9cdbc41808
All 8 tickets' acceptance criteria checked off (incl. the real OpenClaw E2E verified on t480). Trivial import-sort fix from ruff included.
18 lines
1.2 KiB
Markdown
18 lines
1.2 KiB
Markdown
# 01 — 智能评估数据模型 + 迁移
|
||
|
||
**What to build:** 创建智能评估的三个数据表（IntelligentEval、IntelligentEvalSession、IntelligentEvalMessage）及对应的 Repository 类，走 Alembic batch mode 迁移。
|
||
|
||
**Blocked by:** 无
|
||
|
||
**Acceptance criteria:**
|
||
- [x] `IntelligentEvalDB` 表定义：id, name, target_id, status, goal, seeds(JSON), intent, role_description, plan(JSON), plan_feedback, time_window_hours, report(JSON), created_at, updated_at, started_at, completed_at
|
||
- [x] `IntelligentEvalSessionDB` 表定义：id, eval_id, target_id, persona(JSON), goal, dimension, status, verdict(JSON), turn_count, created_at, closed_at
|
||
- [x] `IntelligentEvalMessageDB` 表定义：id, session_id, role, content, latency_ms, created_at
|
||
- [x] `IntelligentEvalStatus` 枚举：draft, planning, pending_approval, executing, completed, cancelled, failed
|
||
- [x] `IntelligentEvalSessionStatus` 枚举：running, completed, failed, expired
|
||
- [x] `IntelligentEvalRepository`：CRUD + list + 状态转换方法
|
||
- [x] `IntelligentEvalSessionRepository`：CRUD + list_by_eval
|
||
- [x] `IntelligentEvalMessageRepository`：create + list_by_session
|
||
- [x] Alembic 迁移（batch mode）通过
|
||
- [x] `init_db()` 的 `create_all()` 能建出三张表
|