wooo/awoooi
1
1
Fork 0
Code Issues 2 Pull Requests 1 Actions 2 Packages Projects Releases Wiki Activity

feat(governance): 新增 AI Agent 市場雷達讀回
Some checks failed
Code Review / ai-code-review (push) Successful in 13s
Details
CD Pipeline / tests (push) Successful in 1m54s
Details
CD Pipeline / build-and-deploy (push) Successful in 5m8s
Details
CD Pipeline / post-deploy-checks (push) Failing after 1m21s
Details
Ansible / Reboot Recovery Contract / validate (push) Has been cancelled
Details

Browse Source
This commit is contained in:
Your Name
2026-06-25 10:43:16 +08:00
parent 279f953144
commit ea0d697e51
19 changed files with 5433 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

71
docs/operations/AI-AGENT-MARKET-RADAR-READBACK-2026-06-25.md Normal file
View File

@@ -0,0 +1,71 @@
# AI Agent 市場雷達與近期變更盤點
- 產生時間：`2026-06-25T02:39:10.207391+00:00`
- 整體治理完成度：`42.2%`
- 市場雷達完成度：`100.0%`
- 候選 Agent：`13`
- 官方 / 主要來源：`34`
- 來源失敗：`0`
- 需要重新審查候選：`13`
- 仍被整合 gate 擋下：`13`
- OpenClaw 取代批准：`0`
## 近期變更盤點
| 優先級 | 工作線 | 狀態 | 進度 | 下一步 |
|---|---|---|---:|---|
| `P0` | Product Governance Owner Response Dashboard / handoff 收斂 | `read_model_ready_runtime_blocked` | `100%` | Owner questions 與 boundary acknowledgements 仍需逐項回覆。 |
| `P0` | Status Cleanup Dashboard read-only API 正式化 | `blocked_status_cleanup_apply_not_authorized` | `100%` | apply_allowed=false 前不得更新 project status 或 memory。 |
| `P0` | Wazuh / IwoooS 可視性邊界 | `blocked_waiting_manager_agent_registry_readback` | `35%` | 等待 manager agent registry readback 與 live route readback。 |
| `P0` | AI Agent market watch 2026-06-25 | `market_refresh_done_integration_blocked` | `100%` | 更新 scorecard 並進入 offline replay gate，不得直接替換。 |
| `P1` | 日報 / 週報 / 月報數據化報告 | `report_contract_defined_runtime_delivery_blocked` | `65%` | 接 Agent 工作量、Telegram receipt 與 human-review queue。 |
| `P1` | 工具 / 套件 / 服務 / 主機版本新鮮度 | `read_only_inventory_defined_update_execution_blocked` | `55%` | 定期產生版本 freshness snapshot；中低風險可 auto proposal，高風險維持人工審核。 |
## 市場主流做法對齊
| 做法 | AWOOOI 判定 | 下一步 |
|---|---|---|
| 多 Agent handoff / specialist delegation | `partially_modeled` | 將 OpenClaw / Hermes / NemoTron handoff 事件寫入可讀 timeline。 |
| Tracing / tool call / guardrail observability | `missing_unified_trace` | 建立 Agent run trace id，串接報告、Telegram receipt 與 replay outcome。 |
| Durable execution / persistence / human-in-the-loop | `needed_for_incident_loop` | 優先把 incident workflow kernel 設計成可暫停、恢復、審核與重放。 |
| MCP / A2A / enterprise multi-agent interoperability | `watch_and_design` | MCP server 先做 read-only tool registry，再開 write adapter。 |
| Evaluation / replay / profiling before integration | `strong_fit_for_nemotron` | NemoTron 維持 smoke / replay / evaluator，不直接接 production routing。 |
| Agent SDK as programmable code/ops remediator | `candidate_for_remediation_lane` | 只允許 no-write replay 與 patch proposal，禁止自動 merge / deploy。 |
| Enterprise-scale ADK with evaluation and observability | `candidate_for_google_stack_review` | 先納入 weekly watch，成本與資料邊界審核後才可 adapter。 |
## Agent 專業角色安排
| Agent / 候選 | 建議角色 | Gate 狀態 | 下一步 |
|---|---|---|---|
| OpenClaw incumbent | 生產仲裁者 / production decision core | `production_baseline` | formal_replacement_adr_and_promotion_gate_required |
| NVIDIA NeMo Agent Toolkit + Nemotron Fabric | 離線 replay、模型能力評估、合約輸出 smoke gate | `integration_blocked` | refresh_source_evidence_then_5_record_smoke_only |
| NousResearch Hermes Agent | 知識記憶、證據草稿、長期技能庫候選 | `watch_only_blocked` | continue_watch_only_until_primary_source_evidence_is_sufficient |
| OpenAI Agents SDK Coordinator | Coordinator / handoff / tracing / guardrail 候選 | `integration_blocked` | refresh_scorecard_then_offline_replay_or_promotion_gate |
| LangGraph Incident Kernel | durable incident workflow kernel 候選 | `integration_blocked` | refresh_scorecard_then_offline_replay_or_promotion_gate |
| Claude Agent SDK Remediator | DevOps / code remediation patch proposal 候選 | `integration_blocked` | refresh_scorecard_then_offline_replay_or_promotion_gate |
| Microsoft Agent Framework | MCP / A2A enterprise workflow 候選 | `integration_blocked` | create_no_sdk_no_api_adapter_then_offline_replay |
| Google Agent Development Kit Stack | Gemini / Vertex agent stack 候選 | `integration_blocked` | create_no_sdk_no_api_adapter_then_offline_replay |
| CrewAI Flows + Crews | 快速多 Agent prototype 候選 | `integration_blocked` | create_no_sdk_no_api_adapter_then_offline_replay |
## 優先工作清單
| 順序 | 工作 | 風險 | 自動化模式 | 完成定義 |
|---:|---|---|---|---|
| 1 | 固定每週 AI Agent market watch 並產生治理 snapshot | `low` | `agent_auto_read_only` | 每週一 09:00 Asia/Taipei 有 watch / integration / discovery / promotion / governance 五份 artifacts。 |
| 2 | 刷新 2026-06-25 market capability scorecard | `medium` | `agent_propose_owner_review` | OpenAI / LangGraph / NeMo-Nemotron / Claude / Microsoft / Google / CrewAI 均有新版官方來源與分數差異。 |
| 3 | 建立 50 筆歷史 incident offline replay queue | `medium` | `agent_auto_prepare_human_approve_run` | replay fixture 不含 secret，候選結果可與 OpenClaw baseline 比較。 |
| 4 | Agent 溝通 / 學習 / 成長可視化 readback | `medium` | `agent_auto_read_model` | 每個 Agent 的 handoff、decision、learning writeback、review score 與 blocked action 可被前端和報告讀到。 |
| 5 | Telegram Bot 報告與高風險審核橋接 | `high` | `human_approve_before_send_or_action` | 低中風險只告警回報，高風險需要 Telegram approval token / owner response 才能執行。 |
| 6 | 工具、套件、服務、主機版本自動 freshness 盤點 | `medium` | `agent_auto_scan_agent_propose` | 套件、服務、主機、MCP、AI provider、模型版本都有 stale / upgrade / rollback / approval gate。 |
## 禁止越界
- `replacement_decisions_approved=0`
- `replay_candidates_approved=0`
- `sdk_installations_approved=0`
- `paid_api_calls_approved=0`
- `shadow_or_canary_approved=0`
- `production_routing_approved=false`
- `status_cleanup_apply_allowed=false`
- `memory_write_authorized=false`
- `telegram_send_approved=false`

71
docs/operations/AI-AGENT-MARKET-RADAR-READBACK.md Normal file
View File

@@ -0,0 +1,71 @@
# AI Agent 市場雷達與近期變更盤點
- 產生時間：`2026-06-25T02:39:10.207391+00:00`
- 整體治理完成度：`42.2%`
- 市場雷達完成度：`100.0%`
- 候選 Agent：`13`
- 官方 / 主要來源：`34`
- 來源失敗：`0`
- 需要重新審查候選：`13`
- 仍被整合 gate 擋下：`13`
- OpenClaw 取代批准：`0`
## 近期變更盤點
| 優先級 | 工作線 | 狀態 | 進度 | 下一步 |
|---|---|---|---:|---|
| `P0` | Product Governance Owner Response Dashboard / handoff 收斂 | `read_model_ready_runtime_blocked` | `100%` | Owner questions 與 boundary acknowledgements 仍需逐項回覆。 |
| `P0` | Status Cleanup Dashboard read-only API 正式化 | `blocked_status_cleanup_apply_not_authorized` | `100%` | apply_allowed=false 前不得更新 project status 或 memory。 |
| `P0` | Wazuh / IwoooS 可視性邊界 | `blocked_waiting_manager_agent_registry_readback` | `35%` | 等待 manager agent registry readback 與 live route readback。 |
| `P0` | AI Agent market watch 2026-06-25 | `market_refresh_done_integration_blocked` | `100%` | 更新 scorecard 並進入 offline replay gate，不得直接替換。 |
| `P1` | 日報 / 週報 / 月報數據化報告 | `report_contract_defined_runtime_delivery_blocked` | `65%` | 接 Agent 工作量、Telegram receipt 與 human-review queue。 |
| `P1` | 工具 / 套件 / 服務 / 主機版本新鮮度 | `read_only_inventory_defined_update_execution_blocked` | `55%` | 定期產生版本 freshness snapshot；中低風險可 auto proposal，高風險維持人工審核。 |
## 市場主流做法對齊
| 做法 | AWOOOI 判定 | 下一步 |
|---|---|---|
| 多 Agent handoff / specialist delegation | `partially_modeled` | 將 OpenClaw / Hermes / NemoTron handoff 事件寫入可讀 timeline。 |
| Tracing / tool call / guardrail observability | `missing_unified_trace` | 建立 Agent run trace id，串接報告、Telegram receipt 與 replay outcome。 |
| Durable execution / persistence / human-in-the-loop | `needed_for_incident_loop` | 優先把 incident workflow kernel 設計成可暫停、恢復、審核與重放。 |
| MCP / A2A / enterprise multi-agent interoperability | `watch_and_design` | MCP server 先做 read-only tool registry，再開 write adapter。 |
| Evaluation / replay / profiling before integration | `strong_fit_for_nemotron` | NemoTron 維持 smoke / replay / evaluator，不直接接 production routing。 |
| Agent SDK as programmable code/ops remediator | `candidate_for_remediation_lane` | 只允許 no-write replay 與 patch proposal，禁止自動 merge / deploy。 |
| Enterprise-scale ADK with evaluation and observability | `candidate_for_google_stack_review` | 先納入 weekly watch，成本與資料邊界審核後才可 adapter。 |
## Agent 專業角色安排
| Agent / 候選 | 建議角色 | Gate 狀態 | 下一步 |
|---|---|---|---|
| OpenClaw incumbent | 生產仲裁者 / production decision core | `production_baseline` | formal_replacement_adr_and_promotion_gate_required |
| NVIDIA NeMo Agent Toolkit + Nemotron Fabric | 離線 replay、模型能力評估、合約輸出 smoke gate | `integration_blocked` | refresh_source_evidence_then_5_record_smoke_only |
| NousResearch Hermes Agent | 知識記憶、證據草稿、長期技能庫候選 | `watch_only_blocked` | continue_watch_only_until_primary_source_evidence_is_sufficient |
| OpenAI Agents SDK Coordinator | Coordinator / handoff / tracing / guardrail 候選 | `integration_blocked` | refresh_scorecard_then_offline_replay_or_promotion_gate |
| LangGraph Incident Kernel | durable incident workflow kernel 候選 | `integration_blocked` | refresh_scorecard_then_offline_replay_or_promotion_gate |
| Claude Agent SDK Remediator | DevOps / code remediation patch proposal 候選 | `integration_blocked` | refresh_scorecard_then_offline_replay_or_promotion_gate |
| Microsoft Agent Framework | MCP / A2A enterprise workflow 候選 | `integration_blocked` | create_no_sdk_no_api_adapter_then_offline_replay |
| Google Agent Development Kit Stack | Gemini / Vertex agent stack 候選 | `integration_blocked` | create_no_sdk_no_api_adapter_then_offline_replay |
| CrewAI Flows + Crews | 快速多 Agent prototype 候選 | `integration_blocked` | create_no_sdk_no_api_adapter_then_offline_replay |
## 優先工作清單
| 順序 | 工作 | 風險 | 自動化模式 | 完成定義 |
|---:|---|---|---|---|
| 1 | 固定每週 AI Agent market watch 並產生治理 snapshot | `low` | `agent_auto_read_only` | 每週一 09:00 Asia/Taipei 有 watch / integration / discovery / promotion / governance 五份 artifacts。 |
| 2 | 刷新 2026-06-25 market capability scorecard | `medium` | `agent_propose_owner_review` | OpenAI / LangGraph / NeMo-Nemotron / Claude / Microsoft / Google / CrewAI 均有新版官方來源與分數差異。 |
| 3 | 建立 50 筆歷史 incident offline replay queue | `medium` | `agent_auto_prepare_human_approve_run` | replay fixture 不含 secret，候選結果可與 OpenClaw baseline 比較。 |
| 4 | Agent 溝通 / 學習 / 成長可視化 readback | `medium` | `agent_auto_read_model` | 每個 Agent 的 handoff、decision、learning writeback、review score 與 blocked action 可被前端和報告讀到。 |
| 5 | Telegram Bot 報告與高風險審核橋接 | `high` | `human_approve_before_send_or_action` | 低中風險只告警回報，高風險需要 Telegram approval token / owner response 才能執行。 |
| 6 | 工具、套件、服務、主機版本自動 freshness 盤點 | `medium` | `agent_auto_scan_agent_propose` | 套件、服務、主機、MCP、AI provider、模型版本都有 stale / upgrade / rollback / approval gate。 |
## 禁止越界
- `replacement_decisions_approved=0`
- `replay_candidates_approved=0`
- `sdk_installations_approved=0`
- `paid_api_calls_approved=0`
- `shadow_or_canary_approved=0`
- `production_routing_approved=false`
- `status_cleanup_apply_allowed=false`
- `memory_write_authorized=false`
- `telegram_send_approved=false`

477
docs/operations/ai-agent-market-radar-readback.snapshot.json Normal file
View File

@@ -0,0 +1,477 @@
{
"blocked_gates": [
"replacement_decisions_approved=0",
"replay_candidates_approved=0",
"sdk_installations_approved=0",
"paid_api_calls_approved=0",
"shadow_or_canary_approved=0",
"production_routing_approved=false",
"status_cleanup_apply_allowed=false",
"memory_write_authorized=false",
"telegram_send_approved=false"
],
"candidate_role_plan": [
{
"candidate_id": "openclaw_incumbent",
"display_name": "OpenClaw incumbent",
"gate_status": "production_baseline",
"next_gate": "formal_replacement_adr_and_promotion_gate_required",
"recommended_role": "生產仲裁者 / production decision core"
},
{
"candidate_id": "nemo_nemotron_fabric",
"display_name": "NVIDIA NeMo Agent Toolkit + Nemotron Fabric",
"gate_status": "integration_blocked",
"next_gate": "refresh_source_evidence_then_5_record_smoke_only",
"recommended_role": "離線 replay、模型能力評估、合約輸出 smoke gate"
},
{
"candidate_id": "hermes_agent_personal_platform",
"display_name": "NousResearch Hermes Agent",
"gate_status": "watch_only_blocked",
"next_gate": "continue_watch_only_until_primary_source_evidence_is_sufficient",
"recommended_role": "知識記憶、證據草稿、長期技能庫候選"
},
{
"candidate_id": "openai_agents_sdk_coordinator",
"display_name": "OpenAI Agents SDK Coordinator",
"gate_status": "integration_blocked",
"next_gate": "refresh_scorecard_then_offline_replay_or_promotion_gate",
"recommended_role": "Coordinator / handoff / tracing / guardrail 候選"
},
{
"candidate_id": "langgraph_incident_kernel",
"display_name": "LangGraph Incident Kernel",
"gate_status": "integration_blocked",
"next_gate": "refresh_scorecard_then_offline_replay_or_promotion_gate",
"recommended_role": "durable incident workflow kernel 候選"
},
{
"candidate_id": "claude_agent_sdk_remediator",
"display_name": "Claude Agent SDK Remediator",
"gate_status": "integration_blocked",
"next_gate": "refresh_scorecard_then_offline_replay_or_promotion_gate",
"recommended_role": "DevOps / code remediation patch proposal 候選"
},
{
"candidate_id": "microsoft_agent_framework",
"display_name": "Microsoft Agent Framework",
"gate_status": "integration_blocked",
"next_gate": "create_no_sdk_no_api_adapter_then_offline_replay",
"recommended_role": "MCP / A2A enterprise workflow 候選"
},
{
"candidate_id": "google_adk_stack",
"display_name": "Google Agent Development Kit Stack",
"gate_status": "integration_blocked",
"next_gate": "create_no_sdk_no_api_adapter_then_offline_replay",
"recommended_role": "Gemini / Vertex agent stack 候選"
},
{
"candidate_id": "crewai_flows_crews",
"display_name": "CrewAI Flows + Crews",
"gate_status": "integration_blocked",
"next_gate": "create_no_sdk_no_api_adapter_then_offline_replay",
"recommended_role": "快速多 Agent prototype 候選"
}
],
"generated_at": "2026-06-25T02:39:10.207391+00:00",
"market_practice_alignment": [
{
"awoooi_status": "partially_modeled",
"next_step": "將 OpenClaw / Hermes / NemoTron handoff 事件寫入可讀 timeline。",
"practice": "多 Agent handoff / specialist delegation",
"source": "https://openai.github.io/openai-agents-python/handoffs/"
},
{
"awoooi_status": "missing_unified_trace",
"next_step": "建立 Agent run trace id，串接報告、Telegram receipt 與 replay outcome。",
"practice": "Tracing / tool call / guardrail observability",
"source": "https://openai.github.io/openai-agents-python/tracing/"
},
{
"awoooi_status": "needed_for_incident_loop",
"next_step": "優先把 incident workflow kernel 設計成可暫停、恢復、審核與重放。",
"practice": "Durable execution / persistence / human-in-the-loop",
"source": "https://docs.langchain.com/oss/python/langgraph/overview"
},
{
"awoooi_status": "watch_and_design",
"next_step": "MCP server 先做 read-only tool registry，再開 write adapter。",
"practice": "MCP / A2A / enterprise multi-agent interoperability",
"source": "https://learn.microsoft.com/en-us/agent-framework/overview/"
},
{
"awoooi_status": "strong_fit_for_nemotron",
"next_step": "NemoTron 維持 smoke / replay / evaluator，不直接接 production routing。",
"practice": "Evaluation / replay / profiling before integration",
"source": "https://docs.nvidia.com/nemo/agent-toolkit/latest/index.html"
},
{
"awoooi_status": "candidate_for_remediation_lane",
"next_step": "只允許 no-write replay 與 patch proposal，禁止自動 merge / deploy。",
"practice": "Agent SDK as programmable code/ops remediator",
"source": "https://code.claude.com/docs/en/agent-sdk/overview"
},
{
"awoooi_status": "candidate_for_google_stack_review",
"next_step": "先納入 weekly watch，成本與資料邊界審核後才可 adapter。",
"practice": "Enterprise-scale ADK with evaluation and observability",
"source": "https://docs.cloud.google.com/gemini-enterprise-agent-platform/build/adk"
}
],
"market_source_freshness": [
{
"candidate_id": "openai_agents_sdk_coordinator",
"changed": true,
"decision": "changed_requires_replay_readiness_review",
"display_name": "OpenAI Agents SDK Coordinator",
"versions": [
{
"changed": true,
"published_at": null,
"source_id": "openai_agents_docs",
"status": "ok",
"version": null
},
{
"changed": false,
"published_at": null,
"source_id": "openai_agents_tracing_docs",
"status": "ok",
"version": null
},
{
"changed": false,
"published_at": null,
"source_id": "openai_agents_handoffs_docs",
"status": "ok",
"version": null
},
{
"changed": true,
"published_at": "2026-06-24T05:15:31.741499Z",
"source_id": "openai_agents_python_pypi",
"status": "ok",
"version": "0.17.7"
},
{
"changed": true,
"published_at": "2026-06-24T04:02:12.610Z",
"source_id": "openai_agents_typescript_npm",
"status": "ok",
"version": "0.12.0"
}
]
},
{
"candidate_id": "langgraph_incident_kernel",
"changed": true,
"decision": "changed_requires_replay_readiness_review",
"display_name": "LangGraph Incident Kernel",
"versions": [
{
"changed": true,
"published_at": null,
"source_id": "langgraph_docs",
"status": "ok",
"version": null
},
{
"changed": true,
"published_at": "2026-06-18T20:58:20.335564Z",
"source_id": "langgraph_pypi",
"status": "ok",
"version": "1.2.6"
},
{
"changed": true,
"published_at": "2026-06-18T20:58:32Z",
"source_id": "langgraph_github_release",
"status": "ok",
"version": "1.2.6"
}
]
},
{
"candidate_id": "nemo_nemotron_fabric",
"changed": true,
"decision": "changed_requires_replay_readiness_review",
"display_name": "NVIDIA NeMo Agent Toolkit + Nemotron Fabric",
"versions": [
{
"changed": true,
"published_at": null,
"source_id": "nvidia_nemo_agent_toolkit_docs",
"status": "ok",
"version": null
},
{
"changed": false,
"published_at": null,
"source_id": "nvidia_nemotron_developer_page",
"status": "ok",
"version": null
},
{
"changed": false,
"published_at": null,
"source_id": "nvidia_nim_llm_docs",
"status": "ok",
"version": null
},
{
"changed": true,
"published_at": null,
"source_id": "nvidia_build_models",
"status": "ok",
"version": null
}
]
},
{
"candidate_id": "claude_agent_sdk_remediator",
"changed": true,
"decision": "changed_requires_replay_readiness_review",
"display_name": "Claude Agent SDK Remediator",
"versions": [
{
"changed": true,
"published_at": null,
"source_id": "claude_agent_sdk_docs",
"status": "ok",
"version": null
},
{
"changed": true,
"published_at": null,
"source_id": "anthropic_api_docs",
"status": "ok",
"version": null
}
]
},
{
"candidate_id": "google_adk_stack",
"changed": true,
"decision": "changed_requires_replay_readiness_review",
"display_name": "Google Agent Development Kit Stack",
"versions": [
{
"changed": true,
"published_at": null,
"source_id": "google_adk_docs",
"status": "ok",
"version": null
},
{
"changed": true,
"published_at": "2026-06-18T18:47:06.323661Z",
"source_id": "google_adk_pypi",
"status": "ok",
"version": "2.3.0"
},
{
"changed": true,
"published_at": "2026-06-18T18:45:04Z",
"source_id": "google_adk_github_release",
"status": "ok",
"version": "v2.3.0"
}
]
},
{
"candidate_id": "microsoft_agent_framework",
"changed": true,
"decision": "changed_requires_replay_readiness_review",
"display_name": "Microsoft Agent Framework",
"versions": [
{
"changed": true,
"published_at": null,
"source_id": "microsoft_agent_framework_docs",
"status": "ok",
"version": null
},
{
"changed": true,
"published_at": "2026-06-23T21:18:26Z",
"source_id": "microsoft_agent_framework_github_release",
"status": "ok",
"version": "dotnet-1.11.0"
}
]
},
{
"candidate_id": "crewai_flows_crews",
"changed": true,
"decision": "changed_requires_replay_readiness_review",
"display_name": "CrewAI Flows + Crews",
"versions": [
{
"changed": true,
"published_at": null,
"source_id": "crewai_docs",
"status": "ok",
"version": null
},
{
"changed": true,
"published_at": "2026-06-11T17:14:39.912300Z",
"source_id": "crewai_pypi",
"status": "ok",
"version": "1.14.7"
},
{
"changed": true,
"published_at": "2026-06-11T17:13:46Z",
"source_id": "crewai_github_release",
"status": "ok",
"version": "1.14.7"
}
]
}
],
"next_report_contract": {
"agent_auto_allowed_for": [
"read-only market watch",
"read-only package/version freshness snapshot",
"low-risk evidence aggregation",
"no-send report draft",
"offline deterministic replay fixture preparation"
],
"daily": "每日彙整 Agent 工作量、告警、blocked gates、低中風險自動處理與高風險待審。",
"human_review_required_for": [
"高風險主機寫入",
"付費 provider 或 token 上限變更",
"新 SDK / 新 MCP server / 新 runtime component",
"OpenClaw production routing replacement",
"Telegram Bot 發送策略變更"
],
"monthly": "每月執行正式 market scorecard review，決定是否提出 replay、shadow 或 replacement ADR。",
"weekly": "每週刷新 market watch、版本新鮮度、replay queue、成本/依賴 gate 與候選優先級。"
},
"policy": {
"host_write_approved": false,
"openclaw_replacement_approved": false,
"paid_api_calls_approved": false,
"production_routing_approved": false,
"raw_chat_history_synced": false,
"read_only": true,
"replay_candidate_approved": false,
"sdk_installation_approved": false,
"shadow_or_canary_approved": false,
"telegram_send_approved": false,
"workflow_modification_approved": false
},
"priority_workplan": [
{
"automation_mode": "agent_auto_read_only",
"done_definition": "每週一 09:00 Asia/Taipei 有 watch / integration / discovery / promotion / governance 五份 artifacts。",
"order": 1,
"risk": "low",
"work_item": "固定每週 AI Agent market watch 並產生治理 snapshot"
},
{
"automation_mode": "agent_propose_owner_review",
"done_definition": "OpenAI / LangGraph / NeMo-Nemotron / Claude / Microsoft / Google / CrewAI 均有新版官方來源與分數差異。",
"order": 2,
"risk": "medium",
"work_item": "刷新 2026-06-25 market capability scorecard"
},
{
"automation_mode": "agent_auto_prepare_human_approve_run",
"done_definition": "replay fixture 不含 secret，候選結果可與 OpenClaw baseline 比較。",
"order": 3,
"risk": "medium",
"work_item": "建立 50 筆歷史 incident offline replay queue"
},
{
"automation_mode": "agent_auto_read_model",
"done_definition": "每個 Agent 的 handoff、decision、learning writeback、review score 與 blocked action 可被前端和報告讀到。",
"order": 4,
"risk": "medium",
"work_item": "Agent 溝通 / 學習 / 成長可視化 readback"
},
{
"automation_mode": "human_approve_before_send_or_action",
"done_definition": "低中風險只告警回報，高風險需要 Telegram approval token / owner response 才能執行。",
"order": 5,
"risk": "high",
"work_item": "Telegram Bot 報告與高風險審核橋接"
},
{
"automation_mode": "agent_auto_scan_agent_propose",
"done_definition": "套件、服務、主機、MCP、AI provider、模型版本都有 stale / upgrade / rollback / approval gate。",
"order": 6,
"risk": "medium",
"work_item": "工具、套件、服務、主機版本自動 freshness 盤點"
}
],
"recent_change_inventory": [
{
"completion_percent": 100,
"next_gate": "Owner questions 與 boundary acknowledgements 仍需逐項回覆。",
"priority": "P0",
"status": "read_model_ready_runtime_blocked",
"title": "Product Governance Owner Response Dashboard / handoff 收斂"
},
{
"completion_percent": 100,
"next_gate": "apply_allowed=false 前不得更新 project status 或 memory。",
"priority": "P0",
"status": "blocked_status_cleanup_apply_not_authorized",
"title": "Status Cleanup Dashboard read-only API 正式化"
},
{
"completion_percent": 35,
"next_gate": "等待 manager agent registry readback 與 live route readback。",
"priority": "P0",
"status": "blocked_waiting_manager_agent_registry_readback",
"title": "Wazuh / IwoooS 可視性邊界"
},
{
"completion_percent": 100,
"next_gate": "更新 scorecard 並進入 offline replay gate，不得直接替換。",
"priority": "P0",
"status": "market_refresh_done_integration_blocked",
"title": "AI Agent market watch 2026-06-25"
},
{
"completion_percent": 65,
"next_gate": "接 Agent 工作量、Telegram receipt 與 human-review queue。",
"priority": "P1",
"status": "report_contract_defined_runtime_delivery_blocked",
"title": "日報 / 週報 / 月報數據化報告"
},
{
"completion_percent": 55,
"next_gate": "定期產生版本 freshness snapshot；中低風險可 auto proposal，高風險維持人工審核。",
"priority": "P1",
"status": "read_only_inventory_defined_update_execution_blocked",
"title": "工具 / 套件 / 服務 / 主機版本新鮮度"
}
],
"schema_version": "ai_agent_market_radar_readback_v1",
"source_scope": {
"gitea_main_deploy_marker": "279f9531",
"market_governance_snapshot": "docs/evaluations/agent_market_governance_snapshot_2026-06-25.json",
"market_watch_report": "docs/evaluations/agent_market_watch_report_2026-06-25.json",
"project_handoff_basis": "Codex Start Here handoff generated 2026-06-25",
"scope_note": "盤點範圍涵蓋近期 Gitea 主線、治理 handoff、AI Agent market watch 與 Status Cleanup gates；不包含 raw chat history。",
"status_cleanup_dashboard": "docs/operations/awoooi-status-cleanup-dashboard.snapshot.json"
},
"summary": {
"changed_candidates": 13,
"integration_blocked_candidates": 13,
"market_candidates": 13,
"market_sources": 34,
"market_watch_completion_percent": 100.0,
"overall_completion_percent": 42.2,
"recommended_watch_additions": 5,
"replacement_decisions_approved": 0,
"source_failures": 0,
"status": "market_refresh_done_integration_blocked",
"status_cleanup_dashboard_percent": 41.9
}
}