fix(phase2+aiops): Phase 2 Agent timeout + AI Router intent hint + signoz incident_id
## Phase 2 Agent timeout(防止單步 LLM 拖垮整場辯證) - critic_agent.py: asyncio.wait_for + PHASE2_STEP_TIMEOUT_SEC=20s - diagnostician_agent.py: 同等超時保護 - solver_agent.py: 同等超時保護 ## AI Router 優化 - ai_router.py: _resolve_intent_from_context() Phase 2 agents 傳 intent_hint → Router 快路徑,不重跑 intent LLM ## SignOz Webhook 修復 - signoz_webhook.py: incident_id 補傳 send_approval_card()(移除 TODO 2026-04-05) ## Alert 處理流程修復 - webhooks.py: _should_bypass_alertmanager_llm() Host 類 NO_ACTION 告警直接走人工排查卡片,不再誤觸 LLM Agent Debate - incident_repository.py: update_incident_status 加 resolved_at 參數 - incident_service.py / proposal_service.py / incident_approval_service.py: 小修 Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
@@ -32,6 +32,36 @@ AWOOOI 已有兩套 agent 系統:
|
||||
| ADR-082 5-agent 辯證 | Alert → incident 鏈 | `ai_router.py` → Ollama/NIM | 保留 |
|
||||
| **本 ADR-095 12 Claude Code agent** | Hermes NL 對話回覆 | **`claude-agent-sdk` → Anthropic API** | **新增** |
|
||||
|
||||
### 日常工作模式(Game Rules v1)
|
||||
|
||||
除 Hermes 對外視覺分派外,12-agent 也定義為專案內部的任務協作模型:
|
||||
|
||||
- `12 agents` = 任務分工角色
|
||||
- `.agents/skills/*.md` 9 份 = 工程守則與落地規範
|
||||
- 實際工作流 = **先用 12-agent 判型與派工,再落到對應 skill 執行**
|
||||
|
||||
對照原則:
|
||||
|
||||
| 類型 | 用途 |
|
||||
|------|------|
|
||||
| `debugger / db-expert / vuln-verifier / migration-engineer` | 問題診斷與高風險鏈路處理 |
|
||||
| `frontend-designer / fullstack-engineer / refactor-specialist` | UI、功能落地與重構 |
|
||||
| `planner / onboarder / critic / web-researcher / tool-expert` | 規劃、導覽、審查、查證、工具整合 |
|
||||
|
||||
常設派工規則:
|
||||
|
||||
| 任務 | 主責 | 預設協作 |
|
||||
|------|------|---------|
|
||||
| 查根因 / 查斷點 | `debugger` | `db-expert`, `tool-expert`, `critic` |
|
||||
| learning / migration / SQL | `db-expert` | `debugger`, `refactor-specialist` |
|
||||
| UI / 戰情中心 / i18n | `frontend-designer` | `fullstack-engineer`, `critic` |
|
||||
| 前後端完整落地 | `fullstack-engineer` | `frontend-designer`, `debugger`, `db-expert` |
|
||||
| 重構 / 抽層 / 技術債 | `refactor-specialist` | `migration-engineer`, `critic`, `db-expert` |
|
||||
| Gitea / CI/CD / deploy | `migration-engineer` | `tool-expert`, `vuln-verifier`, `critic` |
|
||||
| Telegram / approval / callback / 權限 | `vuln-verifier` | `debugger`, `db-expert`, `critic` |
|
||||
|
||||
正式規則文件見:`docs/12-agent-game-rules.md`
|
||||
|
||||
### 視覺規格(單 bot,用文字前綴模擬 12 分身)
|
||||
|
||||
| Agent | Emoji | Hashtag | 中文短稱 | TG handle(文字 prefix,非真 bot)|
|
||||
@@ -128,3 +158,4 @@ CONSENSUS_WEIGHTS = {
|
||||
| 版本 | 日期 | 執行者 | 變更內容 |
|
||||
|------|------|--------|---------|
|
||||
| v1.0 | 2026-04-24 | 12-Agent 全景分析 | 初版 Proposed |
|
||||
| v1.1 | 2026-04-24 | Codex | 補入 12-agent 日常工作模式(Game Rules v1)與 9 skills 對照 |
|
||||
|
||||
Reference in New Issue
Block a user