template(session, "tmpl-deleted", USER_ID, active=False, clip_count=3)
+
+        svc = EditTemplateService(session)
+        with pytest.raises(TemplateNotFoundError):
+            svc.list_clip_configs_for_editor("tmpl-deleted", USER_ID)
+
+    def test_legacy_template_without_configs_returns_empty(self):
+        """模板存在且归属正确但无片段配置 → 返回空列表（不抛异常，路由层映射 422）。"""
+        from app.services.edit_template_service import EditTemplateService
+
+        session = _make_session()
+        _seed_legacy_template(session, "tmpl-noconfig", USER_ID, clip_count=0)
+
+        svc = EditTemplateService(session)
+        configs = svc.list_clip_configs_for_editor("tmpl-noconfig", USER_ID)
+        assert configs == []
+
+    def test_global_template_returns_configs(self):
+        """新表 edit_templates 全局模板（无 user_id）→ 任意用户可读，正常返回。"""
+        from app.services.edit_template_service import EditTemplateService
+
+        session = _make_session()
+        _seed_global_template(session, "tmpl-global", clip_count=2)
+
+        svc = EditTemplateService(session)
+        configs = svc.list_clip_configs_for_editor("tmpl-global", USER_ID)
+
+        assert len(configs) == 2
+        assert [c.order for c in configs] == [0, 1]
+
+    def test_normal_legacy_request_does_not_raise_valueerror(self):
+        """正常旧表模板请求绝不在读路径抛 ValueError: 模板不存在（回归保护）。"""
+        import logging
+
+        from app.services.edit_template_service import EditTemplateService
+
+        session = _make_session()
+        _seed_legacy_template(session, "tmpl-ok", USER_ID, clip_count=3)
+        svc = EditTemplateService(session)
+
+        with pytest.MonkeyPatch.context() as mp:
+            # 若读路径意外抛 ValueError 并被记录为异常堆栈，测试能感知
+            errors: list[str] = []
+            mp.setattr(
+                logging.getLogger("app.services.edit_template_service"),
+                "exception",
+                lambda *a, **k: errors.append(str(a)),
+            )
+            configs = svc.list_clip_configs_for_editor("tmpl-ok", USER_ID)
+
+        assert len(configs) == 3
+        assert errors == []
+
+
+# ---------------------------------------------------------------------------
+# _get_template_segments 映射测试
+# ---------------------------------------------------------------------------
 
 
 def _make_clip_config(order: int, min_dur: float = 3.0, max_dur: float = 8.0):
-    """构造 mock TemplateClipConfig 领域实体."""
     cc = MagicMock()
     cc.order = order
     cc.min_duration = min_dur
@@ -37,207 +218,53 @@ def _make_clip_config(order: int, min_dur: float = 3.0, max_dur: float = 8.0):
     return cc
 
 
-def _make_old_segment(segment_order: int, dur_min: float = 4.0, dur_max: float = 7.0):
-    """构造 mock 旧 TemplateSegment."""
-    s = MagicMock()
-    s.segment_order = segment_order
-    s.duration_min = dur_min
-    s.duration_max = dur_max
-    return s
+class TestGetTemplateSegments:
+    """_get_template_segments 仅做映射/排序，异常与空配置语义明确。"""
 
+    def test_maps_and_sorts_configs(self):
+        from app.api.routes.templates_editor.clips import _get_template_segments
 
-# ---------------------------------------------------------------------------
-# 测试
-# ---------------------------------------------------------------------------
-
-
-class TestGetTemplateSegmentsFallback:
-    """_get_template_segments 三级回退链."""
-
-    def test_new_system_works(self):
-        """路径1：新模板系统正常返回 → 直接使用."""
-        configs = [_make_clip_config(0, 2.0, 6.0), _make_clip_config(1, 3.0, 9.0)]
         tpl_svc = MagicMock()
-        tpl_svc.list_clip_configs.return_value = configs
-        db = MagicMock()
-
-        result = _get_template_segments(TEST_TEMPLATE_ID, tpl_svc, db)
-
-        assert len(result) == 2
-        assert result[0] == (0, 2.0, 6.0)
-        assert result[1] == (1, 3.0, 9.0)
-        tpl_svc.list_clip_configs.assert_called_once_with(TEST_TEMPLATE_ID)
-
-    def test_main_table_missing_direct_query_succeeds(self):
-        """路径2（P0修复）：主表不存在 ValueError → 直接查表成功.
-
-        模拟自建模板在 edit_templates 主表已删除/不存在，
-        但 template_clip_configs 表有记录。
-        """
-        tpl_svc = MagicMock()
-        tpl_svc.list_clip_configs.side_effect = ValueError(f"模板不存在: {TEST_TEMPLATE_ID}")
-        db = MagicMock()
-
-        # Mock SQLAlchemyTemplateClipConfigRepository
-        direct_configs = [
-            _make_clip_config(0, 2.0, 5.0),
-            _make_clip_config(1, 3.0, 7.0),
-            _make_clip_config(2, 4.0, 8.0),
-        ]
-        with (
-            __import__("unittest.mock", fromlist=["patch"]).patch(
-                "app.api.routes.templates_editor.clips.SQLAlchemyTemplateClipConfigRepository"
-            ) as mock_repo_cls,
-        ):
-            mock_repo = MagicMock()
-            mock_repo.list_by_template.return_value = direct_configs
-            mock_repo_cls.return_value = mock_repo
-
-            result = _get_template_segments(TEST_TEMPLATE_ID, tpl_svc, db)
-
-        assert len(result) == 3
-        assert result[0] == (0, 2.0, 5.0)
-        assert result[1] == (1, 3.0, 7.0)
-        assert result[2] == (2, 4.0, 8.0)
-        mock_repo.list_by_template.assert_called_once_with(TEST_TEMPLATE_ID)
-
-    def test_main_table_missing_direct_query_empty_falls_to_old(self):
-        """路径2→3：主表不存在 + 直接查表为空 → 回退旧系统."""
-        tpl_svc = MagicMock()
-        tpl_svc.list_clip_configs.side_effect = ValueError("模板不存在")
-        db = MagicMock()
-
-        old_segments = [_make_old_segment(0, 3.0, 6.0)]
-
-        with __import__("unittest.mock", fromlist=["patch"]).patch(
-            "app.api.routes.templates_editor.clips.SQLAlchemyTemplateClipConfigRepository"
-        ) as mock_repo_cls:
-            mock_repo = MagicMock()
-            mock_repo.list_by_template.return_value = []  # 新表也没记录
-            mock_repo_cls.return_value = mock_repo
-
-            with __import__("unittest.mock", fromlist=["patch"]).patch(
-                "app.api.routes.templates_editor.clips.SQLAlchemyTemplateRepository"
-            ) as mock_old_cls:
-                mock_old = MagicMock()
-                mock_old.list_segments.return_value = old_segments
-                mock_old_cls.return_value = mock_old
-
-                result = _get_template_segments(TEST_TEMPLATE_ID, tpl_svc, db)
-
-        assert len(result) == 1
-        assert result[0] == (0, 3.0, 6.0)
-
-    def test_all_fail_returns_empty(self):
-        """路径4：三级全部失败 → 返回空列表."""
-        tpl_svc = MagicMock()
-        tpl_svc.list_clip_configs.side_effect = ValueError("模板不存在")
-        db = MagicMock()
-
-        with __import__("unittest.mock", fromlist=["patch"]).patch(
-            "app.api.routes.templates_editor.clips.SQLAlchemyTemplateClipConfigRepository"
-        ) as mock_repo_cls:
-            mock_repo = MagicMock()
-            mock_repo.list_by_template.side_effect = Exception("DB error")
-            mock_repo_cls.return_value = mock_repo
-
-            with __import__("unittest.mock", fromlist=["patch"]).patch(
-                "app.api.routes.templates_editor.clips.SQLAlchemyTemplateRepository"
-            ) as mock_old_cls:
-                mock_old = MagicMock()
-                mock_old.list_segments.return_value = []  # 旧表也空
-                mock_old_cls.return_value = mock_old
-
-                result = _get_template_segments(TEST_TEMPLATE_ID, tpl_svc, db)
-
-        assert result == []
-
-    def test_direct_query_sorts_by_order(self):
-        """直接查表返回的结果按 order 排序."""
-        tpl_svc = MagicMock()
-        tpl_svc.list_clip_configs.side_effect = ValueError("模板不存在")
-        db = MagicMock()
-
-        # 故意乱序
-        configs = [
+        tpl_svc.list_clip_configs_for_editor.return_value = [
             _make_clip_config(2, 5.0, 10.0),
             _make_clip_config(0, 2.0, 4.0),
             _make_clip_config(1, 3.0, 6.0),
         ]
 
-        with __import__("unittest.mock", fromlist=["patch"]).patch(
-            "app.api.routes.templates_editor.clips.SQLAlchemyTemplateClipConfigRepository"
-        ) as mock_repo_cls:
-            mock_repo = MagicMock()
-            mock_repo.list_by_template.return_value = configs
-            mock_repo_cls.return_value = mock_repo
-
-            result = _get_template_segments(TEST_TEMPLATE_ID, tpl_svc, db)
+        result = _get_template_segments("tmpl-1", USER_ID, tpl_svc)
 
         assert [r[0] for r in result] == [0, 1, 2]
         assert result[0] == (0, 2.0, 4.0)
-        assert result[1] == (1, 3.0, 6.0)
         assert result[2] == (2, 5.0, 10.0)
+        tpl_svc.list_clip_configs_for_editor.assert_called_once_with("tmpl-1", USER_ID)
+
+    def test_empty_configs_returns_empty(self):
+        from app.api.routes.templates_editor.clips import _get_template_segments
 
-    def test_direct_query_handles_none_durations(self):
-        """直接查表时 min/max_duration 为 None → 使用默认值."""
         tpl_svc = MagicMock()
-        tpl_svc.list_clip_configs.side_effect = ValueError("模板不存在")
-        db = MagicMock()
+        tpl_svc.list_clip_configs_for_editor.return_value = []
 
+        assert _get_template_segments("tmpl-1", USER_ID, tpl_svc) == []
+
+    def test_missing_template_propagates_not_found(self):
+        from app.api.routes.templates_editor.clips import _get_template_segments
+        from app.services.edit_template_service import TemplateNotFoundError
+
+        tpl_svc = MagicMock()
+        tpl_svc.list_clip_configs_for_editor.side_effect = TemplateNotFoundError("tmpl-x")
+
+        with pytest.raises(TemplateNotFoundError):
+            _get_template_segments("tmpl-x", USER_ID, tpl_svc)
+
+    def test_none_durations_use_default(self):
+        from app.api.routes.templates_editor.clips import _get_template_segments
+
+        tpl_svc = MagicMock()
         cc = MagicMock()
         cc.order = 0
         cc.min_duration = None
         cc.max_duration = None
+        tpl_svc.list_clip_configs_for_editor.return_value = [cc]
 
-        with __import__("unittest.mock", fromlist=["patch"]).patch(
-            "app.api.routes.templates_editor.clips.SQLAlchemyTemplateClipConfigRepository"
-        ) as mock_repo_cls:
-            mock_repo = MagicMock()
-            mock_repo.list_by_template.return_value = [cc]
-            mock_repo_cls.return_value = mock_repo
-
-            result = _get_template_segments(TEST_TEMPLATE_ID, tpl_svc, db)
-
-        assert len(result) == 1
-        # None → default (5.0), max(None or None) → default (5.0)
-        assert result[0] == (0, DEFAULT_DUR, DEFAULT_DUR)
-
-    def test_new_system_returns_empty_tries_direct(self):
-        """新模板系统返回空列表（非异常）→ 继续尝试直接查表."""
-        tpl_svc = MagicMock()
-        tpl_svc.list_clip_configs.return_value = []  # 空列表，非异常
-        db = MagicMock()
-
-        direct_configs = [_make_clip_config(0, 3.0, 6.0)]
-
-        with __import__("unittest.mock", fromlist=["patch"]).patch(
-            "app.api.routes.templates_editor.clips.SQLAlchemyTemplateClipConfigRepository"
-        ) as mock_repo_cls:
-            mock_repo = MagicMock()
-            mock_repo.list_by_template.return_value = direct_configs
-            mock_repo_cls.return_value = mock_repo
-
-            result = _get_template_segments(TEST_TEMPLATE_ID, tpl_svc, db)
-
-        # 新系统返回空 → 不走 except → 但也没 return → 继续往下走
-        # 直接查表有数据 → 返回
-        assert len(result) == 1
-        assert result[0] == (0, 3.0, 6.0)
-
-    def test_existing_template_unaffected(self):
-        """正常模板（主表存在）行为不变."""
-        configs = [_make_clip_config(0, 2.0, 5.0)]
-        tpl_svc = MagicMock()
-        tpl_svc.list_clip_configs.return_value = configs
-        db = MagicMock()
-
-        with __import__("unittest.mock", fromlist=["patch"]).patch(
-            "app.api.routes.templates_editor.clips.SQLAlchemyTemplateClipConfigRepository"
-        ) as mock_repo_cls:
-            result = _get_template_segments(TEST_TEMPLATE_ID, tpl_svc, db)
-            # 直接查表不应被调用（新系统已返回）
-            mock_repo_cls.assert_not_called()
-
-        assert len(result) == 1
-        assert result[0] == (0, 2.0, 5.0)
+        result = _get_template_segments("tmpl-1", USER_ID, tpl_svc)
+        assert result == [(0, DEFAULT_DUR, DEFAULT_DUR)]

tests/unit/test_in_memory_asset_library_repository.py

@@ -128,12 +128,12 @@ class TestAssetLibraryRepoCounting:
         assert lib.asset_count == 0
         assert lib.total_size == 0
 
-        repo.increment_asset_count(lib.id, 1024)
+        repo.increment_asset_count(lib.id, size_delta=1024)
         fetched = repo.get(lib.id)
         assert fetched.asset_count == 1
         assert fetched.total_size == 1024
 
-        repo.increment_asset_count(lib.id, 2048)
+        repo.increment_asset_count(lib.id, size_delta=2048)
         fetched = repo.get(lib.id)
         assert fetched.asset_count == 2
         assert fetched.total_size == 3072
@@ -144,7 +144,7 @@ class TestAssetLibraryRepoCounting:
         lib.total_size = 3000
         repo.create(lib)
 
-        repo.decrement_asset_count(lib.id, 1000)
+        repo.decrement_asset_count(lib.id, size_delta=1000)
         fetched = repo.get(lib.id)
         assert fetched.asset_count == 2
         assert fetched.total_size == 2000
@@ -157,14 +157,14 @@ class TestAssetLibraryRepoCounting:
         repo.create(lib)
 
         # 减 2 次，应该被钳制到 0
-        repo.decrement_asset_count(lib.id, 200)
+        repo.decrement_asset_count(lib.id, size_delta=200)
         fetched = repo.get(lib.id)
         assert fetched.asset_count == 0
         assert fetched.total_size == 0
 
     def test_increment_nonexistent_library_no_error(self, repo):
         """对不存在的素材库操作，不抛异常也无效果."""
-        repo.increment_asset_count("nonexistent", 100)
+        repo.increment_asset_count("nonexistent", size_delta=100)
         # 不报错
         assert repo.get("nonexistent") is None
 

tests/unit/test_inmemory_small_repos.py

@@ -97,40 +97,40 @@ class TestInMemoryAssetLibraryRepository:
 
     def test_increment_asset_count(self, repo, lib_video):
         repo.create(lib_video)
-        repo.increment_asset_count(lib_video.id, 1024)
+        repo.increment_asset_count(lib_video.id, size_delta=1024)
 
         lib = repo.get(lib_video.id)
         assert lib.asset_count == 1
         assert lib.total_size == 1024
 
-        repo.increment_asset_count(lib_video.id, 512)
+        repo.increment_asset_count(lib_video.id, size_delta=512)
         lib = repo.get(lib_video.id)
         assert lib.asset_count == 2
         assert lib.total_size == 1536
 
     def test_increment_asset_count_nonexistent(self, repo):
         # 不报错，静默忽略
-        repo.increment_asset_count("nonexistent", 100)
+        repo.increment_asset_count("nonexistent", size_delta=100)
 
     def test_decrement_asset_count(self, repo, lib_video):
         repo.create(lib_video)
-        repo.increment_asset_count(lib_video.id, 1024)
-        repo.increment_asset_count(lib_video.id, 512)
+        repo.increment_asset_count(lib_video.id, size_delta=1024)
+        repo.increment_asset_count(lib_video.id, size_delta=512)
 
-        repo.decrement_asset_count(lib_video.id, 512)
+        repo.decrement_asset_count(lib_video.id, size_delta=512)
         lib = repo.get(lib_video.id)
         assert lib.asset_count == 1
         assert lib.total_size == 1024
 
     def test_decrement_asset_count_not_below_zero(self, repo, lib_video):
         repo.create(lib_video)
-        repo.decrement_asset_count(lib_video.id, 9999)
+        repo.decrement_asset_count(lib_video.id, size_delta=9999)
         lib = repo.get(lib_video.id)
         assert lib.asset_count == 0
         assert lib.total_size == 0
 
     def test_decrement_asset_count_nonexistent(self, repo):
-        repo.decrement_asset_count("nonexistent", 100)
+        repo.decrement_asset_count("nonexistent", size_delta=100)
 
 
 # ==================== Tag ====================

tests/unit/test_mediakit_smart_clips.py

@@ -226,10 +226,10 @@ class TestGetMediakitRecommendations:
 
 
 class TestGetTemplateSegments:
-    """测试模板片段配置查询。"""
+    """测试模板片段配置查询（单一数据源：template_clip_configs）。"""
 
-    def test_returns_segments_from_new_template_system(self):
-        """新模板系统（clip_configs）有数据时优先使用。"""
+    def test_returns_segments_from_clip_configs(self):
+        """片段配置主表（clip_configs）有数据时按 order 排序返回。"""
         from app.api.routes.templates_editor.clips import _get_template_segments
 
         mock_tpl_svc = MagicMock()
@@ -241,66 +241,34 @@ class TestGetTemplateSegments:
         cc2.order = 1
         cc2.min_duration = 4.0
         cc2.max_duration = 8.0
-        mock_tpl_svc.list_clip_configs.return_value = [cc2, cc1]  # 乱序返回
+        mock_tpl_svc.list_clip_configs_for_editor.return_value = [cc2, cc1]  # 乱序返回
 
-        result = _get_template_segments("tmpl-1", mock_tpl_svc, MagicMock())
+        result = _get_template_segments("tmpl-1", "user-1", mock_tpl_svc)
         assert len(result) == 2
         assert result[0] == (0, 3.0, 5.0)
         assert result[1] == (1, 4.0, 8.0)
+        mock_tpl_svc.list_clip_configs_for_editor.assert_called_once_with("tmpl-1", "user-1")
 
-    def test_falls_back_to_old_template_segments(self):
-        """新模板系统无数据时回退到旧系统。"""
+    def test_returns_empty_when_no_configs(self):
+        """模板存在但没有片段配置时返回空列表（路由层据此返回 422）。"""
         from app.api.routes.templates_editor.clips import _get_template_segments
 
         mock_tpl_svc = MagicMock()
-        mock_tpl_svc.list_clip_configs.return_value = []
+        mock_tpl_svc.list_clip_configs_for_editor.return_value = []
 
-        with patch("app.api.routes.templates_editor.clips.SQLAlchemyTemplateRepository") as MockRepo:
-            mock_repo = MagicMock()
-            seg1 = MagicMock()
-            seg1.segment_order = 0
-            seg1.duration_min = 2.0
-            seg1.duration_max = 4.0
-            mock_repo.list_segments.return_value = [seg1]
-            MockRepo.return_value = mock_repo
+        result = _get_template_segments("tmpl-1", "user-1", mock_tpl_svc)
+        assert result == []
 
-            result = _get_template_segments("tmpl-1", mock_tpl_svc, MagicMock())
-            assert len(result) == 1
-            assert result[0] == (0, 2.0, 4.0)
-
-    def test_returns_empty_when_no_segments(self):
-        """两套系统都没有片段配置时返回空列表。"""
+    def test_missing_template_raises(self):
+        """模板不存在/无权限时服务层抛 TemplateNotFoundError（路由层据此返回 404）。"""
         from app.api.routes.templates_editor.clips import _get_template_segments
+        from app.services.edit_template_service import TemplateNotFoundError
 
         mock_tpl_svc = MagicMock()
-        mock_tpl_svc.list_clip_configs.return_value = []
+        mock_tpl_svc.list_clip_configs_for_editor.side_effect = TemplateNotFoundError("tmpl-x")
 
-        with patch("app.api.routes.templates_editor.clips.SQLAlchemyTemplateRepository") as MockRepo:
-            mock_repo = MagicMock()
-            mock_repo.list_segments.return_value = []
-            MockRepo.return_value = mock_repo
-
-            result = _get_template_segments("tmpl-1", mock_tpl_svc, MagicMock())
-            assert result == []
-
-    def test_new_system_exception_falls_back(self):
-        """新模板系统异常时回退到旧系统。"""
-        from app.api.routes.templates_editor.clips import _get_template_segments
-
-        mock_tpl_svc = MagicMock()
-        mock_tpl_svc.list_clip_configs.side_effect = RuntimeError("db error")
-
-        with patch("app.api.routes.templates_editor.clips.SQLAlchemyTemplateRepository") as MockRepo:
-            mock_repo = MagicMock()
-            seg = MagicMock()
-            seg.segment_order = 0
-            seg.duration_min = 1.0
-            seg.duration_max = 3.0
-            mock_repo.list_segments.return_value = [seg]
-            MockRepo.return_value = mock_repo
-
-            result = _get_template_segments("tmpl-1", mock_tpl_svc, MagicMock())
-            assert len(result) == 1
+        with pytest.raises(TemplateNotFoundError):
+            _get_template_segments("tmpl-x", "user-1", mock_tpl_svc)
 
 
 # ── from-assets 端点集成测试 ────────────────────────────────────────────────
@@ -358,7 +326,7 @@ def _make_tpl_svc_with_segments(segments):
     """segments: list of (order, min_dur, max_dur)"""
     svc = MagicMock()
     clip_configs = [_make_clip_config(o, mn, mx) for o, mn, mx in segments]
-    svc.list_clip_configs.return_value = clip_configs
+    svc.list_clip_configs_for_editor.return_value = clip_configs
     return svc
 
 
@@ -540,34 +508,56 @@ class TestFromAssetsByTemplateSegments:
         orders = [c["order"] for c in clips_data]
         assert orders == [0, 1, 2]
 
-    def test_no_segments_raises_400(self):
-        """模板没有 segment 配置时返回 400。"""
+    def test_no_segments_raises_422(self):
+        """模板存在但未配置片段时返回 422（与模板不存在的 404 区分）。"""
         from app.api.routes.templates_editor.clips import create_clips_from_assets_editor
         from app.api.routes.templates_editor.schemas import ClipsFromAssetsRequest
         from fastapi import HTTPException
 
         mock_tpl_svc = MagicMock()
-        mock_tpl_svc.list_clip_configs.return_value = []
+        mock_tpl_svc.list_clip_configs_for_editor.return_value = []
         mock_plan_svc = _make_plan_svc()
 
-        with patch("app.api.routes.templates_editor.clips.SQLAlchemyTemplateRepository") as MockRepo:
-            mock_repo = MagicMock()
-            mock_repo.list_segments.return_value = []
-            MockRepo.return_value = mock_repo
+        body = ClipsFromAssetsRequest(asset_ids=["a1"])
+        with pytest.raises(HTTPException) as exc_info:
+            create_clips_from_assets_editor(
+                template_id="tmpl-1",
+                body=body,
+                background_tasks=MagicMock(),
+                plan_id="plan-1",
+                services=(mock_tpl_svc, mock_plan_svc),
+                asset_repo=MagicMock(),
+                db=MagicMock(),
+                current_user=_make_auth_user(),
+            )
+        assert exc_info.value.status_code == 422
 
-            body = ClipsFromAssetsRequest(asset_ids=["a1"])
-            with pytest.raises(HTTPException) as exc_info:
-                create_clips_from_assets_editor(
-                    template_id="tmpl-1",
-                    body=body,
-                    background_tasks=MagicMock(),
-                    plan_id="plan-1",
-                    services=(mock_tpl_svc, mock_plan_svc),
-                    asset_repo=MagicMock(),
-                    db=MagicMock(),
-                    current_user=_make_auth_user(),
-                )
-            assert exc_info.value.status_code == 400
+        mock_plan_svc.replace_all_clips_transactional.assert_not_called()
+
+    def test_template_not_found_raises_404(self):
+        """模板不存在/已删除/无权限时返回 404。"""
+        from app.api.routes.templates_editor.clips import create_clips_from_assets_editor
+        from app.api.routes.templates_editor.schemas import ClipsFromAssetsRequest
+        from app.services.edit_template_service import TemplateNotFoundError
+        from fastapi import HTTPException
+
+        mock_tpl_svc = MagicMock()
+        mock_tpl_svc.list_clip_configs_for_editor.side_effect = TemplateNotFoundError("tmpl-x")
+        mock_plan_svc = _make_plan_svc()
+
+        body = ClipsFromAssetsRequest(asset_ids=["a1"])
+        with pytest.raises(HTTPException) as exc_info:
+            create_clips_from_assets_editor(
+                template_id="tmpl-x",
+                body=body,
+                background_tasks=MagicMock(),
+                plan_id="plan-1",
+                services=(mock_tpl_svc, mock_plan_svc),
+                asset_repo=MagicMock(),
+                db=MagicMock(),
+                current_user=_make_auth_user(),
+            )
+        assert exc_info.value.status_code == 404
 
         mock_plan_svc.replace_all_clips_transactional.assert_not_called()
 

tests/unit/test_pixel_perturbation.py

@@ -110,3 +110,32 @@ class TestPixelPerturbationAcceptance:
         # 至少 2 种不同组合
         unique = len(set(results))
         assert unique >= 2, f"Expected >= 2 unique filter combos, got {unique}: {results}"
+
+
+class TestIntSeedSupport:
+    """int seed 入参支持（与 get_rhythm_template(seed) 接口一致）。"""
+
+    def test_int_seed_returns_dict(self):
+        """int seed 正常返回 dict。"""
+        result = generate_pixel_perturbation(42)
+        assert isinstance(result, dict)
+        assert "filters" in result
+
+    def test_int_seed_reproducible(self):
+        """相同 int seed 结果一致。"""
+        assert generate_pixel_perturbation(42) == generate_pixel_perturbation(42)
+
+    def test_int_seed_differs_across_seeds(self):
+        """不同 int seed 大概率不同（遍历确认至少 2 种组合）。"""
+        results = {tuple(generate_pixel_perturbation(s)["filters"]) for s in range(30)}
+        assert len(results) >= 2
+
+    def test_int_seed_matches_random_obj(self):
+        """int seed 与等价 random.Random(seed) 结果一致。"""
+        assert generate_pixel_perturbation(7) == generate_pixel_perturbation(random.Random(7))
+
+    def test_none_seed_works(self):
+        """None 入参（默认随机）正常返回。"""
+        result = generate_pixel_perturbation(None)
+        assert isinstance(result, dict)
+        assert len(result["filters"]) in [2, 3]

tests/unit/test_rhythm_templates.py

@@ -1,7 +1,7 @@
 """节奏模板单元测试（Issue #1764）。
 
 覆盖：
-- RHYTHM_TEMPLATES 池定义（6 种模板）
+- RHYTHM_TEMPLATES 池定义（8 种模板，#1764 原始 6 种 + #1768 新增 2 种）
 - get_rhythm_template：根据 seed 选择模板
 - adapt_template_length：适配不同片段数
 - plan_clip_durations：按权重分配时长
@@ -26,8 +26,8 @@ class TestRhythmTemplates:
     """节奏模板池测试。"""
 
     def test_six_templates_defined(self):
-        """预设 6 种节奏模板。"""
-        assert len(RHYTHM_TEMPLATES) == 6
+        """预设 8 种节奏模板（#1764 原始 6 种 + #1768 新增 2 种）。"""
+        assert len(RHYTHM_TEMPLATES) == 8
 
     def test_average_template_is_all_ones(self):
         """第一种模板是平均（全 1）。"""

tests/unit/test_template_use_cases.py

@@ -188,6 +188,7 @@ class TestListTemplatesUseCase:
             tag="tag1",
             keyword="test",
             mode="one_take",
+            valid_only=False,
         )
 
     def test_list_pagination(self):
@@ -226,6 +227,7 @@ class TestCountTemplatesUseCase:
             tag="tag1",
             keyword="kw",
             mode="pip",
+            valid_only=False,
         )
 
 

tests/unit/test_transition_randomizer_1766.py

@@ -0,0 +1,319 @@
+"""#1766 转场位置与类型随机化测试 (packages/domain/transition_randomizer.py).
+
+覆盖：
+- TRANSITION_POOL 定义（5 种效果）
+- generate_transition_plan：硬切比例 30%-50%
+- generate_transition_plan：转场时长 0.3s ~ 0.8s
+- generate_transition_plan：jitter 在 ±0.5s 范围内
+- 不同 seed 产生不同转场序列
+- 与片段时长协同：长片段间转场更长
+- 边界情况：num_transitions=0、1 个片段
+"""
+
+from __future__ import annotations
+
+import random
+import sys
+from pathlib import Path
+
+import pytest
+
+REPO_ROOT = Path(__file__).resolve().parents[2]
+for sub in ("packages", ""):
+    p = str(REPO_ROOT / sub) if sub else str(REPO_ROOT)
+    if p not in sys.path:
+        sys.path.insert(0, p)
+
+from packages.domain import transition_randomizer as tr  # noqa: E402
+from packages.domain.transition_randomizer import (  # noqa: E402
+    CUT_RATIO_MAX,
+    CUT_RATIO_MIN,
+    HARD_CUT,
+    TIMING_JITTER_MAX,
+    TRANSITION_DURATION_MAX,
+    TRANSITION_DURATION_MIN,
+    TRANSITION_POOL,
+    _cut_probability_for_pair,
+    generate_transition_plan,
+)
+
+
+class TestTransitionPool:
+    """转场类型池定义测试。"""
+
+    def test_pool_has_5_effects(self):
+        assert len(TRANSITION_POOL) == 5
+
+    def test_pool_contains_expected_effects(self):
+        assert "dissolve" in TRANSITION_POOL
+        assert "zoomin" in TRANSITION_POOL
+        assert "slideleft" in TRANSITION_POOL
+        assert "wipeleft" in TRANSITION_POOL
+        assert "fade" in TRANSITION_POOL
+
+    def test_pool_does_not_contain_cut(self):
+        assert HARD_CUT not in TRANSITION_POOL
+
+
+class TestConstants:
+    """常量约束测试。"""
+
+    def test_duration_range(self):
+        assert TRANSITION_DURATION_MIN == 0.3
+        assert TRANSITION_DURATION_MAX == 0.8
+
+    def test_cut_ratio_range(self):
+        assert CUT_RATIO_MIN == 0.3
+        assert CUT_RATIO_MAX == 0.5
+
+    def test_jitter_max(self):
+        assert TIMING_JITTER_MAX == 0.5
+
+
+class TestCutProbability:
+    """硬切概率计算测试。"""
+
+    def test_short_clips_higher_cut_prob(self):
+        """短片段（<3s）→ 硬切概率偏高。"""
+        prob = _cut_probability_for_pair(2.0, 2.5)
+        assert prob > 0.4  # 高于基准
+
+    def test_long_clips_lower_cut_prob(self):
+        """长片段（>6s）→ 硬切概率偏低。"""
+        prob = _cut_probability_for_pair(8.0, 7.0)
+        assert prob < 0.4  # 低于基准
+
+    def test_medium_clips_base_prob(self):
+        """中等片段（3-6s）→ 基准概率。"""
+        prob = _cut_probability_for_pair(4.0, 5.0)
+        assert abs(prob - 0.4) < 1e-6
+
+    def test_probability_in_range(self):
+        """概率始终在 [MIN, MAX] 范围内。"""
+        for prev in [1.0, 3.0, 5.0, 8.0, 15.0]:
+            for next_ in [1.0, 3.0, 5.0, 8.0, 15.0]:
+                prob = _cut_probability_for_pair(prev, next_)
+                assert CUT_RATIO_MIN <= prob <= CUT_RATIO_MAX
+
+
+class TestGenerateTransitionPlan:
+    """generate_transition_plan 核心测试。"""
+
+    def test_zero_transitions_returns_empty(self):
+        assert generate_transition_plan(0) == []
+
+    def test_negative_transitions_returns_empty(self):
+        assert generate_transition_plan(-1) == []
+
+    def test_returns_correct_count(self):
+        plan = generate_transition_plan(5, rng=random.Random(42))
+        assert len(plan) == 5
+
+    def test_each_item_has_required_keys(self):
+        plan = generate_transition_plan(3, rng=random.Random(42))
+        for item in plan:
+            assert "effect" in item
+            assert "duration" in item
+            assert "jitter" in item
+
+    def test_effects_are_valid(self):
+        """所有 effect 要么是 cut 要么是 TRANSITION_POOL 中的。"""
+        plan = generate_transition_plan(20, rng=random.Random(42))
+        valid_effects = set(TRANSITION_POOL) | {HARD_CUT}
+        for item in plan:
+            assert item["effect"] in valid_effects
+
+    def test_hard_cut_ratio_in_range_many_samples(self):
+        """100 个转场点，硬切比例在 30%-50%（统计保证）。"""
+        plan = generate_transition_plan(
+            100,
+            clip_durations=[5.0] * 101,
+            rng=random.Random(42),
+        )
+        num_cuts = sum(1 for item in plan if item["effect"] == HARD_CUT)
+        ratio = num_cuts / len(plan)
+        # 统计波动允许 ±10% 的宽松范围
+        assert 0.20 <= ratio <= 0.60, f"硬切比例 {ratio:.2%} 超出宽松范围"
+        # 更严格的范围检查（±5%）
+        assert CUT_RATIO_MIN - 0.05 <= ratio <= CUT_RATIO_MAX + 0.05, f"硬切比例 {ratio:.2%} 超出 [25%, 55%] 范围"
+
+    def test_transition_duration_in_range(self):
+        """非硬切转场的时长在 [0.3, 0.8] 范围内。"""
+        plan = generate_transition_plan(30, rng=random.Random(42))
+        for item in plan:
+            if item["effect"] != HARD_CUT:
+                assert (
+                    TRANSITION_DURATION_MIN <= item["duration"] <= TRANSITION_DURATION_MAX
+                ), f"转场时长 {item['duration']} 超出 [{TRANSITION_DURATION_MIN}, {TRANSITION_DURATION_MAX}]"
+
+    def test_cut_duration_is_zero(self):
+        """硬切转场的时长必须为 0。"""
+        plan = generate_transition_plan(20, rng=random.Random(42))
+        for item in plan:
+            if item["effect"] == HARD_CUT:
+                assert item["duration"] == 0.0
+
+    def test_jitter_in_range(self):
+        """jitter 在 [-0.5, +0.5] 范围内。"""
+        plan = generate_transition_plan(30, rng=random.Random(42))
+        for item in plan:
+            assert -TIMING_JITTER_MAX <= item["jitter"] <= TIMING_JITTER_MAX, f"jitter {item['jitter']} 超出范围"
+
+    def test_cut_jitter_is_zero(self):
+        """硬切转场的 jitter 必须为 0。"""
+        plan = generate_transition_plan(20, rng=random.Random(42))
+        for item in plan:
+            if item["effect"] == HARD_CUT:
+                assert item["jitter"] == 0.0
+
+    def test_different_seeds_produce_different_plans(self):
+        """不同 seed 产生不同的转场序列（至少 2 组不同）。"""
+        plans_seen = set()
+        for seed in range(20):
+            plan = generate_transition_plan(5, rng=random.Random(seed))
+            plan_sig = tuple((item["effect"], item["duration"]) for item in plan)
+            plans_seen.add(plan_sig)
+        assert len(plans_seen) >= 2, "20 个 seed 只产生 1 种转场序列"
+
+    def test_same_seed_same_plan(self):
+        """相同 seed 产生相同的转场序列（确定性）。"""
+        plan1 = generate_transition_plan(5, rng=random.Random(42))
+        plan2 = generate_transition_plan(5, rng=random.Random(42))
+        assert plan1 == plan2
+
+    def test_long_clips_longer_transitions(self):
+        """长片段（>6s）之间的转场倾向于比短片段更长。"""
+        # 长片段
+        long_plan = generate_transition_plan(
+            20,
+            clip_durations=[10.0] * 21,
+            rng=random.Random(42),
+        )
+        # 短片段
+        short_plan = generate_transition_plan(
+            20,
+            clip_durations=[2.0] * 21,
+            rng=random.Random(42),
+        )
+        # 长片段的非硬切转场平均时长
+        long_durs = [item["duration"] for item in long_plan if item["effect"] != HARD_CUT]
+        short_durs = [item["duration"] for item in short_plan if item["effect"] != HARD_CUT]
+
+        if long_durs and short_durs:
+            avg_long = sum(long_durs) / len(long_durs)
+            avg_short = sum(short_durs) / len(short_durs)
+            # 长片段平均转场时长 >= 短片段（协同节奏）
+            assert avg_long >= avg_short * 0.95, f"长片段转场 {avg_long:.3f}s 不应显著短于短片段 {avg_short:.3f}s"
+
+    def test_clip_durations_none_uses_default(self):
+        """clip_durations=None 时使用默认值 5.0。"""
+        plan = generate_transition_plan(3, rng=random.Random(42))
+        assert len(plan) == 3
+
+    def test_fewer_clip_durations_than_needed(self):
+        """clip_durations 长度不足时用默认值补齐。"""
+        plan = generate_transition_plan(
+            5,
+            clip_durations=[4.0, 5.0],  # 只需前 2 个
+            rng=random.Random(42),
+        )
+        assert len(plan) == 5
+
+
+class TestTransitionRandomizationIntegration:
+    """转场随机化与变体生成集成测试。"""
+
+    def test_reselect_produces_different_transitions(self):
+        """多次 reselect_clips_for_variant 产生不同的转场序列。"""
+        from packages.domain.variant_plan_selector import reselect_clips_for_variant
+
+        source_clips = [
+            {
+                "order": i,
+                "asset_id": f"asset_{i}",
+                "start_time": 0.0,
+                "duration": 5.0,
+                "clip_type": "main",
+                "playback_speed": 1.0,
+                "transition_effect": "cut",
+                "transition_duration": 0.0,
+                "text_content": "",
+                "config": {},
+            }
+            for i in range(4)
+        ]
+        asset_durations = {f"asset_{i}": 30.0 for i in range(4)}
+
+        transition_seqs = set()
+        for seed in range(5):
+            rng = random.Random(seed)
+            result = reselect_clips_for_variant(
+                source_clips,
+                list(asset_durations.keys()),
+                asset_durations=asset_durations,
+                rng=rng,
+            )
+            seq = tuple(
+                (c.get("transition_effect"), round(c.get("transition_duration", 0), 2))
+                for c in result
+                if c.get("clip_type") == "main"
+            )
+            transition_seqs.add(seq)
+
+        assert len(transition_seqs) >= 2, f"5 个 seed 只产生 {len(transition_seqs)} 种转场序列"
+
+    def test_reselect_preserves_non_main_transitions(self):
+        """非 main 片段（intro/outro）的转场不被随机化。"""
+        from packages.domain.variant_plan_selector import reselect_clips_for_variant
+
+        source_clips = [
+            {
+                "order": 0,
+                "asset_id": "intro_asset",
+                "start_time": 0.0,
+                "duration": 3.0,
+                "clip_type": "intro",
+                "playback_speed": 1.0,
+                "transition_effect": "fade",
+                "transition_duration": 0.5,
+                "text_content": "",
+                "config": {},
+            },
+            {
+                "order": 1,
+                "asset_id": "a1",
+                "start_time": 0.0,
+                "duration": 5.0,
+                "clip_type": "main",
+                "playback_speed": 1.0,
+                "transition_effect": "cut",
+                "transition_duration": 0.0,
+                "text_content": "",
+                "config": {},
+            },
+            {
+                "order": 2,
+                "asset_id": "a2",
+                "start_time": 0.0,
+                "duration": 5.0,
+                "clip_type": "main",
+                "playback_speed": 1.0,
+                "transition_effect": "cut",
+                "transition_duration": 0.0,
+                "text_content": "",
+                "config": {},
+            },
+        ]
+        asset_durations = {"intro_asset": 10.0, "a1": 30.0, "a2": 30.0}
+
+        result = reselect_clips_for_variant(
+            source_clips,
+            list(asset_durations.keys()),
+            asset_durations=asset_durations,
+            rng=random.Random(42),
+        )
+
+        # intro 片段的转场保持不变
+        intro_clip = next(c for c in result if c["clip_type"] == "intro")
+        assert intro_clip["transition_effect"] == "fade"
+        assert intro_clip["transition_duration"] == 0.5

tests/unit/test_unify_template_segments.py

@@ -158,6 +158,52 @@ class TestListByUser:
         assert len(result[0].segments) == 1
         assert result[0].segments[0].duration_min == 2.0
 
+    def test_valid_only_filters_templates_without_segments(self, repo, session):
+        """#1769: valid_only=True 时排除两张片段表都没有记录的无效模板."""
+        # 有效模板：有 clip_configs
+        valid_clip = _make_template(name="有效模板-clip_configs")
+        repo.create(valid_clip)
+        repo.create_segments([_make_segment(valid_clip.id, order=1)])
+        # 有效模板：仅有旧表 template_segments 记录
+        valid_old = _make_template(name="有效模板-old_segments")
+        repo.create(valid_old)
+        old = TemplateSegmentModel(
+            id=str(uuid.uuid4()),
+            template_id=valid_old.id,
+            segment_order=1,
+            duration_min=2.0,
+            duration_max=6.0,
+        )
+        session.add(old)
+        session.commit()
+        # 无效模板：两张表都没有记录
+        invalid = _make_template(name="无效模板-无片段")
+        repo.create(invalid)
+
+        # 默认不过滤：编辑器视角能看到全部 3 个模板
+        all_templates = repo.list_by_user("u1")
+        assert len(all_templates) == 3
+        assert repo.count_by_user("u1") == 3
+
+        # valid_only=True：剪辑页视角只返回 2 个有效模板
+        valid_templates = repo.list_by_user("u1", valid_only=True)
+        assert {t.name for t in valid_templates} == {"有效模板-clip_configs", "有效模板-old_segments"}
+        assert all(len(t.segments) > 0 for t in valid_templates)
+        assert repo.count_by_user("u1", valid_only=True) == 2
+
+    def test_valid_only_with_filters_and_pagination(self, repo, session):
+        """valid_only 与其他过滤/分页条件组合使用."""
+        tpl = _make_template(name="口播模板", mode="voice_over")
+        repo.create(tpl)
+        repo.create_segments([_make_segment(tpl.id, order=1, material_type="人物")])
+        _invalid = _make_template(name="口播无效模板", mode="voice_over")
+        repo.create(_invalid)
+
+        result = repo.list_by_user("u1", mode="voice_over", valid_only=True)
+        assert len(result) == 1
+        assert result[0].name == "口播模板"
+        assert repo.count_by_user("u1", mode="voice_over", valid_only=True) == 1
+
 
 class TestCopyTemplate:
     def test_copy_writes_to_clip_configs(self, repo, session):

tests/unit/test_video_compose_service.py

@@ -8,6 +8,7 @@ from __future__ import annotations
 import sys
 from pathlib import Path
 from unittest import TestCase
+from unittest.mock import patch
 
 # 修正 import 路径
 sys.path.insert(0, str(Path(__file__).resolve().parents[2] / "apps" / "api"))
@@ -62,13 +63,14 @@ class _StubPlan:
         self,
         plan_id: str = "plan-1",
         status: EditPlanStatus = EditPlanStatus.EDITING,
+        config: dict | None = None,
     ):
         self.id = plan_id
         self.template_id = "tpl-1"
         self.name = "测试计划"
         self.status = status
         self.total_duration = 0.0
-        self.config = {}
+        self.config = config or {}
 
 
 # ── Stub 仓储 ─────────────────────────────────────────────────────────────────
@@ -712,6 +714,65 @@ class TestHasAudioTitleSubtitleFix(TestCase):
         self.assertEqual(chain.audio_label, "a0")
 
 
+# ── #1789 标题 drawtext 集成测试 ──────────────────────────────────────────────
+
+
+class TestComposeCommandTitleDrawtext(TestCase):
+    """build_compose_command 中标题 drawtext 集成测试。"""
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_title_config_injected(self, mock_font):
+        """plan.config 有 title 时，filter_complex 包含 drawtext。"""
+        mock_font.return_value = ""
+        plan = _StubPlan(config={"title": {"text": "测试标题", "font_size": 48, "position": "top"}})
+        clips = [_make_ready_clip(plan_id=plan.id)]
+        svc = _make_service(plan, clips)
+        cmd = svc.build_compose_command(plan.id, "/tmp/out.mp4")
+
+        self.assertIn("drawtext=", cmd.filter_complex)
+        self.assertIn("[composed]", cmd.filter_complex)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_title_config_alt_key(self, mock_font):
+        """plan.config['title'] 无文本时回退到 title_config。"""
+        mock_font.return_value = ""
+        plan = _StubPlan(config={"title": {}, "title_config": {"text": "备用标题", "font_size": 36}})
+        clips = [_make_ready_clip(plan_id=plan.id)]
+        svc = _make_service(plan, clips)
+        cmd = svc.build_compose_command(plan.id, "/tmp/out.mp4")
+
+        self.assertIn("drawtext=", cmd.filter_complex)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_no_title_no_drawtext(self, mock_font):
+        """无标题配置时，filter_complex 不包含 drawtext。"""
+        mock_font.return_value = ""
+        plan = _StubPlan(config={})
+        clips = [_make_ready_clip(plan_id=plan.id)]
+        svc = _make_service(plan, clips)
+        cmd = svc.build_compose_command(plan.id, "/tmp/out.mp4")
+
+        self.assertNotIn("drawtext=", cmd.filter_complex)
+
+    def test_title_config_not_dict(self):
+        """title config 为非 dict 值时不崩溃。"""
+        plan = _StubPlan(config={"title": "not a dict"})
+        clips = [_make_ready_clip(plan_id=plan.id)]
+        svc = _make_service(plan, clips)
+        cmd = svc.build_compose_command(plan.id, "/tmp/out.mp4")
+        self.assertIsNotNone(cmd)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_title_content_field(self, mock_font):
+        """title config 使用 content 字段（前端 TitleConfig 命名）。"""
+        mock_font.return_value = ""
+        plan = _StubPlan(config={"title": {"content": "内容标题", "font_size": 36}})
+        clips = [_make_ready_clip(plan_id=plan.id)]
+        svc = _make_service(plan, clips)
+        cmd = svc.build_compose_command(plan.id, "/tmp/out.mp4")
+        self.assertIn("drawtext=", cmd.filter_complex)
+
+
 if __name__ == "__main__":
     import unittest
 

tests/unit/test_video_filter_builder.py

@@ -15,6 +15,7 @@ from __future__ import annotations
 
 import unittest
 from dataclasses import FrozenInstanceError
+from unittest.mock import patch
 
 from packages.domain.edit_plan_clip import EditPlanClip, EditPlanClipStatus
 from packages.domain.template_clip_config import TransitionEffect
@@ -26,9 +27,12 @@ from packages.domain.video_filter_builder import (
     DEFAULT_TRANSITION_DURATION,
     XFADE_TRANSITION_MAP,
     ClipFilterChain,
+    _escape_drawtext_text,
+    _resolve_font_path,
     build_clip_filter,
     build_concat_filter,
     build_filter_complex,
+    build_title_drawtext_filter,
     build_xfade_filter,
     chain_filters,
     has_audio,
@@ -852,5 +856,265 @@ class TestEndToEndFilterBuilding(unittest.TestCase):
         self.assertNotIn("[a0]", filter_str)
 
 
+# ── #1789 标题 drawtext 滤镜补充覆盖率 ──────────────────────────────────────
+
+
+class TestEscapeDrawtextText(unittest.TestCase):
+    """直接测试转义函数，覆盖每一行。"""
+
+    def test_backslash_escape(self):
+        result = _escape_drawtext_text("a\\b")
+        self.assertIn("\\\\", result)
+
+    def test_single_quote_escape(self):
+        result = _escape_drawtext_text("it's")
+        self.assertIn("\\'", result)
+
+    def test_colon_escape(self):
+        result = _escape_drawtext_text("a:b")
+        self.assertIn("\\:", result)
+
+    def test_percent_escape(self):
+        result = _escape_drawtext_text("100%")
+        self.assertIn("%%", result)
+
+    def test_all_special_chars_combined(self):
+        result = _escape_drawtext_text("\\':%")
+        self.assertIn("\\\\", result)
+        self.assertIn("\\'", result)
+        self.assertIn("\\:", result)
+        self.assertIn("%%", result)
+
+    def test_no_special_chars(self):
+        result = _escape_drawtext_text("hello world")
+        self.assertEqual(result, "hello world")
+
+
+class TestResolveFontPath(unittest.TestCase):
+    """测试字体路径解析逻辑。"""
+
+    @patch("os.path.isfile")
+    def test_known_font_found(self, mock_isfile):
+        mock_isfile.side_effect = lambda p: "NotoSansCJK" in p
+        result = _resolve_font_path("思源黑体")
+        self.assertNotEqual(result, "")
+        self.assertIn("NotoSansCJK", result)
+
+    @patch("os.path.isfile")
+    def test_unknown_font_fallback(self, mock_isfile):
+        mock_isfile.side_effect = lambda p: "DejaVu" in p
+        result = _resolve_font_path("UnknownFont")
+        self.assertIn("DejaVu", result)
+
+    @patch("os.path.isfile")
+    def test_no_fonts_available(self, mock_isfile):
+        mock_isfile.return_value = False
+        result = _resolve_font_path("思源黑体")
+        self.assertEqual(result, "")
+
+    @patch("os.path.isfile")
+    def test_passthrough_font_name(self, mock_isfile):
+        mock_isfile.side_effect = lambda p: "NotoSansCJK" in p
+        result = _resolve_font_path("NotoSansCJK")
+        self.assertNotEqual(result, "")
+
+    @patch("os.path.isfile")
+    def test_font_search_first_match(self, mock_isfile):
+        mock_isfile.side_effect = lambda p: "opentype" in p
+        result = _resolve_font_path("思源黑体")
+        self.assertNotEqual(result, "")
+        self.assertIn("opentype", result)
+
+    @patch("os.path.isfile")
+    def test_font_fallback_skips_nonexistent(self, mock_isfile):
+        mock_isfile.side_effect = lambda p: "DejaVu" in p
+        result = _resolve_font_path("不存在字体")
+        self.assertIn("DejaVu", result)
+
+
+class TestDrawtextFontFileIncluded(unittest.TestCase):
+    """当字体文件存在时，fontfile 参数出现在输出中。"""
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_fontfile_in_output(self, mock_font):
+        mock_font.return_value = "/usr/share/fonts/opentype/noto/NotoSansCJK-Regular.ttc"
+        result = build_title_drawtext_filter({"text": "标题"})
+        self.assertIsNotNone(result)
+        self.assertIn("fontfile=", result)
+        self.assertIn("NotoSansCJK", result)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_fontfile_escaped(self, mock_font):
+        mock_font.return_value = "/path/with:special'chars.ttf"
+        result = build_title_drawtext_filter({"text": "标题"})
+        self.assertIsNotNone(result)
+        self.assertIn("fontfile=", result)
+
+
+class TestDrawtextFontFileNotIncluded(unittest.TestCase):
+    """当字体文件不存在时，无 fontfile 参数。"""
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_no_fontfile(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"text": "标题"})
+        self.assertIsNotNone(result)
+        self.assertNotIn("fontfile=", result)
+
+
+class TestDrawtextStrokeBranches(unittest.TestCase):
+    """stroke 各分支覆盖。"""
+
+    def test_stroke_non_bool_non_dict(self):
+        result = build_title_drawtext_filter({"text": "标题", "stroke": "yes"})
+        self.assertIsNotNone(result)
+        self.assertNotIn("borderw", result)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_stroke_dict_default_color(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"text": "标题", "stroke": {"width": 4}})
+        self.assertIsNotNone(result)
+        self.assertIn("borderw=4", result)
+        self.assertIn("bordercolor=000000", result)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_stroke_dict_enabled_false(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"text": "标题", "stroke": {"enabled": False, "width": 5}})
+        self.assertIsNotNone(result)
+        self.assertNotIn("borderw", result)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_stroke_dict_custom_color(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"text": "标题", "stroke": {"width": 2, "color": "#ff0000"}})
+        self.assertIsNotNone(result)
+        self.assertIn("bordercolor=ff0000", result)
+
+
+class TestDrawtextShadowBranches(unittest.TestCase):
+    """shadow 各分支覆盖。"""
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_shadow_dict_default_color(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"text": "标题", "shadow": {"offset_x": 5, "offset_y": 5}})
+        self.assertIsNotNone(result)
+        self.assertIn("shadowcolor=000000", result)
+        self.assertIn("shadowx=5", result)
+        self.assertIn("shadowy=5", result)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_shadow_dict_disabled(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"text": "标题", "shadow": {"enabled": False}})
+        self.assertIsNotNone(result)
+        self.assertNotIn("shadowcolor", result)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_shadow_dict_custom_color(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter(
+            {"text": "标题", "shadow": {"color": "#555555", "offset_x": 1, "offset_y": 1}}
+        )
+        self.assertIsNotNone(result)
+        self.assertIn("shadowcolor=555555", result)
+
+
+class TestDrawtextBoldFalse(unittest.TestCase):
+    def test_bold_false(self):
+        result = build_title_drawtext_filter({"text": "标题", "bold": False})
+        self.assertIsNotNone(result)
+        self.assertNotIn("font=bold", result)
+
+
+class TestDrawtextPositionBranches(unittest.TestCase):
+    """位置相关分支覆盖。"""
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_position_top_explicit(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"text": "标题", "position": "top"})
+        self.assertIsNotNone(result)
+        self.assertIn("y=50", result)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_position_center_explicit(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"text": "标题", "position": "center"})
+        self.assertIsNotNone(result)
+        self.assertIn("y=(h-text_h)/2", result)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_position_bottom_explicit(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"text": "标题", "position": "bottom"})
+        self.assertIsNotNone(result)
+        self.assertIn("y=h-text_h-50", result)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_position_custom_with_float_coords(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"text": "标题", "position": "custom", "pos_x": 100.7, "pos_y": 200.3})
+        self.assertIsNotNone(result)
+        self.assertIn("x=100", result)
+        self.assertIn("y=200", result)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_position_custom_bool_coords_fallback(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"text": "标题", "position": "custom", "pos_x": True, "pos_y": True})
+        self.assertIsNotNone(result)
+        self.assertIn("x=(w-text_w)/2", result)
+        self.assertIn("y=50", result)
+
+
+class TestDrawtextColorNoHash(unittest.TestCase):
+    def test_color_without_hash(self):
+        result = build_title_drawtext_filter({"text": "标题", "font_color": "red"})
+        self.assertIsNotNone(result)
+        self.assertIn("fontcolor=red", result)
+
+
+class TestDrawtextFieldNormalization(unittest.TestCase):
+    """字段归一化覆盖更多分支。"""
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_content_fallback(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"content": "备用标题"})
+        self.assertIsNotNone(result)
+        self.assertIn("备用标题", result)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_font_preset_fallback(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"text": "标题", "font_preset": "楷体"})
+        self.assertIsNotNone(result)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_size_fallback(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"text": "标题", "size": 72})
+        self.assertIsNotNone(result)
+        self.assertIn("fontsize=72", result)
+
+    @patch("packages.domain.video_filter_builder._resolve_font_path")
+    def test_color_fallback(self, mock_font):
+        mock_font.return_value = ""
+        result = build_title_drawtext_filter({"text": "标题", "color": "#abcdef"})
+        self.assertIsNotNone(result)
+        self.assertIn("fontcolor=abcdef", result)
+
+
+class TestDrawtextNotDictConfig(unittest.TestCase):
+    def test_string_config_returns_none(self):
+        self.assertIsNone(build_title_drawtext_filter("not a dict"))
+
+    def test_list_config_returns_none(self):
+        self.assertIsNone(build_title_drawtext_filter([1, 2, 3]))
+
+
 if __name__ == "__main__":
     unittest.main()

tests/unit/test_xfade_per_transition_1766.py

@@ -0,0 +1,261 @@
+"""#1766 xfade_builder 逐转场时长与位置微调测试.
+
+覆盖：
+- build_xfade_filter_chain：transition_durations 参数（逐转场独立时长）
+- build_xfade_filter_chain：jitters 参数（位置微调偏移）
+- 向后兼容：不传新参数时行为不变
+- TransitionEngine.build_xfade_chain：透传新参数
+"""
+
+from __future__ import annotations
+
+import sys
+from pathlib import Path
+
+import pytest
+
+REPO_ROOT = Path(__file__).resolve().parents[2]
+for sub in ("packages", ""):
+    p = str(REPO_ROOT / sub) if sub else str(REPO_ROOT)
+    if p not in sys.path:
+        sys.path.insert(0, p)
+
+from packages.domain.xfade_builder import (  # noqa: E402
+    DEFAULT_TRANSITION_DURATION,
+    build_xfade_filter_chain,
+)
+
+
+class TestBackwardCompatibility:
+    """向后兼容测试：不传新参数时行为不变。"""
+
+    def test_single_transition_duration(self):
+        """全局 transition_duration 仍有效。"""
+        result, dur = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "fade"],
+            transition_duration=0.5,
+        )
+        assert "xfade" in result
+        assert "duration=0.500" in result
+        assert dur > 0
+
+    def test_default_transition_duration(self):
+        """不传 transition_duration 时使用默认值 0.5。"""
+        result, dur = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "fade"],
+        )
+        assert "duration=0.500" in result
+
+
+class TestPerTransitionDurations:
+    """逐转场独立时长测试。"""
+
+    def test_different_durations_per_transition(self):
+        """每个转场使用不同的时长。"""
+        result, dur = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0, 5.0],
+            clip_video_labels=["v0", "v1", "v2"],
+            transitions=["cut", "fade", "dissolve"],
+            transition_durations=[0.3, 0.8],  # 第 1 个转场 0.3s，第 2 个 0.8s
+        )
+        assert "duration=0.300" in result
+        assert "duration=0.800" in result
+
+    def test_partial_durations_fallback_to_global(self):
+        """transition_durations 长度不足时 fallback 到 transition_duration。"""
+        result, dur = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0, 5.0],
+            clip_video_labels=["v0", "v1", "v2"],
+            transitions=["cut", "fade", "dissolve"],
+            transition_duration=0.5,
+            transition_durations=[0.3],  # 只有第一个，第二个 fallback 到 0.5
+        )
+        assert "duration=0.300" in result
+        assert "duration=0.500" in result
+
+    def test_empty_durations_uses_global(self):
+        """transition_durations=[] 时使用全局 transition_duration。"""
+        result, dur = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "fade"],
+            transition_duration=0.6,
+            transition_durations=[],
+        )
+        assert "duration=0.600" in result
+
+    def test_none_durations_uses_global(self):
+        """transition_durations=None 时使用全局 transition_duration。"""
+        result, dur = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "fade"],
+            transition_duration=0.6,
+            transition_durations=None,
+        )
+        assert "duration=0.600" in result
+
+    def test_duration_clamped_by_clip_length(self):
+        """转场时长不能超过相邻片段时长。"""
+        result, dur = build_xfade_filter_chain(
+            clip_durations=[2.0, 2.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "fade"],
+            transition_durations=[1.5],  # 1.5s > 2.0s * 0.4，会被钳制
+        )
+        # 钳制到可用范围内
+        assert "duration=" in result
+
+    def test_total_duration_reflects_per_transition(self):
+        """总时长反映逐转场的重叠量。"""
+        # 使用 0.3s 转场
+        _, dur_short = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "fade"],
+            transition_durations=[0.3],
+        )
+        # 使用 0.8s 转场
+        _, dur_long = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "fade"],
+            transition_durations=[0.8],
+        )
+        # 更长转场 → 更多重叠 → 总时长更短
+        assert dur_long < dur_short
+
+
+class TestJitters:
+    """位置微调 jitter 测试。"""
+
+    def test_positive_jitter_delays_transition(self):
+        """正 jitter 推迟转场（offset 增大）。"""
+        result_no_jitter, _ = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "fade"],
+            transition_duration=0.5,
+        )
+        result_with_jitter, _ = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "fade"],
+            transition_duration=0.5,
+            jitters=[0.3],  # 正 jitter：推迟转场
+        )
+        # 提取 offset 值
+        import re
+
+        offset_no = float(re.search(r"offset=([\d.]+)", result_no_jitter).group(1))
+        offset_yes = float(re.search(r"offset=([\d.]+)", result_with_jitter).group(1))
+        assert offset_yes > offset_no
+
+    def test_negative_jitter_advances_transition(self):
+        """负 jitter 提前转场（offset 减小）。"""
+        result_no_jitter, _ = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "fade"],
+            transition_duration=0.5,
+        )
+        result_with_jitter, _ = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "fade"],
+            transition_duration=0.5,
+            jitters=[-0.3],  # 负 jitter：提前转场
+        )
+        import re
+
+        offset_no = float(re.search(r"offset=([\d.]+)", result_no_jitter).group(1))
+        offset_yes = float(re.search(r"offset=([\d.]+)", result_with_jitter).group(1))
+        assert offset_yes < offset_no
+
+    def test_jitter_clamped_to_valid_range(self):
+        """jitter 不会使 offset 超出有效范围（>=0）。"""
+        result, _ = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "fade"],
+            transition_duration=0.5,
+            jitters=[-100.0],  # 极大负 jitter
+        )
+        import re
+
+        offset = float(re.search(r"offset=([\d.]+)", result).group(1))
+        assert offset >= 0.0
+
+    def test_per_transition_jitters(self):
+        """每个转场可以有独立的 jitter。"""
+        result, _ = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0, 5.0],
+            clip_video_labels=["v0", "v1", "v2"],
+            transitions=["cut", "fade", "dissolve"],
+            transition_duration=0.5,
+            jitters=[0.2, -0.1],
+        )
+        import re
+
+        offsets = [float(m.group(1)) for m in re.finditer(r"offset=([\d.]+)", result)]
+        assert len(offsets) == 2
+
+    def test_empty_jitters_no_effect(self):
+        """jitters=[] 等同于无 jitter。"""
+        result_no, _ = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "fade"],
+            transition_duration=0.5,
+        )
+        result_empty, _ = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "fade"],
+            transition_duration=0.5,
+            jitters=[],
+        )
+        assert result_no == result_empty
+
+
+class TestEdgeCases:
+    """边界情况测试。"""
+
+    def test_single_clip_with_durations(self):
+        """单片段传入 transition_durations 不报错。"""
+        result, dur = build_xfade_filter_chain(
+            clip_durations=[5.0],
+            clip_video_labels=["v0"],
+            transitions=["cut"],
+            transition_durations=[0.5],
+        )
+        assert "copy" in result
+        assert dur == 5.0
+
+    def test_empty_clips(self):
+        """空片段列表返回空字符串。"""
+        result, dur = build_xfade_filter_chain(
+            clip_durations=[],
+            clip_video_labels=[],
+            transitions=[],
+            transition_durations=[],
+            jitters=[],
+        )
+        assert result == ""
+        assert dur == 0.0
+
+    def test_all_cut_transitions(self):
+        """全硬切场景（不进入 xfade，由调用方处理 concat）。"""
+        result, dur = build_xfade_filter_chain(
+            clip_durations=[5.0, 5.0],
+            clip_video_labels=["v0", "v1"],
+            transitions=["cut", "cut"],
+            transition_durations=[0.0, 0.0],
+        )
+        # cut 转场仍然会生成 xfade 滤镜（因为底层不区分 cut）
+        # 调用方（unified_render_service）负责检测全硬切并走 concat
+        assert dur > 0