diff --git a/docs/LOGBOOK.md b/docs/LOGBOOK.md
index 2bfaa31b2..b7c99d441 100644
--- a/docs/LOGBOOK.md
+++ b/docs/LOGBOOK.md
@@ -1,3 +1,88 @@
+## 2026-05-25｜T163 KM completion state in Telegram callback evidence
+
+**背景**：
+
+- T162 已把 KM stale completion queue 狀態加到 Work Items summary 與 Telegram detail/history 回覆。
+- 但 Operator Console 的 `/awooop/runs` TG Callback Evidence 只看得到「詳情 / 歷史有沒有送達」，還不能直接看到該 callback 對應的 KM owner-review completion 狀態。
+- 本階段目標是把 callback reply evidence 也接上 `km_stale_owner_review_completion_callback_summary_v1`，讓「送達狀態」和「後續 KM owner-review 卡點」同屏可查。
+
+**完成變更**：
+
+- `GET /api/v1/platform/runs/callback-replies` 每筆 item 新增 `km_stale_completion_summary`：
+  - read-only 查詢 completion queue；不改 incident、run、KM、governance audit。
+  - 以 callback `incident_id` 對應 completion queue item 的 `related_incident_id`。
+  - 顯示 ready / blocked / completed / failed counts、`manual_review_required`、`batch_writes_allowed`、`writes_on_read`。
+  - 若沒有對應 item，明確回 `status=no_related_owner_review`，避免 operator 以為流程消失。
+- `/zh-TW/awooop/runs` TG Callback Evidence card 新增 `KM Owner Review` 區塊：
+  - 顯示 matched / no-related / fetch-failed 狀態。
+  - 顯示 completion counts 與 guardrail。
+  - 保留 AwoooP status chain 原本 execution / MCP / source-provider evidence。
+
+**本地驗證**：
+
+```text
+python3 -m py_compile apps/api/src/services/platform_operator_service.py apps/api/src/api/v1/platform/operator_runs.py apps/api/tests/test_awooop_operator_timeline_labels.py
+/Users/ogt/.pyenv/versions/3.11.7/bin/ruff check --select F,E9 apps/api/src/services/platform_operator_service.py apps/api/src/api/v1/platform/operator_runs.py apps/api/tests/test_awooop_operator_timeline_labels.py
+DATABASE_URL='postgresql+asyncpg://test:test@localhost:5432/test?ssl=disable' /Users/ogt/.pyenv/versions/3.11.7/bin/pytest apps/api/tests/test_awooop_operator_timeline_labels.py apps/api/tests/test_ai_governance_endpoints.py -q
+  -> 104 passed
+pnpm --filter @awoooi/web exec tsc --noEmit --incremental false
+pnpm --dir apps/web exec next lint --file 'src/app/[locale]/awooop/runs/page.tsx'
+  -> exit 0, only pre-existing i18n literal warnings in this page
+python3 -m json.tool apps/web/messages/zh-TW.json
+python3 -m json.tool apps/web/messages/en.json
+git diff --check
+```
+
+**production deploy / smoke**：
+
+```text
+Code commit:
+760d6745 feat(governance): surface km completion callback evidence
+
+Deploy marker:
+fcaaad87 chore(cd): deploy 760d674 [skip ci]
+
+Gitea Actions:
+3022 / run_number 2079 Code Review -> success
+3021 / run_number 2078 CD -> success
+  3991 tests -> success
+  3992 build-and-deploy -> success
+  3993 post-deploy-checks -> success
+
+Production API:
+GET https://awoooi.wooo.work/api/v1/platform/runs/callback-replies?project_id=awoooi&per_page=3
+  -> total=2, items=2
+  -> every item includes km_stale_completion_summary
+  -> sample incident INC-20260524-16109D: status=no_related_owner_review, ready_count=10
+
+Production frontend:
+https://awoooi.wooo.work/zh-TW/awooop/runs?project_id=awoooi&incident_id=INC-20260524-16109D&callback_reply_status=sent
+  -> HTTP 200
+  -> TG Callback Evidence visible
+  -> KM Owner Review visible
+  -> no-related owner review state visible
+  -> guardrail visible: writes_on_read=false, batch_writes_allowed=false, manual_review_required=true
+  -> consoleErrors=0, pageErrors=0
+  -> screenshot: /tmp/t163-runs-callback-km-production.png
+```
+
+**production health note**：
+
+- `GET /api/v1/health` 回 `status=degraded`，不是本階段功能造成。
+- 子狀態：API / PostgreSQL / Redis / OpenClaw / SigNoz up；`ollama_gcp_a=down timeout`，`ollama_gcp_b=up`。
+- `GET /api/v1/platform/ai-route-status?workload_type=deep_rca` 顯示 selected_provider=`ollama_gcp_b`，policy order 仍為 `ollama_gcp_a -> ollama_gcp_b -> ollama_local -> gemini`，符合既定 fallback 順序。
+- 本機到 120 K8s API / SSH 暫時不可達；K8s rollout 以 Gitea CD post-deploy success 與 production API / frontend smoke 作為本階段部署證據。
+
+**目前整體進度**：
+
+- AwoooP 告警可觀測鏈：約 97.5%。
+- 低風險自動修復閉環：約 95%。
+- 前端 AI 自動化管理介面同步：約 93%。
+- CI/CD 通知與部署證據鏈：約 99%。
+- 治理告警可讀性 / 可處置性：約 91%。
+- KM owner-review / completion 可治理鏈：約 74%。
+- 完整 AI 自動化管理產品化：約 90.2%。
+
 ## 2026-05-24｜T162 KM stale completion state in Work Items summary + Telegram detail/history
 
 **觸發**：