補齊 AI 自動化可觀測性指標
All checks were successful
CD Pipeline / deploy (push) Successful in 1m15s

This commit is contained in:
OoO
2026-04-29 23:41:03 +08:00
parent d58e4d0d11
commit e6a1c9d09f
14 changed files with 341 additions and 18 deletions

View File

@@ -137,13 +137,14 @@ L1 Hermes 掛 → L0 模板直出 + 🟡 「AI 分析暫不可用」
- **Phase 2**Hermes L1 接入 scheduler.run_auto_import_task + run_momo_task 兩個 exception
- **Phase 3**NemoTron 擴充 3 個新 tool (retry/query_km/silence)
- **Phase 4**:依需求擴 L3 HITL 按鈕
- **Phase 5**Prometheus metric 接入(`agent_action_total{tier,agent,event_type}``agent_latency_seconds`
- **Phase 5**Prometheus metric 接入(`momo_ai_event_router_dispatch_total``momo_ai_event_router_latency_ms_*`
## 2026-04-29 Implementation Update
- Phase 1~3 核心已落地EventRouter 分流、L2 safe action、NemoTron fallback、OpenClaw memory 寫入與通知 replay 已具測試覆蓋。
- L3 已擴展為 OpenClaw + ElephantAlphaOpenClaw 負責策略/記憶ElephantAlpha 負責 orchestration/HITL/AutoHeal bridge。
- 尚未完成Prometheus metrics 與完整線上 smoke dashboard
- 2026-04-29 已補 `/metrics` 匯出EventRouter dispatch、L2 safe action、Telegram replay、AutoHeal action 與 latency/duration
- 尚未完成:完整線上 smoke dashboard 與 Grafana/Superset 視覺化面板。
## References
- `services/event_router.py` — 分流入口Phase 1