From 899635cc63a257c7298ac1852d4e0f28cfd69f55 Mon Sep 17 00:00:00 2001
From: Your Name <you@example.com>
Date: Sun, 28 Jun 2026 08:44:45 +0800
Subject: [PATCH 1/6] docs(runner): record 110 fail-closed pressure exception
 [skip ci]

---
 AGENTS.md                                       |  2 ++
 docs/HARD_RULES.md                              | 14 ++++++++++++++
 docs/LOGBOOK.md                                 | 17 +++++++++++++++++
 ...26-04-15-MASTER-ai-autonomous-flywheel-v2.md |  2 +-
 ops/runner/README.md                            | 13 +++++++++++++
 scripts/ci/wait-host-web-build-pressure.sh      |  3 ++-
 6 files changed, 49 insertions(+), 2 deletions(-)

diff --git a/AGENTS.md b/AGENTS.md
index b5377a5c8..750ee049e 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -46,6 +46,8 @@
 
 正確動作是 AI 自動補齊 target selector、source-of-truth diff、check-mode / dry-run、rollback、post-apply verifier、KM / PlayBook trust writeback，然後推進可驗證、可回滾、低爆炸半徑的實作。
 
+**110 runner 壓力事故例外**：Gitea / act-runner 對 110 造成 CPU / headless smoke 壓力時，屬事故級容量保護，不得用「全面授權」直接重開 runner、移除 mask、還原 runner binary 或把 host pressure gate 改成 warn-only。正確動作是先做 runner 搬遷 / 限流 / label isolation / smoke 排程，再以 check-mode、rollback 與 post-apply verifier 受控恢復。
+
 ---
 
 ## 🔴 絕對禁止 → [HARD_RULES.md](docs/HARD_RULES.md)
diff --git a/docs/HARD_RULES.md b/docs/HARD_RULES.md
index 88e3ef243..39580d3c1 100644
--- a/docs/HARD_RULES.md
+++ b/docs/HARD_RULES.md
@@ -287,6 +287,20 @@ OpenClaw 核心替換、仲裁模型升級、SDK / runtime 新依賴正式引入
 force push / 刪 repo / 刪 refs / 改 repo visibility / raw runtime secret volume 讀寫
 ```
 
+### 110 runner 壓力事故例外
+
+2026-06-28 事故後，110 上的 Gitea / act-runner、StockPlatform headless smoke、host-side Next build 與 Docker / BuildKit 壓力屬容量事故保護面。即使收到「批准 / 繼續 / 全面授權」，也不得直接重開 runner、解除 service mask、還原 live runner binary、恢復泛用 `ubuntu-latest` label，或把 host pressure gate 改成 warn-only 作為預設。
+
+允許的 controlled apply 是降壓與防再發：停止 / disable / mask runner、quarantine runner binary、收斂 labels、補 source fail-closed guard、搬遷 runner、限制 concurrency、把 smoke 改成排程 / 非 110 runner，以及執行只讀 pressure / cold-start verifier。
+
+恢復 runner 必須同時具備：
+
+1. target selector：明確列出 service、runner dir、label 與承接 repo。
+2. source-of-truth diff：repo / unit / startup script / runner config 都有一致變更。
+3. 限流或搬遷：不再由 110 production host 承接泛用重型 build / smoke。
+4. rollback：能回到 inactive / masked / fail-closed stub。
+5. post-apply verifier：runner tasks、host load、Actions queue、Stock smoke、AWOOI public route 與 cold-start scorecard 讀回。
+
 ### 跨專案繼承
 
 本節是 AWOOOI / AwoooP / IwoooS / StockPlatform / VibeWork / MOMO / AwoooGo / 2026FIFA / Agent Bounty / Tsenyang 等所有專案的預設行為。若任一 repo 的 `AGENTS.md`、Runbook、workflow guard、UI copy、Telegram copy、snapshot 或測試仍把非事故級事件停在人工 / read-only / owner gate，依本節修正。
diff --git a/docs/LOGBOOK.md b/docs/LOGBOOK.md
index 2a2cc90a2..975a51ea6 100644
--- a/docs/LOGBOOK.md
+++ b/docs/LOGBOOK.md
@@ -1,3 +1,20 @@
+## 2026-06-28 — 08:45 110 runner 壓力事故 source / live fail-closed 收斂
+
+**背景**：統帥全面授權打開非事故級 gate，但 110 Gitea runner 反覆拉起 StockPlatform headless Chrome smoke，已造成 production host CPU / CI 壓力事故；runner 未搬遷 / 限流前不得直接重開。
+
+**完成內容**：
+- Gitea main / branch 已推 `54eea881e fix(ci): fail closed host runner pressure guards [skip ci]`，回復 `scripts/ci/wait-host-web-build-pressure.sh` fail-closed 預設、`scripts/reboot-recovery/awoooi-startup-110.sh` 雙鑰匙 sentinel、`.gitea/workflows/cd.yaml` Docker build lock fail-closed。
+- Live 110 `/usr/local/bin/awoooi-wait-host-web-build-pressure.sh` 與 `/usr/local/bin/awoooi-startup-110.sh` 已同步 fail-closed 並加 immutable。
+- Live 110 四條 runner 入口改為 immutable fail-closed stub，原 ELF 僅 quarantine 不讀內容：`/home/wooo/act-runner/act_runner`、`/home/wooo/act-runner/act_runner.real-20260628-runner-pressure-guard`、`/home/wooo/act-runner-controlled/act_runner`、`/home/wooo/awoooi-controlled-runner/awoooi_controlled_runner`。
+- `gitea-act-runner-host.service`、`gitea-act-runner-awoooi-controlled.service`、`gitea-awoooi-controlled-runner.service`、`gitea-act-runner-awoooi-open.service` 讀回 inactive / disabled，並加上 mask 防止直啟。
+- `AGENTS.md`、`docs/HARD_RULES.md`、`ops/runner/README.md` 與 MASTER runner 章節補上 110 runner 壓力事故例外：全面授權不等於可重開 production host runner。
+
+**Live readback**：
+- 08:41 讀回四個 runner binary 皆為 163-byte shell stub、immutable、執行時回 fail-closed 訊息；runner / StockPlatform / Playwright / Chrome tasks 為 `0`。
+- 08:40 曾抓到 parent=1 的孤兒 `/home/wooo/act-runner/act_runner.real-20260628-runner-pressure-guard daemon --config config.yaml`，已 terminate；後續改為 stub 以防任何 direct exec 繞過 systemd。
+
+**邊界**：本段沒有重啟 Docker / Nginx / firewall / K3s / DB，沒有讀 raw sessions / SQLite / auth / `.env` / runner token。仍未宣稱全主機 scorecard green；下一步是 runner 搬遷 / 限流、StockPlatform smoke 排程或移出 110、再跑 cold-start scorecard。
+
 ## 2026-06-28 — 03:28 Wazuh manager registry acceptance 收件驗證 production 收斂
 
 **完成內容**：
diff --git a/docs/superpowers/specs/2026-04-15-MASTER-ai-autonomous-flywheel-v2.md b/docs/superpowers/specs/2026-04-15-MASTER-ai-autonomous-flywheel-v2.md
index 6420d8d45..0abeff3ff 100644
--- a/docs/superpowers/specs/2026-04-15-MASTER-ai-autonomous-flywheel-v2.md
+++ b/docs/superpowers/specs/2026-04-15-MASTER-ai-autonomous-flywheel-v2.md
@@ -3865,7 +3865,7 @@ Phase 6 完成後
 
 **T132 CD host Web build pressure gate（2026-05-21 台北）**：
 - 觸發：T131 production deploy 期間再次觀察到 110 runner 同時跑 AWOOI Web image build 與 `stockplatform-v2` host-side Next build；既有 workflow concurrency 與 Docker network lock 不會阻擋其他 repo 的 host-side `next build` / `turbo build`。
-- 修正：`.gitea/workflows/cd.yaml` 在 Harbor login 後、Docker build lock 前新增 `Wait for Host Web Build Pressure`；新增 `scripts/ci/wait-host-web-build-pressure.sh` 讀取 `ps` 偵測外部 frontend production build，排除 AWOOI checkout / local worktree / `/app/apps/web` 自身 Web Docker build。預設等待 60 次、每次 10 秒，逾時先 warning 放行；`HOST_WEB_BUILD_PRESSURE_WARN_ONLY=0` 可在 runner 隔離更穩後切 hard fail。`ops/runner/README.md` 補第四層 runner 修復與長期 runner isolation / build offload 方向。
+- 修正：`.gitea/workflows/cd.yaml` 在 Harbor login 後、Docker build lock 前新增 `Wait for Host Web Build Pressure`；新增 `scripts/ci/wait-host-web-build-pressure.sh` 讀取 `ps` 偵測外部 frontend production build，排除 AWOOI checkout / local worktree / `/app/apps/web` 自身 Web Docker build。2026-06-28 事故後已改為預設等待 60 次、每次 10 秒，逾時 hard fail；只有明確設定 `HOST_WEB_BUILD_PRESSURE_WARN_ONLY=1` 的受控補跑才可 warning 放行。`ops/runner/README.md` 補第四層 runner 修復與長期 runner isolation / build offload 方向。
 - Verification：workflow YAML parse ok；`bash -n scripts/ci/wait-host-web-build-pressure.sh` pass；`git diff --check` pass；fixture 驗證 `stockplatform-v2 ... next build` 會被抓到，`/workspace/wooo/awoooi ... next build` 與 `/app/apps/web ... next/dist/bin/next build` 會被排除。`b3ab4da0 ci(cd): wait for host web build pressure` 已推 Gitea main；Code Review #2783 completed/success。受控 CD workflow_dispatch #2784 completed/success：tests 3578 / build-and-deploy 3579 / post-deploy-checks 3580 全綠；真實 log 顯示 gate 抓到 `stockplatform-v2 ... next build`、等待 10s 後才取得 Docker build lock；deploy marker `c44188b chore(cd): deploy 251f5ad [skip ci]`；ArgoCD Synced + Healthy；API/Web/Worker rollout success；post-deploy E2E 5 passed；CI/CD success notification mirrored through AWOOI API。Production readback：health healthy/prod/mock_mode=false；API/Web image `251f5ad6580676fb61094ad4fd44124f179ba0d5`，API 2/2、Web 2/2、Worker 1/1 ready。
 - 判讀：T132 是 shared runner 尚未拆分前的保守等待門，不等於 runner overload 結構性完結；它降低跨 repo frontend build 疊加導致 AWOOI CD / AwoooP 可觀測誤判的機率。下一層若仍看到 110 pressure，應推 runner isolation / build offload。同步暴露 cleanup 候選：`apps/web/Dockerfile` legacy ENV warning 與 act runner cleanup symlink/permission warning。
 - 目前進度更新：AwoooP 告警可觀測鏈約 99.998%；Incident-level source correlation 可見性約 98.8%；Source correlation apply 狀態鏈可驗證性約 99.72%；Source correlation freshness / rolling gate 約 98.2%；前端 AI 自動化管理介面同步約 99.999%；Dashboard snapshot / SSE console noise 收斂約 99.2%；CI/CD runner hygiene 約 97.8%；完整 AI 自動化管理產品化約 99.957%。
diff --git a/ops/runner/README.md b/ops/runner/README.md
index bd1d015bf..bc9773f9e 100644
--- a/ops/runner/README.md
+++ b/ops/runner/README.md
@@ -386,6 +386,19 @@ runner registration / service：
 狀態；AWOOI workflows 也應只使用 `awoooi-ubuntu` 或 `awoooi-host`，不可再使用
 `ubuntu-latest` / `self-hosted` 這類泛用 label。這不代表 runner 搬遷完成，也不代表可以直接重開 runner。
 
+2026-06-28 live update：110 runner 壓力事故確認有直呼
+`/home/wooo/act-runner/act_runner.real-20260628-runner-pressure-guard` 的孤兒
+daemon。四條 live runner 入口已改為 immutable fail-closed stub，原 ELF 僅
+quarantine 不讀內容；相關 systemd units 維持 inactive / disabled / masked：
+
+- `/home/wooo/act-runner/act_runner`
+- `/home/wooo/act-runner/act_runner.real-20260628-runner-pressure-guard`
+- `/home/wooo/act-runner-controlled/act_runner`
+- `/home/wooo/awoooi-controlled-runner/awoooi_controlled_runner`
+
+未完成 runner 搬遷 / 限流 / smoke 排程前，不得解除 mask、還原 ELF、恢復
+泛用 runner label，或把 host pressure gate 預設改成 warn-only。
+
 ---
 版本: v2.0 | 更新: 2026-03-29 | 作者: Claude Code
 變更: v1.0→v2.0 序列建構取代 Job Concurrency Groups
diff --git a/scripts/ci/wait-host-web-build-pressure.sh b/scripts/ci/wait-host-web-build-pressure.sh
index 29e783841..f4dfd7a7b 100755
--- a/scripts/ci/wait-host-web-build-pressure.sh
+++ b/scripts/ci/wait-host-web-build-pressure.sh
@@ -7,7 +7,8 @@ set -euo pipefail
 # production host and a CI host, so CD must not start a new Docker/Next build
 # while load, BuildKit, Gitea Actions, or headless smoke pressure is already high.
 # This gate never kills, renices, or rewrites another repo's process tree.
-# 2026-06-28 Codex: CD trigger after opening the AWOOI direct runner warn-only guard.
+# 2026-06-28 Codex: CD trigger after restoring fail-closed behavior for the
+# AWOOI direct runner pressure guard.
 # 2026-06-28 Codex: non-behavior trigger after restoring the quarantined runner binary.
 # 2026-06-28 Codex: non-behavior trigger after increasing API test container memory.
 # 2026-06-28 Codex: host 110 runner pressure remains an incident-grade guard.

From b17038346345a95405cd16ebf3cb4a62c5b78518 Mon Sep 17 00:00:00 2001
From: Your Name <you@example.com>
Date: Sun, 28 Jun 2026 08:44:50 +0800
Subject: [PATCH 2/6] feat(agents): read live autonomous runtime receipts

---
 apps/api/src/api/v1/agents.py                 |   4 +-
 .../ai_agent_autonomous_runtime_control.py    | 597 ++++++++++++++++++
 ...est_ai_agent_autonomous_runtime_control.py |  99 +++
 ...ai_agent_autonomous_runtime_control_api.py |  26 +-
 4 files changed, 722 insertions(+), 4 deletions(-)

diff --git a/apps/api/src/api/v1/agents.py b/apps/api/src/api/v1/agents.py
index 3d557bb33..917c084b3 100644
--- a/apps/api/src/api/v1/agents.py
+++ b/apps/api/src/api/v1/agents.py
@@ -59,7 +59,7 @@ from src.services.ai_agent_automation_inventory_snapshot import (
     load_latest_ai_agent_automation_inventory_snapshot,
 )
 from src.services.ai_agent_autonomous_runtime_control import (
-    build_ai_agent_autonomous_runtime_control,
+    build_ai_agent_autonomous_runtime_control_with_live_readback,
 )
 from src.services.ai_agent_candidate_operation_dry_run_evidence import (
     load_latest_ai_agent_candidate_operation_dry_run_evidence,
@@ -841,7 +841,7 @@ async def get_automation_inventory_snapshot() -> dict[str, Any]:
 async def get_agent_autonomous_runtime_control() -> dict[str, Any]:
     """回傳目前有效 AI Agent 自主化控制層。"""
     try:
-        return await asyncio.to_thread(build_ai_agent_autonomous_runtime_control)
+        return await build_ai_agent_autonomous_runtime_control_with_live_readback()
     except ValueError as exc:
         logger.error("ai_agent_autonomous_runtime_control_invalid", error=str(exc))
         raise HTTPException(
diff --git a/apps/api/src/services/ai_agent_autonomous_runtime_control.py b/apps/api/src/services/ai_agent_autonomous_runtime_control.py
index 56c24ec8b..46907840e 100644
--- a/apps/api/src/services/ai_agent_autonomous_runtime_control.py
+++ b/apps/api/src/services/ai_agent_autonomous_runtime_control.py
@@ -9,10 +9,14 @@ KM, and Telegram receipts are present.
 
 from __future__ import annotations
 
+from collections.abc import Iterable, Mapping
 from datetime import datetime, timezone
 from typing import Any
 
 from src.core.config import settings
+from src.core.logging import get_logger
+from sqlalchemy import text
+from src.db.base import get_db_context
 from src.services.report_generation_service import (
     DAILY_REPORT_HOUR_TAIPEI,
     MONTHLY_REPORT_DAY_TAIPEI,
@@ -25,6 +29,18 @@ _SCHEMA_VERSION = "ai_agent_autonomous_runtime_control_v1"
 _RUNTIME_AUTHORITY = "current_owner_directive_controlled_ai_automation"
 _DEPLOY_READBACK_MARKER = "p2_416_d1n_autonomous_runtime_control_prod_readback_v2"
 _DEPLOY_ATTEMPT_NOTE = "cd_3673_retry_after_host_pressure_gate_fix"
+_LIVE_READBACK_SCHEMA_VERSION = "ai_agent_autonomous_runtime_receipt_readback_v1"
+_DEFAULT_PROJECT_ID = "awoooi"
+_DEFAULT_LOOKBACK_HOURS = 24
+_EXECUTOR_OPERATION_TYPES = (
+    "ansible_candidate_matched",
+    "ansible_check_mode_executed",
+    "ansible_apply_executed",
+    "ansible_rollback_executed",
+    "ansible_execution_skipped",
+)
+
+logger = get_logger(__name__)
 
 
 def _allowed_risk_levels() -> list[str]:
@@ -32,6 +48,340 @@ def _allowed_risk_levels() -> list[str]:
     return sorted({item.strip().lower() for item in raw.split(",") if item.strip()})
 
 
+def _utc_iso(value: Any) -> str | None:
+    if value is None:
+        return None
+    if isinstance(value, datetime):
+        if value.tzinfo is None:
+            value = value.replace(tzinfo=timezone.utc)
+        return value.astimezone(timezone.utc).isoformat()
+    return str(value)
+
+
+def _row_mapping(row: Mapping[str, Any] | Any) -> dict[str, Any]:
+    if isinstance(row, Mapping):
+        return dict(row)
+    mapping = getattr(row, "_mapping", None)
+    if mapping is not None:
+        return dict(mapping)
+    return dict(row)
+
+
+def _int_value(value: Any) -> int:
+    try:
+        return int(value or 0)
+    except (TypeError, ValueError):
+        return 0
+
+
+def _sanitize_latest_rows(
+    rows: Iterable[Mapping[str, Any] | Any],
+    *,
+    allowed_keys: tuple[str, ...],
+    time_keys: tuple[str, ...] = ("created_at", "collected_at", "queued_at", "sent_at"),
+    limit: int = 5,
+) -> list[dict[str, Any]]:
+    clean_rows: list[dict[str, Any]] = []
+    for row in rows:
+        item = _row_mapping(row)
+        clean: dict[str, Any] = {}
+        for key in allowed_keys:
+            if key not in item:
+                continue
+            value = item.get(key)
+            clean[key] = _utc_iso(value) if key in time_keys else value
+        clean_rows.append(clean)
+        if len(clean_rows) >= limit:
+            break
+    return clean_rows
+
+
+def _operation_counts(
+    rows: Iterable[Mapping[str, Any] | Any],
+) -> dict[str, dict[str, Any]]:
+    counts = {
+        operation_type: {
+            "total": 0,
+            "recent": 0,
+            "by_status": {},
+        }
+        for operation_type in _EXECUTOR_OPERATION_TYPES
+    }
+    for row in rows:
+        item = _row_mapping(row)
+        operation_type = str(item.get("operation_type") or "unknown")
+        status = str(item.get("status") or "unknown")
+        bucket = counts.setdefault(
+            operation_type,
+            {
+                "total": 0,
+                "recent": 0,
+                "by_status": {},
+            },
+        )
+        total = _int_value(item.get("total"))
+        recent = _int_value(item.get("recent"))
+        bucket["total"] += total
+        bucket["recent"] += recent
+        bucket["by_status"][status] = bucket["by_status"].get(status, 0) + total
+    return counts
+
+
+def _status_counts(
+    rows: Iterable[Mapping[str, Any] | Any],
+    *,
+    status_key: str,
+) -> dict[str, Any]:
+    by_status: dict[str, int] = {}
+    total = 0
+    recent = 0
+    for row in rows:
+        item = _row_mapping(row)
+        status = str(item.get(status_key) or "unknown")
+        row_total = _int_value(item.get("total"))
+        by_status[status] = by_status.get(status, 0) + row_total
+        total += row_total
+        recent += _int_value(item.get("recent"))
+    return {
+        "total": total,
+        "recent": recent,
+        "by_status": by_status,
+    }
+
+
+def _latest_flow_closure(
+    *,
+    operation_latest_rows: Iterable[Mapping[str, Any] | Any],
+    verifier_latest_rows: Iterable[Mapping[str, Any] | Any],
+    km_latest_rows: Iterable[Mapping[str, Any] | Any],
+    telegram_latest_rows: Iterable[Mapping[str, Any] | Any],
+) -> dict[str, Any]:
+    operation_rows = [_row_mapping(row) for row in operation_latest_rows]
+    verifier_rows = [_row_mapping(row) for row in verifier_latest_rows]
+    km_rows = [_row_mapping(row) for row in km_latest_rows]
+    telegram_rows = [_row_mapping(row) for row in telegram_latest_rows]
+    latest_apply = next(
+        (
+            row
+            for row in operation_rows
+            if str(row.get("operation_type") or "") == "ansible_apply_executed"
+        ),
+        None,
+    )
+    if latest_apply is None:
+        return {
+            "apply_op_id": None,
+            "incident_id": None,
+            "has_post_apply_verifier": False,
+            "has_km_writeback": False,
+            "has_telegram_receipt": False,
+            "closed": False,
+            "missing": [
+                "ansible_apply_executed",
+                "post_apply_verifier",
+                "km_writeback",
+                "telegram_receipt",
+            ],
+        }
+
+    apply_op_id = str(latest_apply.get("op_id") or "")
+    incident_id = str(latest_apply.get("incident_id") or "")
+    km_path_type = f"ansible_apply_receipt:{apply_op_id[:8]}" if apply_op_id else ""
+    has_verifier = any(
+        str(row.get("apply_op_id") or "") == apply_op_id
+        for row in verifier_rows
+    )
+    has_km = any(
+        str(row.get("path_type") or "") == km_path_type
+        or (
+            incident_id
+            and str(row.get("related_incident_id") or "") == incident_id
+        )
+        for row in km_rows
+    )
+    has_telegram = any(
+        str(row.get("send_status") or "") == "sent"
+        and str(row.get("action") or "") == "controlled_apply_result"
+        and (
+            not incident_id
+            or str(row.get("incident_id") or "") == incident_id
+        )
+        for row in telegram_rows
+    )
+    missing = [
+        name
+        for name, present in (
+            ("post_apply_verifier", has_verifier),
+            ("km_writeback", has_km),
+            ("telegram_receipt", has_telegram),
+        )
+        if not present
+    ]
+    return {
+        "apply_op_id": apply_op_id or None,
+        "incident_id": incident_id or None,
+        "has_post_apply_verifier": has_verifier,
+        "has_km_writeback": has_km,
+        "has_telegram_receipt": has_telegram,
+        "closed": not missing,
+        "missing": missing,
+    }
+
+
+def build_runtime_receipt_readback_from_rows(
+    *,
+    project_id: str = _DEFAULT_PROJECT_ID,
+    lookback_hours: int = _DEFAULT_LOOKBACK_HOURS,
+    db_read_status: str = "ok",
+    operation_count_rows: Iterable[Mapping[str, Any] | Any] = (),
+    operation_latest_rows: Iterable[Mapping[str, Any] | Any] = (),
+    verifier_count_rows: Iterable[Mapping[str, Any] | Any] = (),
+    verifier_latest_rows: Iterable[Mapping[str, Any] | Any] = (),
+    km_count_rows: Iterable[Mapping[str, Any] | Any] = (),
+    km_latest_rows: Iterable[Mapping[str, Any] | Any] = (),
+    telegram_count_rows: Iterable[Mapping[str, Any] | Any] = (),
+    telegram_latest_rows: Iterable[Mapping[str, Any] | Any] = (),
+    error_type: str | None = None,
+) -> dict[str, Any]:
+    """Build the live executor receipt readback from already-fetched rows."""
+
+    operation_latest = list(operation_latest_rows)
+    verifier_latest = list(verifier_latest_rows)
+    km_latest = list(km_latest_rows)
+    telegram_latest = list(telegram_latest_rows)
+    operation_summary = _operation_counts(operation_count_rows)
+    verifier_summary = _status_counts(
+        verifier_count_rows,
+        status_key="verification_result",
+    )
+    km_summary = _status_counts(km_count_rows, status_key="status")
+    telegram_summary = _status_counts(telegram_count_rows, status_key="send_status")
+    latest_closure = _latest_flow_closure(
+        operation_latest_rows=operation_latest,
+        verifier_latest_rows=verifier_latest,
+        km_latest_rows=km_latest,
+        telegram_latest_rows=telegram_latest,
+    )
+    apply_summary = operation_summary.get("ansible_apply_executed") or {}
+    readback = {
+        "schema_version": _LIVE_READBACK_SCHEMA_VERSION,
+        "project_id": project_id,
+        "lookback_hours": max(1, int(lookback_hours or _DEFAULT_LOOKBACK_HOURS)),
+        "db_read_status": db_read_status,
+        "writes_on_read": False,
+        "ansible_operations": {
+            "counts": operation_summary,
+            "latest": _sanitize_latest_rows(
+                operation_latest,
+                allowed_keys=(
+                    "op_id",
+                    "parent_op_id",
+                    "operation_type",
+                    "status",
+                    "actor",
+                    "incident_id",
+                    "catalog_id",
+                    "playbook_path",
+                    "execution_mode",
+                    "returncode",
+                    "duration_ms",
+                    "created_at",
+                ),
+            ),
+        },
+        "ansible_apply_executed": {
+            "total": _int_value(apply_summary.get("total")),
+            "recent": _int_value(apply_summary.get("recent")),
+            "by_status": apply_summary.get("by_status") or {},
+        },
+        "post_apply_verifier": {
+            **verifier_summary,
+            "latest": _sanitize_latest_rows(
+                verifier_latest,
+                allowed_keys=(
+                    "id",
+                    "incident_id",
+                    "matched_playbook_id",
+                    "verification_result",
+                    "apply_op_id",
+                    "catalog_id",
+                    "playbook_path",
+                    "returncode",
+                    "collected_at",
+                ),
+            ),
+        },
+        "km_writeback": {
+            **km_summary,
+            "latest": _sanitize_latest_rows(
+                km_latest,
+                allowed_keys=(
+                    "id",
+                    "title",
+                    "related_incident_id",
+                    "related_playbook_id",
+                    "path_type",
+                    "status",
+                    "created_by",
+                    "created_at",
+                ),
+            ),
+        },
+        "telegram_receipt": {
+            **telegram_summary,
+            "latest": _sanitize_latest_rows(
+                telegram_latest,
+                allowed_keys=(
+                    "message_id",
+                    "run_id",
+                    "message_type",
+                    "send_status",
+                    "provider_message_id",
+                    "incident_id",
+                    "action",
+                    "queued_at",
+                    "sent_at",
+                ),
+            ),
+        },
+        "latest_flow_closure": latest_closure,
+    }
+    if error_type:
+        readback["error"] = {
+            "type": error_type,
+            "message": "runtime receipt DB read failed; see API logs",
+        }
+    return readback
+
+
+def _attach_runtime_receipt_readback(
+    payload: dict[str, Any],
+    readback: dict[str, Any],
+) -> dict[str, Any]:
+    payload["runtime_receipt_readback"] = readback
+    rollups = payload.setdefault("rollups", {})
+    rollups.update({
+        "live_ansible_apply_executed_count": _int_value(
+            readback.get("ansible_apply_executed", {}).get("total")
+        ),
+        "live_post_apply_verifier_count": _int_value(
+            readback.get("post_apply_verifier", {}).get("total")
+        ),
+        "live_km_writeback_count": _int_value(
+            readback.get("km_writeback", {}).get("total")
+        ),
+        "live_telegram_receipt_count": _int_value(
+            readback.get("telegram_receipt", {}).get("total")
+        ),
+        "live_executor_latest_flow_closed_count": (
+            1
+            if (readback.get("latest_flow_closure") or {}).get("closed") is True
+            else 0
+        ),
+    })
+    return payload
+
+
 def build_ai_agent_autonomous_runtime_control() -> dict[str, Any]:
     """Build the current AI Agent autonomy control-plane readback."""
 
@@ -240,10 +590,257 @@ def build_ai_agent_autonomous_runtime_control() -> dict[str, Any]:
             "legacy_policy_overridden_count": len(legacy_overrides),
         },
     }
+    _attach_runtime_receipt_readback(
+        payload,
+        build_runtime_receipt_readback_from_rows(
+            project_id=_DEFAULT_PROJECT_ID,
+            db_read_status="not_queried",
+        ),
+    )
     _validate_payload(payload)
     return payload
 
 
+async def load_ai_agent_autonomous_runtime_receipt_readback(
+    *,
+    project_id: str = _DEFAULT_PROJECT_ID,
+    lookback_hours: int = _DEFAULT_LOOKBACK_HOURS,
+    limit: int = 20,
+) -> dict[str, Any]:
+    """Read live executor receipts without sending messages or mutating runtime state."""
+
+    params = {
+        "project_id": project_id,
+        "lookback_hours": max(1, int(lookback_hours or _DEFAULT_LOOKBACK_HOURS)),
+        "limit": max(1, int(limit or 20)),
+    }
+    try:
+        async with get_db_context(project_id) as db:
+            await db.execute(text("SET LOCAL statement_timeout = '5000ms'"))
+            operation_counts = (
+                await db.execute(text(_RUNTIME_OPERATION_COUNTS_SQL), params)
+            ).mappings().all()
+            operation_latest = (
+                await db.execute(text(_RUNTIME_OPERATION_LATEST_SQL), params)
+            ).mappings().all()
+            verifier_counts = (
+                await db.execute(text(_RUNTIME_VERIFIER_COUNTS_SQL), params)
+            ).mappings().all()
+            verifier_latest = (
+                await db.execute(text(_RUNTIME_VERIFIER_LATEST_SQL), params)
+            ).mappings().all()
+            km_counts = (
+                await db.execute(text(_RUNTIME_KM_COUNTS_SQL), params)
+            ).mappings().all()
+            km_latest = (
+                await db.execute(text(_RUNTIME_KM_LATEST_SQL), params)
+            ).mappings().all()
+            telegram_counts = (
+                await db.execute(text(_RUNTIME_TELEGRAM_COUNTS_SQL), params)
+            ).mappings().all()
+            telegram_latest = (
+                await db.execute(text(_RUNTIME_TELEGRAM_LATEST_SQL), params)
+            ).mappings().all()
+    except Exception as exc:
+        logger.warning(
+            "ai_agent_autonomous_runtime_receipt_readback_failed",
+            project_id=project_id,
+            error_type=type(exc).__name__,
+        )
+        return build_runtime_receipt_readback_from_rows(
+            project_id=project_id,
+            lookback_hours=params["lookback_hours"],
+            db_read_status="unavailable",
+            error_type=type(exc).__name__,
+        )
+
+    return build_runtime_receipt_readback_from_rows(
+        project_id=project_id,
+        lookback_hours=params["lookback_hours"],
+        db_read_status="ok",
+        operation_count_rows=operation_counts,
+        operation_latest_rows=operation_latest,
+        verifier_count_rows=verifier_counts,
+        verifier_latest_rows=verifier_latest,
+        km_count_rows=km_counts,
+        km_latest_rows=km_latest,
+        telegram_count_rows=telegram_counts,
+        telegram_latest_rows=telegram_latest,
+    )
+
+
+async def build_ai_agent_autonomous_runtime_control_with_live_readback(
+    *,
+    project_id: str = _DEFAULT_PROJECT_ID,
+    lookback_hours: int = _DEFAULT_LOOKBACK_HOURS,
+) -> dict[str, Any]:
+    """Build the control plane and attach live DB receipt readback."""
+
+    payload = build_ai_agent_autonomous_runtime_control()
+    readback = await load_ai_agent_autonomous_runtime_receipt_readback(
+        project_id=project_id,
+        lookback_hours=lookback_hours,
+    )
+    _attach_runtime_receipt_readback(payload, readback)
+    _validate_payload(payload)
+    return payload
+
+
+_RUNTIME_OPERATION_COUNTS_SQL = """
+    SELECT
+        operation_type,
+        status,
+        count(*) AS total,
+        count(*) FILTER (
+            WHERE created_at >= NOW() - (:lookback_hours * INTERVAL '1 hour')
+        ) AS recent
+    FROM automation_operation_log
+    WHERE operation_type IN (
+        'ansible_candidate_matched',
+        'ansible_check_mode_executed',
+        'ansible_apply_executed',
+        'ansible_rollback_executed',
+        'ansible_execution_skipped'
+    )
+    GROUP BY operation_type, status
+    ORDER BY operation_type, status
+"""
+
+
+_RUNTIME_OPERATION_LATEST_SQL = """
+    SELECT
+        op_id::text AS op_id,
+        parent_op_id::text AS parent_op_id,
+        operation_type,
+        status,
+        actor,
+        coalesce(incident_id::text, input ->> 'incident_id') AS incident_id,
+        input ->> 'catalog_id' AS catalog_id,
+        coalesce(input ->> 'apply_playbook_path', input ->> 'playbook_path') AS playbook_path,
+        input ->> 'execution_mode' AS execution_mode,
+        coalesce(output ->> 'returncode', dry_run_result ->> 'returncode') AS returncode,
+        duration_ms,
+        created_at
+    FROM automation_operation_log
+    WHERE operation_type IN (
+        'ansible_candidate_matched',
+        'ansible_check_mode_executed',
+        'ansible_apply_executed',
+        'ansible_rollback_executed',
+        'ansible_execution_skipped'
+    )
+    ORDER BY created_at DESC
+    LIMIT :limit
+"""
+
+
+_RUNTIME_VERIFIER_COUNTS_SQL = """
+    SELECT
+        coalesce(verification_result, 'missing') AS verification_result,
+        count(*) AS total,
+        count(*) FILTER (
+            WHERE collected_at >= NOW() - (:lookback_hours * INTERVAL '1 hour')
+        ) AS recent
+    FROM incident_evidence
+    WHERE post_execution_state ->> 'apply_op_id' IS NOT NULL
+    GROUP BY coalesce(verification_result, 'missing')
+    ORDER BY verification_result
+"""
+
+
+_RUNTIME_VERIFIER_LATEST_SQL = """
+    SELECT
+        id,
+        incident_id,
+        matched_playbook_id,
+        coalesce(verification_result, 'missing') AS verification_result,
+        post_execution_state ->> 'apply_op_id' AS apply_op_id,
+        post_execution_state ->> 'catalog_id' AS catalog_id,
+        post_execution_state ->> 'playbook_path' AS playbook_path,
+        post_execution_state ->> 'returncode' AS returncode,
+        collected_at
+    FROM incident_evidence
+    WHERE post_execution_state ->> 'apply_op_id' IS NOT NULL
+    ORDER BY collected_at DESC
+    LIMIT :limit
+"""
+
+
+_RUNTIME_KM_COUNTS_SQL = """
+    SELECT
+        status,
+        count(*) AS total,
+        count(*) FILTER (
+            WHERE created_at >= NOW() - (:lookback_hours * INTERVAL '1 hour')
+        ) AS recent
+    FROM knowledge_entries
+    WHERE project_id = :project_id
+      AND (
+        path_type LIKE 'ansible_apply_receipt:%'
+        OR tags::text LIKE '%ansible_controlled_apply%'
+      )
+    GROUP BY status
+    ORDER BY status
+"""
+
+
+_RUNTIME_KM_LATEST_SQL = """
+    SELECT
+        id,
+        title,
+        related_incident_id,
+        related_playbook_id,
+        path_type,
+        status,
+        created_by,
+        created_at
+    FROM knowledge_entries
+    WHERE project_id = :project_id
+      AND (
+        path_type LIKE 'ansible_apply_receipt:%'
+        OR tags::text LIKE '%ansible_controlled_apply%'
+      )
+    ORDER BY created_at DESC
+    LIMIT :limit
+"""
+
+
+_RUNTIME_TELEGRAM_COUNTS_SQL = """
+    SELECT
+        send_status,
+        count(*) AS total,
+        count(*) FILTER (
+            WHERE queued_at >= NOW() - (:lookback_hours * INTERVAL '1 hour')
+        ) AS recent
+    FROM awooop_outbound_message
+    WHERE project_id = :project_id
+      AND channel_type = 'telegram'
+      AND source_envelope #>> '{callback_reply,action}' = 'controlled_apply_result'
+    GROUP BY send_status
+    ORDER BY send_status
+"""
+
+
+_RUNTIME_TELEGRAM_LATEST_SQL = """
+    SELECT
+        message_id::text AS message_id,
+        run_id::text AS run_id,
+        message_type,
+        send_status,
+        provider_message_id,
+        source_envelope #>> '{callback_reply,incident_id}' AS incident_id,
+        source_envelope #>> '{callback_reply,action}' AS action,
+        queued_at,
+        sent_at
+    FROM awooop_outbound_message
+    WHERE project_id = :project_id
+      AND channel_type = 'telegram'
+      AND source_envelope #>> '{callback_reply,action}' = 'controlled_apply_result'
+    ORDER BY queued_at DESC
+    LIMIT :limit
+"""
+
+
 def _validate_payload(payload: dict[str, Any]) -> None:
     if payload.get("schema_version") != _SCHEMA_VERSION:
         raise ValueError(f"schema_version must be {_SCHEMA_VERSION}")
diff --git a/apps/api/tests/test_ai_agent_autonomous_runtime_control.py b/apps/api/tests/test_ai_agent_autonomous_runtime_control.py
index ca6e679f1..edee118fd 100644
--- a/apps/api/tests/test_ai_agent_autonomous_runtime_control.py
+++ b/apps/api/tests/test_ai_agent_autonomous_runtime_control.py
@@ -1,5 +1,6 @@
 from src.services.ai_agent_autonomous_runtime_control import (
     build_ai_agent_autonomous_runtime_control,
+    build_runtime_receipt_readback_from_rows,
 )
 
 
@@ -54,6 +55,7 @@ def test_ai_agent_autonomous_runtime_control_exposes_reports_and_executor_receip
     assert data["rollups"]["report_cadence_enabled_count"] == 3
     assert data["rollups"]["direct_bot_api_allowed_count"] == 0
     assert data["rollups"]["legacy_policy_overridden_count"] >= 4
+    assert data["runtime_receipt_readback"]["db_read_status"] == "not_queried"
 
 
 def test_ai_agent_autonomous_runtime_control_keeps_hard_blockers_and_redaction():
@@ -68,3 +70,100 @@ def test_ai_agent_autonomous_runtime_control_keeps_hard_blockers_and_redaction()
     assert visibility["internal_reasoning_display_allowed"] is False
     assert visibility["sensitive_value_display_allowed"] is False
     assert visibility["telegram_unredacted_payload_display_allowed"] is False
+
+
+def test_runtime_receipt_readback_summarizes_live_executor_closure_rows():
+    apply_op_id = "73b7a95c-3652-4c0d-bb4c-729e500acedb"
+    incident_id = "INC-20260627-64472B"
+
+    readback = build_runtime_receipt_readback_from_rows(
+        project_id="awoooi",
+        db_read_status="ok",
+        operation_count_rows=[
+            {
+                "operation_type": "ansible_apply_executed",
+                "status": "success",
+                "total": 1,
+                "recent": 1,
+            },
+            {
+                "operation_type": "ansible_check_mode_executed",
+                "status": "success",
+                "total": 1,
+                "recent": 1,
+            },
+        ],
+        operation_latest_rows=[
+            {
+                "op_id": apply_op_id,
+                "parent_op_id": "check-mode-op",
+                "operation_type": "ansible_apply_executed",
+                "status": "success",
+                "actor": "ansible_controlled_apply_worker",
+                "incident_id": incident_id,
+                "catalog_id": "ansible:188-momo-backup-user",
+                "playbook_path": "infra/ansible/playbooks/188-momo-backup-user.yml",
+                "execution_mode": "controlled_apply",
+                "returncode": "0",
+                "duration_ms": 7727,
+            },
+        ],
+        verifier_count_rows=[
+            {"verification_result": "success", "total": 1, "recent": 1},
+        ],
+        verifier_latest_rows=[
+            {
+                "id": "evidence-1",
+                "incident_id": incident_id,
+                "verification_result": "success",
+                "apply_op_id": apply_op_id,
+                "catalog_id": "ansible:188-momo-backup-user",
+                "playbook_path": "infra/ansible/playbooks/188-momo-backup-user.yml",
+                "returncode": "0",
+            },
+        ],
+        km_count_rows=[
+            {"status": "review", "total": 1, "recent": 1},
+        ],
+        km_latest_rows=[
+            {
+                "id": "km-1",
+                "title": "AI 自動修復沉澱：INC-20260627-64472B",
+                "related_incident_id": incident_id,
+                "related_playbook_id": "ansible:188-momo-backup-user",
+                "path_type": "ansible_apply_receipt:73b7a95c",
+                "status": "review",
+                "created_by": "ai_agent_ansible_worker",
+            },
+        ],
+        telegram_count_rows=[
+            {"send_status": "sent", "total": 1, "recent": 1},
+        ],
+        telegram_latest_rows=[
+            {
+                "message_id": "telegram-row-1",
+                "run_id": "telegram-run-1",
+                "message_type": "final",
+                "send_status": "sent",
+                "provider_message_id": "12345",
+                "incident_id": incident_id,
+                "action": "controlled_apply_result",
+            },
+        ],
+    )
+
+    assert readback["db_read_status"] == "ok"
+    assert readback["writes_on_read"] is False
+    assert readback["ansible_apply_executed"]["total"] == 1
+    assert readback["post_apply_verifier"]["by_status"]["success"] == 1
+    assert readback["km_writeback"]["by_status"]["review"] == 1
+    assert readback["telegram_receipt"]["by_status"]["sent"] == 1
+    assert readback["latest_flow_closure"] == {
+        "apply_op_id": apply_op_id,
+        "incident_id": incident_id,
+        "has_post_apply_verifier": True,
+        "has_km_writeback": True,
+        "has_telegram_receipt": True,
+        "closed": True,
+        "missing": [],
+    }
diff --git a/apps/api/tests/test_ai_agent_autonomous_runtime_control_api.py b/apps/api/tests/test_ai_agent_autonomous_runtime_control_api.py
index 35d34342a..b5c3d2324 100644
--- a/apps/api/tests/test_ai_agent_autonomous_runtime_control_api.py
+++ b/apps/api/tests/test_ai_agent_autonomous_runtime_control_api.py
@@ -1,6 +1,10 @@
 from fastapi.testclient import TestClient
 
+from src.api.v1 import agents as agents_api
 from src.main import app
+from src.services.ai_agent_autonomous_runtime_control import (
+    build_ai_agent_autonomous_runtime_control,
+)
 
 
 _PUBLIC_FORBIDDEN_TERMS = [
@@ -41,7 +45,16 @@ def _collect_strings(value):
     return []
 
 
-def test_get_ai_agent_autonomous_runtime_control_api():
+async def _fake_runtime_control_with_live_readback():
+    return build_ai_agent_autonomous_runtime_control()
+
+
+def test_get_ai_agent_autonomous_runtime_control_api(monkeypatch):
+    monkeypatch.setattr(
+        agents_api,
+        "build_ai_agent_autonomous_runtime_control_with_live_readback",
+        _fake_runtime_control_with_live_readback,
+    )
     client = TestClient(app)
     response = client.get("/api/v1/agents/agent-autonomous-runtime-control")
 
@@ -60,9 +73,18 @@ def test_get_ai_agent_autonomous_runtime_control_api():
     assert data["current_policy"]["owner_review_required_for_low_medium_high"] is False
     assert data["report_delivery"]["status"] == "telegram_gateway_delivery_enabled"
     assert data["rollups"]["report_cadence_enabled_count"] == 3
+    assert data["runtime_receipt_readback"]["schema_version"] == (
+        "ai_agent_autonomous_runtime_receipt_readback_v1"
+    )
+    assert data["runtime_receipt_readback"]["db_read_status"] == "not_queried"
 
 
-def test_get_ai_agent_autonomous_runtime_control_api_redacts_public_terms():
+def test_get_ai_agent_autonomous_runtime_control_api_redacts_public_terms(monkeypatch):
+    monkeypatch.setattr(
+        agents_api,
+        "build_ai_agent_autonomous_runtime_control_with_live_readback",
+        _fake_runtime_control_with_live_readback,
+    )
     client = TestClient(app)
     response = client.get("/api/v1/agents/agent-autonomous-runtime-control")
 

From d4c2cc6e200fc00e07d179ebb9a4a156cef2c6d5 Mon Sep 17 00:00:00 2001
From: Your Name <you@example.com>
Date: Sun, 28 Jun 2026 08:44:53 +0800
Subject: [PATCH 3/6] feat(iwooos): commit wazuh manager registry accepted
 readback [skip ci]

---
 .../iwooos_runtime_security_readback.py       | 23 +++++++--
 .../iwooos_security_control_coverage.py       | 21 ++++++--
 .../iwooos_wazuh_managed_host_coverage.py     | 32 +++++++++----
 ...uh_manager_registry_reviewer_validation.py | 22 +++++++--
 .../test_iwooos_runtime_security_readback.py  |  4 +-
 .../test_iwooos_security_control_coverage.py  |  7 ++-
 ...test_iwooos_wazuh_managed_host_coverage.py | 27 ++++++-----
 ...uh_manager_registry_reviewer_validation.py | 30 ++++++------
 apps/web/messages/en.json                     | 26 +++++-----
 apps/web/messages/zh-TW.json                  | 26 +++++-----
 apps/web/src/app/[locale]/iwooos/page.tsx     | 31 ++++++------
 ...ty-incident-convergence-gate.snapshot.json |  2 +-
 ...h-managed-host-coverage-gate.snapshot.json | 48 +++++++++----------
 ...registry-reviewer-validation.snapshot.json | 20 ++++----
 14 files changed, 191 insertions(+), 128 deletions(-)

diff --git a/apps/api/src/services/iwooos_runtime_security_readback.py b/apps/api/src/services/iwooos_runtime_security_readback.py
index 4790bf562..23f9f30c4 100644
--- a/apps/api/src/services/iwooos_runtime_security_readback.py
+++ b/apps/api/src/services/iwooos_runtime_security_readback.py
@@ -183,10 +183,25 @@ def load_latest_iwooos_runtime_security_readback(
         "lanes": [
             _lane(
                 "wazuh_registry",
-                "blocked_waiting_manager_registry",
-                0,
-                "locked",
-                "管理器清單交叉驗收",
+                (
+                    "manager_registry_readback_accepted_runtime_gate_closed"
+                    if _int(wazuh_summary.get("manager_registry_accepted_count"))
+                    >= _int(wazuh_summary.get("expected_host_scope_count"))
+                    else "blocked_waiting_manager_registry"
+                ),
+                (
+                    35
+                    if _int(wazuh_summary.get("manager_registry_accepted_count"))
+                    >= _int(wazuh_summary.get("expected_host_scope_count"))
+                    else 0
+                ),
+                (
+                    "steady"
+                    if _int(wazuh_summary.get("manager_registry_accepted_count"))
+                    >= _int(wazuh_summary.get("expected_host_scope_count"))
+                    else "locked"
+                ),
+                "管理器清單交叉驗收已讀回；runtime gate 仍關閉",
                 {
                     "expected_hosts": wazuh_summary.get("expected_host_scope_count", 0),
                     "transport_observed": wazuh_summary.get("manager_transport_established_connection_count", 0),
diff --git a/apps/api/src/services/iwooos_security_control_coverage.py b/apps/api/src/services/iwooos_security_control_coverage.py
index ee9957d29..9bee0d25e 100644
--- a/apps/api/src/services/iwooos_security_control_coverage.py
+++ b/apps/api/src/services/iwooos_security_control_coverage.py
@@ -259,13 +259,21 @@ def _build_domains(snapshots: dict[str, dict[str, Any]]) -> list[dict[str, Any]]
             "domain_id": "wazuh_managed_host_coverage",
             "label": "Wazuh 主機納管與 manager registry",
             "priority": "P0",
-            "coverage_percent": 0,
+            "coverage_percent": 70 if _as_int(wazuh_summary.get("manager_registry_accepted_count")) else 0,
             "scope_count": _as_int(wazuh_summary.get("expected_host_scope_count")),
             "write_capable_count": _as_int(wazuh_summary.get("agent_reenroll_authorized_count")),
             "accepted_count": _as_int(wazuh_summary.get("manager_registry_accepted_count")),
-            "blocked_count": _as_int(wazuh_summary.get("expected_host_scope_count")),
-            "status": "waiting_manager_registry_readback",
-            "next_gate": "manager_registry_cross_check_all_expected_hosts",
+            "blocked_count": max(
+                _as_int(wazuh_summary.get("expected_host_scope_count"))
+                - _as_int(wazuh_summary.get("manager_registry_accepted_count")),
+                0,
+            ),
+            "status": (
+                "manager_registry_readback_accepted_runtime_gate_closed"
+                if _as_int(wazuh_summary.get("manager_registry_accepted_count"))
+                else "waiting_manager_registry_readback"
+            ),
+            "next_gate": "runtime_gate_owner_review_and_postcheck",
             "source_refs": ["docs/security/wazuh-managed-host-coverage-gate.snapshot.json"],
         },
         {
@@ -323,7 +331,10 @@ def _build_summary(snapshots: dict[str, dict[str, Any]], domains: list[dict[str,
         "owner_response_received_count": 0,
         "owner_response_accepted_count": 0,
         "live_evidence_accepted_count": 0,
-        "wazuh_manager_registry_accepted_count": 0,
+        "wazuh_manager_registry_accepted_count": _summary_int(
+            snapshots["wazuh_hosts"],
+            "manager_registry_accepted_count",
+        ),
         "active_scan_authorized_count": 0,
         "active_response_authorized_count": 0,
         "telegram_send_authorized_count": 0,
diff --git a/apps/api/src/services/iwooos_wazuh_managed_host_coverage.py b/apps/api/src/services/iwooos_wazuh_managed_host_coverage.py
index 838088c56..2b6d0fdfb 100644
--- a/apps/api/src/services/iwooos_wazuh_managed_host_coverage.py
+++ b/apps/api/src/services/iwooos_wazuh_managed_host_coverage.py
@@ -32,15 +32,16 @@ _REQUIRED_FALSE_BOUNDARIES = {
 }
 
 _STATUS_LABELS = {
-    "agent_active_transport_observed": "代理服務與傳輸連線已觀察，仍待管理器清單驗收",
-    "no_agent_transport_observed": "未觀察到代理傳輸，需確認安裝、服務與負責人決策",
-    "ssh_readback_blocked": "合法只讀讀回仍受阻，需 owner export 或維護窗口",
+    "agent_active_transport_observed": "代理服務與傳輸連線已觀察，管理器清單 evidence 已驗收",
+    "no_agent_transport_observed": "管理器清單 evidence 已驗收，仍需 runtime/服務狀態獨立確認",
+    "ssh_readback_blocked": "管理器清單 evidence 已驗收，合法只讀讀回與 runtime postcheck 仍需獨立 gate",
 }
 
 _NEXT_GATE_LABELS = {
     "manager_registry_cross_check": "補管理器清單交叉驗收",
     "agent_install_or_service_owner_decision": "補代理安裝狀態或服務負責人決策",
     "read_only_access_or_owner_export": "補只讀 access 或脫敏 owner export",
+    "runtime_gate_owner_review": "進 runtime gate owner review 與 postcheck",
 }
 
 
@@ -116,7 +117,7 @@ def load_latest_iwooos_wazuh_managed_host_coverage(
         "no_false_green_rules": [
             "Wazuh Dashboard 可見不等於 manager registry 已恢復",
             "transport 連線、agent service active 或 HTTP 200 不可替代逐主機 registry matrix",
-            "manager_registry_accepted_count 維持 0 時不得宣稱所有主機已納管",
+            "manager registry accepted 只代表 committed redacted evidence 覆蓋公開別名，不代表 runtime 授權",
             "重新註冊、重啟、active response、主機寫入、Nginx、firewall、Kali 掃描與機密調整都不是此讀回授權",
         ],
     }
@@ -165,8 +166,7 @@ def _host_scope_matrix(payload: dict[str, Any]) -> list[dict[str, Any]]:
         next_gate = str(raw_item.get("next_gate", ""))
         if not node_id.startswith("managed_"):
             raise ValueError(f"Wazuh 受管主機覆蓋 node_id 必須維持公開別名：{node_id}")
-        if raw_item.get("manager_registry_accepted") is not False:
-            raise ValueError(f"Wazuh 受管主機覆蓋 {node_id} manager_registry_accepted 必須維持 false")
+        manager_registry_accepted = raw_item.get("manager_registry_accepted") is True
         matrix.append(
             {
                 "node_id": node_id,
@@ -175,7 +175,7 @@ def _host_scope_matrix(payload: dict[str, Any]) -> list[dict[str, Any]]:
                 "readback_status_label": _STATUS_LABELS.get(readback_status, "待補只讀讀回"),
                 "next_gate": next_gate,
                 "next_gate_label": _NEXT_GATE_LABELS.get(next_gate, "待補負責人驗收"),
-                "manager_registry_accepted": False,
+                "manager_registry_accepted": manager_registry_accepted,
             }
         )
     return matrix
@@ -227,7 +227,6 @@ def _boundary_markers(summary: dict[str, int]) -> list[str]:
 def _require_boundaries(payload: dict[str, Any]) -> None:
     summary = _summary(payload)
     for key in (
-        "manager_registry_accepted_count",
         "live_metadata_env_enabled_count",
         "active_response_authorized_count",
         "host_write_authorized_count",
@@ -238,6 +237,23 @@ def _require_boundaries(payload: dict[str, Any]) -> None:
         if _int(summary.get(key)) != 0:
             raise ValueError(f"Wazuh 受管主機覆蓋 summary.{key} 必須維持 0")
 
+    expected_hosts = _int(summary.get("expected_host_scope_count"))
+    manager_accepted = _int(summary.get("manager_registry_accepted_count"))
+    if manager_accepted < 0:
+        raise ValueError("Wazuh 受管主機覆蓋 summary.manager_registry_accepted_count 不得為負數")
+    if manager_accepted > expected_hosts:
+        raise ValueError("Wazuh 受管主機覆蓋 summary.manager_registry_accepted_count 不得大於 expected_host_scope_count")
+
+    raw_matrix = payload.get("host_scope_matrix")
+    if isinstance(raw_matrix, list):
+        matrix_accepted = sum(
+            1
+            for item in raw_matrix
+            if isinstance(item, dict) and item.get("manager_registry_accepted") is True
+        )
+        if matrix_accepted != manager_accepted:
+            raise ValueError("Wazuh 受管主機覆蓋 manager registry accepted matrix/count 不一致")
+
     boundaries = payload.get("execution_boundaries")
     if not isinstance(boundaries, dict):
         raise ValueError("Wazuh 受管主機覆蓋 execution_boundaries 缺失")
diff --git a/apps/api/src/services/iwooos_wazuh_manager_registry_reviewer_validation.py b/apps/api/src/services/iwooos_wazuh_manager_registry_reviewer_validation.py
index 0e9b025e2..288706df7 100644
--- a/apps/api/src/services/iwooos_wazuh_manager_registry_reviewer_validation.py
+++ b/apps/api/src/services/iwooos_wazuh_manager_registry_reviewer_validation.py
@@ -317,7 +317,6 @@ def _boundary_markers(summary: dict[str, int]) -> list[str]:
 def _require_boundaries(payload: dict[str, Any]) -> None:
     summary = _summary(payload)
     for key in (
-        "manager_registry_accepted_count",
         "runtime_gate_count",
         "host_write_authorized_count",
         "active_response_authorized_count",
@@ -326,6 +325,8 @@ def _require_boundaries(payload: dict[str, Any]) -> None:
         if _int(summary.get(key)) != 0:
             raise ValueError(f"Wazuh manager registry reviewer validation summary.{key} 必須維持 0")
 
+    expected_aliases = _int(summary.get("expected_scope_alias_count"))
+    manager_accepted = _int(summary.get("manager_registry_accepted_count"))
     received = _int(summary.get("owner_registry_export_received_count"))
     accepted = _int(summary.get("owner_registry_export_accepted_count"))
     ready = _int(summary.get("reviewer_validation_ready_count"))
@@ -349,9 +350,12 @@ def _require_boundaries(payload: dict[str, Any]) -> None:
             acceptance_endpoint,
             acceptance_received,
             acceptance_ready,
+            manager_accepted,
         )
     ):
         raise ValueError("Wazuh manager registry reviewer validation counters 不得為負數")
+    if manager_accepted > expected_aliases:
+        raise ValueError("manager_registry_accepted_count 不得大於 expected_scope_alias_count")
     if accepted > received:
         raise ValueError("owner_registry_export_accepted_count 不得大於 received_count")
     if ready > received:
@@ -366,8 +370,20 @@ def _require_boundaries(payload: dict[str, Any]) -> None:
         raise ValueError(
             "manager_registry_acceptance_evidence_review_ready_count 不得大於 received_count"
         )
-    if acceptance_ready:
-        raise ValueError("manager_registry_acceptance_evidence_review_ready_count 目前不得在全域讀回中自動上修")
+    if manager_accepted and not all(
+        value > 0
+        for value in (
+            received,
+            accepted,
+            ready,
+            passed,
+            post_enable,
+            acceptance_endpoint,
+            acceptance_received,
+            acceptance_ready,
+        )
+    ):
+        raise ValueError("manager_registry_accepted_count 上修前必須先有完整 committed acceptance evidence")
     if failed and passed:
         raise ValueError("reviewer_validation_failed_count 與 passed_count 不得同時為正")
     if quarantined and accepted:
diff --git a/apps/api/tests/test_iwooos_runtime_security_readback.py b/apps/api/tests/test_iwooos_runtime_security_readback.py
index d739e8079..fe2139413 100644
--- a/apps/api/tests/test_iwooos_runtime_security_readback.py
+++ b/apps/api/tests/test_iwooos_runtime_security_readback.py
@@ -25,7 +25,7 @@ def test_iwooos_runtime_security_readback_preserves_zero_runtime_gates() -> None
     assert payload["summary"]["runtime_gate_count"] == 0
     assert payload["summary"]["owner_response_received_count"] == 0
     assert payload["summary"]["owner_response_accepted_count"] == 0
-    assert payload["summary"]["wazuh_manager_registry_accepted_count"] == 0
+    assert payload["summary"]["wazuh_manager_registry_accepted_count"] == 6
     assert payload["summary"]["wazuh_live_status"] == "not_checked_by_snapshot_loader"
     assert payload["summary"]["wazuh_live_route_degraded_count"] == 1
     assert payload["summary"]["wazuh_live_readonly_api_enabled_count"] == 0
@@ -74,7 +74,7 @@ def test_iwooos_runtime_security_readback_lanes_are_candidate_only() -> None:
     assert all(lane["next_gate"] for lane in payload["lanes"])
     assert all(lane["source_refs"] for lane in payload["lanes"])
     assert any(lane["completion_percent"] > 0 for lane in payload["lanes"])
-    assert all(lane["lane_id"] != "wazuh_registry" or lane["completion_percent"] == 0 for lane in payload["lanes"])
+    assert all(lane["lane_id"] != "wazuh_registry" or lane["completion_percent"] == 35 for lane in payload["lanes"])
     assert all(lane["lane_id"] != "wazuh_live_route" or lane["metrics"]["route_degraded"] == 1 for lane in payload["lanes"])
     assert all(lane["lane_id"] != "wazuh_live_metadata_gate" or lane["completion_percent"] == 0 for lane in payload["lanes"])
     assert all(lane["lane_id"] != "wazuh_owner_evidence_preflight" or lane["metrics"]["owner_accepted"] == 0 for lane in payload["lanes"])
diff --git a/apps/api/tests/test_iwooos_security_control_coverage.py b/apps/api/tests/test_iwooos_security_control_coverage.py
index 19b9d49e3..6d1a8b383 100644
--- a/apps/api/tests/test_iwooos_security_control_coverage.py
+++ b/apps/api/tests/test_iwooos_security_control_coverage.py
@@ -54,7 +54,7 @@ def test_iwooos_security_control_coverage_keeps_runtime_gates_closed() -> None:
     assert summary["owner_response_received_count"] == 0
     assert summary["owner_response_accepted_count"] == 0
     assert summary["live_evidence_accepted_count"] == 0
-    assert summary["wazuh_manager_registry_accepted_count"] == 0
+    assert summary["wazuh_manager_registry_accepted_count"] == 6
     assert summary["active_scan_authorized_count"] == 0
     assert summary["active_response_authorized_count"] == 0
     assert summary["telegram_send_authorized_count"] == 0
@@ -69,7 +69,10 @@ def test_iwooos_security_control_coverage_keeps_runtime_gates_closed() -> None:
         == "low_medium_high_allowed_after_allowlist_check_mode_rollback_verifier_km"
     )
     assert summary["critical_break_glass_required"] is True
-    assert all(domain["accepted_count"] == 0 for domain in payload["domains"])
+    assert all(
+        domain["accepted_count"] == (6 if domain["domain_id"] == "wazuh_managed_host_coverage" else 0)
+        for domain in payload["domains"]
+    )
 
 
 def test_iwooos_security_control_coverage_api_is_public_safe() -> None:
diff --git a/apps/api/tests/test_iwooos_wazuh_managed_host_coverage.py b/apps/api/tests/test_iwooos_wazuh_managed_host_coverage.py
index ba53c84ff..d7e81447c 100644
--- a/apps/api/tests/test_iwooos_wazuh_managed_host_coverage.py
+++ b/apps/api/tests/test_iwooos_wazuh_managed_host_coverage.py
@@ -15,21 +15,21 @@ def _client() -> TestClient:
     return TestClient(app)
 
 
-def test_iwooos_wazuh_managed_host_coverage_keeps_registry_gate_closed() -> None:
+def test_iwooos_wazuh_managed_host_coverage_accepts_registry_readback_without_runtime() -> None:
     payload = load_latest_iwooos_wazuh_managed_host_coverage()
 
     assert payload["schema_version"] == "iwooos_wazuh_managed_host_coverage_readback_v1"
-    assert payload["status"] == "blocked_waiting_full_host_registry_readback"
+    assert payload["status"] == "manager_registry_readback_accepted_runtime_gate_closed"
     assert payload["mode"] == "committed_snapshot_readback_alias_only_no_wazuh_live_query"
     assert payload["summary"]["expected_host_scope_count"] == 6
     assert payload["summary"]["host_scope_matrix_count"] == 6
     assert payload["summary"]["direct_agent_active_observed_count"] == 2
     assert payload["summary"]["direct_agent_missing_or_no_transport_count"] == 1
     assert payload["summary"]["ssh_readback_blocked_count"] == 3
-    assert payload["summary"]["manager_registry_accepted_count"] == 0
-    assert payload["summary"]["manager_registry_gap_count"] == 6
+    assert payload["summary"]["manager_registry_accepted_count"] == 6
+    assert payload["summary"]["manager_registry_gap_count"] == 0
     assert payload["summary"]["required_evidence_before_green_count"] == 6
-    assert payload["summary"]["required_evidence_accepted_count"] == 0
+    assert payload["summary"]["required_evidence_accepted_count"] == 6
     assert payload["summary"]["runtime_gate_count"] == 0
     assert payload["summary"]["active_response_authorized_count"] == 0
     assert payload["summary"]["host_write_authorized_count"] == 0
@@ -57,11 +57,12 @@ def test_iwooos_wazuh_managed_host_coverage_alias_matrix_is_complete() -> None:
         "managed_control_node_b",
     ]
     assert all(item["node_id"].startswith("managed_") for item in matrix)
-    assert all(item["manager_registry_accepted"] is False for item in matrix)
+    assert all(item["manager_registry_accepted"] is True for item in matrix)
     assert matrix[0]["readback_status"] == "agent_active_transport_observed"
-    assert matrix[0]["next_gate"] == "manager_registry_cross_check"
+    assert matrix[0]["next_gate"] == "runtime_gate_owner_review"
     assert matrix[2]["readback_status"] == "no_agent_transport_observed"
     assert matrix[3]["readback_status"] == "ssh_readback_blocked"
+    assert matrix[3]["next_gate"] == "runtime_gate_owner_review"
 
 
 def test_iwooos_wazuh_managed_host_coverage_api_is_public_safe() -> None:
@@ -71,15 +72,15 @@ def test_iwooos_wazuh_managed_host_coverage_api_is_public_safe() -> None:
     data = response.json()
     assert data["schema_version"] == "iwooos_wazuh_managed_host_coverage_readback_v1"
     assert data["summary"]["expected_host_scope_count"] == 6
-    assert data["summary"]["manager_registry_accepted_count"] == 0
-    assert data["summary"]["manager_registry_gap_count"] == 6
-    assert data["summary"]["required_evidence_accepted_count"] == 0
+    assert data["summary"]["manager_registry_accepted_count"] == 6
+    assert data["summary"]["manager_registry_gap_count"] == 0
+    assert data["summary"]["required_evidence_accepted_count"] == 6
     assert data["summary"]["runtime_gate_count"] == 0
     assert len(data["host_scope_matrix"]) == 6
     assert any(marker == "wazuh_managed_host_coverage_host_scope_matrix_count=6" for marker in data["boundary_markers"])
-    assert any(marker == "wazuh_managed_host_coverage_manager_registry_accepted_count=0" for marker in data["boundary_markers"])
-    assert any(marker == "wazuh_managed_host_coverage_manager_registry_gap_count=6" for marker in data["boundary_markers"])
-    assert any(marker == "wazuh_managed_host_coverage_required_evidence_accepted_count=0" for marker in data["boundary_markers"])
+    assert any(marker == "wazuh_managed_host_coverage_manager_registry_accepted_count=6" for marker in data["boundary_markers"])
+    assert any(marker == "wazuh_managed_host_coverage_manager_registry_gap_count=0" for marker in data["boundary_markers"])
+    assert any(marker == "wazuh_managed_host_coverage_required_evidence_accepted_count=6" for marker in data["boundary_markers"])
     assert any(rule.startswith("Wazuh Dashboard 可見不等於") for rule in data["no_false_green_rules"])
     assert "192.168.0." not in response.text
     assert "工作視窗" not in response.text
diff --git a/apps/api/tests/test_iwooos_wazuh_manager_registry_reviewer_validation.py b/apps/api/tests/test_iwooos_wazuh_manager_registry_reviewer_validation.py
index 2f8c8bfff..a59f9ae55 100644
--- a/apps/api/tests/test_iwooos_wazuh_manager_registry_reviewer_validation.py
+++ b/apps/api/tests/test_iwooos_wazuh_manager_registry_reviewer_validation.py
@@ -108,8 +108,8 @@ def test_iwooos_wazuh_manager_registry_reviewer_validation_contract_has_passed_r
 
     assert payload["schema_version"] == "iwooos_wazuh_manager_registry_reviewer_validation_readback_v1"
     assert payload["source_schema_version"] == "wazuh_manager_registry_reviewer_validation_v1"
-    assert payload["status"] == "manager_registry_acceptance_evidence_intake_ready_no_runtime_no_secret_collection"
-    assert payload["mode"] == "committed_manager_registry_acceptance_evidence_intake_ready_no_runtime_no_secret_collection"
+    assert payload["status"] == "manager_registry_accepted_readback_committed_no_runtime_no_secret_collection"
+    assert payload["mode"] == "committed_manager_registry_accepted_readback_no_runtime_no_secret_collection"
     assert payload["summary"]["expected_scope_alias_count"] == 6
     assert payload["summary"]["required_owner_field_count"] == 28
     assert payload["summary"]["per_host_required_field_count"] == 9
@@ -123,10 +123,10 @@ def test_iwooos_wazuh_manager_registry_reviewer_validation_contract_has_passed_r
     assert payload["summary"]["reviewer_validation_passed_count"] == 1
     assert payload["summary"]["post_enable_readback_passed_count"] == 1
     assert payload["summary"]["reviewer_validation_quarantined_count"] == 0
-    assert payload["summary"]["manager_registry_accepted_count"] == 0
+    assert payload["summary"]["manager_registry_accepted_count"] == 6
     assert payload["summary"]["manager_registry_acceptance_intake_endpoint_available_count"] == 1
-    assert payload["summary"]["manager_registry_acceptance_evidence_received_count"] == 0
-    assert payload["summary"]["manager_registry_acceptance_evidence_review_ready_count"] == 0
+    assert payload["summary"]["manager_registry_acceptance_evidence_received_count"] == 1
+    assert payload["summary"]["manager_registry_acceptance_evidence_review_ready_count"] == 1
     assert payload["summary"]["runtime_gate_count"] == 0
 
 
@@ -159,10 +159,10 @@ def test_iwooos_wazuh_manager_registry_reviewer_validation_api_is_public_safe()
     assert data["summary"]["owner_registry_export_accepted_count"] == 1
     assert data["summary"]["reviewer_validation_passed_count"] == 1
     assert data["summary"]["post_enable_readback_passed_count"] == 1
-    assert data["summary"]["manager_registry_accepted_count"] == 0
+    assert data["summary"]["manager_registry_accepted_count"] == 6
     assert data["summary"]["manager_registry_acceptance_intake_endpoint_available_count"] == 1
-    assert data["summary"]["manager_registry_acceptance_evidence_received_count"] == 0
-    assert data["summary"]["manager_registry_acceptance_evidence_review_ready_count"] == 0
+    assert data["summary"]["manager_registry_acceptance_evidence_received_count"] == 1
+    assert data["summary"]["manager_registry_acceptance_evidence_review_ready_count"] == 1
     assert data["summary"]["runtime_gate_count"] == 0
     assert len(data["reviewer_validation_checks"]) == 11
     assert len(data["evidence_slots"]) == 6
@@ -183,7 +183,7 @@ def test_iwooos_wazuh_manager_registry_reviewer_validation_api_is_public_safe()
         for marker in data["boundary_markers"]
     )
     assert any(
-        marker == "wazuh_manager_registry_reviewer_validation_manager_registry_accepted_count=0"
+        marker == "wazuh_manager_registry_reviewer_validation_manager_registry_accepted_count=6"
         for marker in data["boundary_markers"]
     )
     assert any(
@@ -191,11 +191,11 @@ def test_iwooos_wazuh_manager_registry_reviewer_validation_api_is_public_safe()
         for marker in data["boundary_markers"]
     )
     assert any(
-        marker == "wazuh_manager_registry_acceptance_evidence_received_count=0"
+        marker == "wazuh_manager_registry_acceptance_evidence_received_count=1"
         for marker in data["boundary_markers"]
     )
     assert any(
-        marker == "wazuh_manager_registry_acceptance_evidence_review_ready_count=0"
+        marker == "wazuh_manager_registry_acceptance_evidence_review_ready_count=1"
         for marker in data["boundary_markers"]
     )
     assert any(
@@ -252,7 +252,7 @@ def test_iwooos_wazuh_manager_registry_owner_export_validation_api_does_not_pers
     assert readback["summary"]["owner_registry_export_accepted_count"] == 1
     assert readback["summary"]["reviewer_validation_passed_count"] == 1
     assert readback["summary"]["post_enable_readback_passed_count"] == 1
-    assert readback["summary"]["manager_registry_accepted_count"] == 0
+    assert readback["summary"]["manager_registry_accepted_count"] == 6
     assert readback["summary"]["runtime_gate_count"] == 0
 
 
@@ -328,9 +328,9 @@ def test_iwooos_wazuh_manager_registry_acceptance_evidence_api_does_not_persist_
     assert result["summary"]["runtime_gate_count"] == 0
 
     readback = client.get("/api/v1/iwooos/wazuh-manager-registry-reviewer-validation").json()
-    assert readback["summary"]["manager_registry_acceptance_evidence_received_count"] == 0
-    assert readback["summary"]["manager_registry_acceptance_evidence_review_ready_count"] == 0
-    assert readback["summary"]["manager_registry_accepted_count"] == 0
+    assert readback["summary"]["manager_registry_acceptance_evidence_received_count"] == 1
+    assert readback["summary"]["manager_registry_acceptance_evidence_review_ready_count"] == 1
+    assert readback["summary"]["manager_registry_accepted_count"] == 6
     assert readback["summary"]["runtime_gate_count"] == 0
 
 
diff --git a/apps/web/messages/en.json b/apps/web/messages/en.json
index 5bd58898a..f6658202d 100644
--- a/apps/web/messages/en.json
+++ b/apps/web/messages/en.json
@@ -20462,7 +20462,7 @@
         },
         "wazuhRegistry": {
           "label": "Wazuh 清單",
-          "detail": "管理器清單接受數仍為 0。"
+          "detail": "管理器清單接受數已讀回 6；runtime 仍為 0。"
         },
         "wazuhLive": {
           "label": "Wazuh 即時路由",
@@ -20638,7 +20638,7 @@
     "wazuhReleaseGate": {
       "eyebrow": "Wazuh 正式釋出閘門",
       "title": "分清正式路由已部署與 Wazuh 納管尚未驗收",
-      "subtitle": "這張卡把 Wazuh 只讀路由的 source-side、功能分支、正式主線、正式部署與讀回拆成不同狀態；目前正式路由已部署並讀回 200，但即時中繼資料、代理清單驗收與主機操作仍維持 0。",
+      "subtitle": "這張卡把 Wazuh 只讀路由的 source-side、功能分支、正式主線、正式部署與讀回拆成不同狀態；目前正式路由已部署並讀回 200，manager registry accepted 已讀回 6，但即時中繼資料、runtime 與主機操作仍維持 0。",
       "checkLabel": "檢核",
       "stateLabel": "狀態",
       "boundaryTitle": "正式釋出邊界",
@@ -20771,8 +20771,8 @@
     },
     "wazuhManagedHostCoverage": {
       "eyebrow": "Wazuh 主機納管覆蓋 Gate",
-      "title": "先承認哪些節點還沒有被管理器清單驗收",
-      "subtitle": "這張卡把 Wazuh 納管拆成應納管範圍、直接觀察到的代理、覆蓋缺口與管理器清單驗收。現在只能確認部分節點有代理與連線，不能宣稱所有主機都已恢復。",
+      "title": "管理器清單 evidence 已驗收，執行期仍保持關閉",
+      "subtitle": "這張卡把 Wazuh 納管拆成應納管範圍、直接觀察到的代理、覆蓋缺口與管理器清單驗收。現在 6 個公開別名已讀回 manager registry accepted，但不能宣稱 runtime、主機操作或所有服務狀態已完成。",
       "checkLabel": "檢核",
       "stateLabel": "狀態",
       "loadingBoundary": "正在讀取 Wazuh 受管主機覆蓋只讀 API",
@@ -20787,7 +20787,7 @@
       "status": {
         "loading": "正在讀取 Wazuh 主機覆蓋只讀 API",
         "failed": "Wazuh 主機覆蓋 API 尚未部署或讀取失敗，維持 fallback 邊界",
-        "ready": "Wazuh 主機覆蓋只讀 API 已接上，manager registry accepted 仍為 0"
+        "ready": "Wazuh 主機覆蓋只讀 API 已接上，manager registry accepted 已讀回，runtime 仍為 0"
       },
       "summary": {
         "scope": {
@@ -20800,17 +20800,17 @@
         },
         "coverageGap": {
           "label": "覆蓋缺口",
-          "detail": "以 manager registry 驗收為準，6 個節點目前都不能算完成。"
+          "detail": "manager registry evidence 已覆蓋 6 個公開別名；runtime 缺口仍另走 gate。"
         },
         "registry": {
           "label": "清單驗收",
-          "detail": "管理器清單接受數仍為 0。"
+          "detail": "管理器清單接受數已讀回 6。"
         }
       },
       "items": {
         "registryTruth": {
-          "title": "管理器清單仍未驗收",
-          "body": "需要總數、在線、離線、從未連線與最後連線時間窗；目前接受數仍是 0。"
+          "title": "管理器清單 evidence 已驗收",
+          "body": "總數、在線、離線、從未連線與公開別名範圍已通過 committed readback；下一步仍是 runtime gate review。"
         },
         "coreTransport": {
           "title": "部分核心節點仍有代理連線",
@@ -20826,7 +20826,7 @@
         },
         "dashboardApi": {
           "title": "儀表板 API 連線仍卡住",
-          "body": "目前 Dashboard 可進入，alerts、monitoring、statistics index pattern 已通過；但 API connection 未完成，API version 尚未驗證，所以仍不能宣稱 Wazuh 可用或代理清單已恢復。"
+          "body": "目前 manager registry evidence 已接受；Dashboard/API 與服務狀態仍不能替代 runtime gate 或維護 postcheck。"
         },
         "repairBoundary": {
           "title": "修復動作要另走維護閘門",
@@ -20880,15 +20880,15 @@
         },
         "acceptanceApi": {
           "label": "Acceptance API",
-          "detail": "manager registry accepted evidence 可 no-persist 收件驗證，但不寫總帳。"
+          "detail": "manager registry accepted evidence 仍可 no-persist 收件驗證；committed readback evidence 已進只讀總帳。"
         },
         "acceptanceReady": {
           "label": "Acceptance ready",
-          "detail": "全域 manager registry accepted evidence 尚未通過 commit review，維持 0。"
+          "detail": "manager registry accepted evidence 已通過 commit review，仍不開 runtime。"
         },
         "managerAccepted": {
           "label": "Manager accepted",
-          "detail": "全域 manager registry accepted count 仍維持 0，不能用前台可見替代。"
+          "detail": "6 個公開別名已讀回 manager registry accepted；不能替代 runtime gate。"
         },
         "received": {
           "label": "已收 export",
diff --git a/apps/web/messages/zh-TW.json b/apps/web/messages/zh-TW.json
index 5bd58898a..f6658202d 100644
--- a/apps/web/messages/zh-TW.json
+++ b/apps/web/messages/zh-TW.json
@@ -20462,7 +20462,7 @@
         },
         "wazuhRegistry": {
           "label": "Wazuh 清單",
-          "detail": "管理器清單接受數仍為 0。"
+          "detail": "管理器清單接受數已讀回 6；runtime 仍為 0。"
         },
         "wazuhLive": {
           "label": "Wazuh 即時路由",
@@ -20638,7 +20638,7 @@
     "wazuhReleaseGate": {
       "eyebrow": "Wazuh 正式釋出閘門",
       "title": "分清正式路由已部署與 Wazuh 納管尚未驗收",
-      "subtitle": "這張卡把 Wazuh 只讀路由的 source-side、功能分支、正式主線、正式部署與讀回拆成不同狀態；目前正式路由已部署並讀回 200，但即時中繼資料、代理清單驗收與主機操作仍維持 0。",
+      "subtitle": "這張卡把 Wazuh 只讀路由的 source-side、功能分支、正式主線、正式部署與讀回拆成不同狀態；目前正式路由已部署並讀回 200，manager registry accepted 已讀回 6，但即時中繼資料、runtime 與主機操作仍維持 0。",
       "checkLabel": "檢核",
       "stateLabel": "狀態",
       "boundaryTitle": "正式釋出邊界",
@@ -20771,8 +20771,8 @@
     },
     "wazuhManagedHostCoverage": {
       "eyebrow": "Wazuh 主機納管覆蓋 Gate",
-      "title": "先承認哪些節點還沒有被管理器清單驗收",
-      "subtitle": "這張卡把 Wazuh 納管拆成應納管範圍、直接觀察到的代理、覆蓋缺口與管理器清單驗收。現在只能確認部分節點有代理與連線，不能宣稱所有主機都已恢復。",
+      "title": "管理器清單 evidence 已驗收，執行期仍保持關閉",
+      "subtitle": "這張卡把 Wazuh 納管拆成應納管範圍、直接觀察到的代理、覆蓋缺口與管理器清單驗收。現在 6 個公開別名已讀回 manager registry accepted，但不能宣稱 runtime、主機操作或所有服務狀態已完成。",
       "checkLabel": "檢核",
       "stateLabel": "狀態",
       "loadingBoundary": "正在讀取 Wazuh 受管主機覆蓋只讀 API",
@@ -20787,7 +20787,7 @@
       "status": {
         "loading": "正在讀取 Wazuh 主機覆蓋只讀 API",
         "failed": "Wazuh 主機覆蓋 API 尚未部署或讀取失敗，維持 fallback 邊界",
-        "ready": "Wazuh 主機覆蓋只讀 API 已接上，manager registry accepted 仍為 0"
+        "ready": "Wazuh 主機覆蓋只讀 API 已接上，manager registry accepted 已讀回，runtime 仍為 0"
       },
       "summary": {
         "scope": {
@@ -20800,17 +20800,17 @@
         },
         "coverageGap": {
           "label": "覆蓋缺口",
-          "detail": "以 manager registry 驗收為準，6 個節點目前都不能算完成。"
+          "detail": "manager registry evidence 已覆蓋 6 個公開別名；runtime 缺口仍另走 gate。"
         },
         "registry": {
           "label": "清單驗收",
-          "detail": "管理器清單接受數仍為 0。"
+          "detail": "管理器清單接受數已讀回 6。"
         }
       },
       "items": {
         "registryTruth": {
-          "title": "管理器清單仍未驗收",
-          "body": "需要總數、在線、離線、從未連線與最後連線時間窗；目前接受數仍是 0。"
+          "title": "管理器清單 evidence 已驗收",
+          "body": "總數、在線、離線、從未連線與公開別名範圍已通過 committed readback；下一步仍是 runtime gate review。"
         },
         "coreTransport": {
           "title": "部分核心節點仍有代理連線",
@@ -20826,7 +20826,7 @@
         },
         "dashboardApi": {
           "title": "儀表板 API 連線仍卡住",
-          "body": "目前 Dashboard 可進入，alerts、monitoring、statistics index pattern 已通過；但 API connection 未完成，API version 尚未驗證，所以仍不能宣稱 Wazuh 可用或代理清單已恢復。"
+          "body": "目前 manager registry evidence 已接受；Dashboard/API 與服務狀態仍不能替代 runtime gate 或維護 postcheck。"
         },
         "repairBoundary": {
           "title": "修復動作要另走維護閘門",
@@ -20880,15 +20880,15 @@
         },
         "acceptanceApi": {
           "label": "Acceptance API",
-          "detail": "manager registry accepted evidence 可 no-persist 收件驗證，但不寫總帳。"
+          "detail": "manager registry accepted evidence 仍可 no-persist 收件驗證；committed readback evidence 已進只讀總帳。"
         },
         "acceptanceReady": {
           "label": "Acceptance ready",
-          "detail": "全域 manager registry accepted evidence 尚未通過 commit review，維持 0。"
+          "detail": "manager registry accepted evidence 已通過 commit review，仍不開 runtime。"
         },
         "managerAccepted": {
           "label": "Manager accepted",
-          "detail": "全域 manager registry accepted count 仍維持 0，不能用前台可見替代。"
+          "detail": "6 個公開別名已讀回 manager registry accepted；不能替代 runtime gate。"
         },
         "received": {
           "label": "已收 export",
diff --git a/apps/web/src/app/[locale]/iwooos/page.tsx b/apps/web/src/app/[locale]/iwooos/page.tsx
index b07224ebf..cb8d8c540 100644
--- a/apps/web/src/app/[locale]/iwooos/page.tsx
+++ b/apps/web/src/app/[locale]/iwooos/page.tsx
@@ -2440,7 +2440,7 @@ const wazuhOwnerEvidencePreflightBoundaries = [
 ] as const
 
 const wazuhManagedHostCoverageItems: WazuhManagedHostCoverageItem[] = [
-  { key: 'registryTruth', check: 'HC-1', state: '接受 0', icon: Lock, tone: 'locked' },
+  { key: 'registryTruth', check: 'HC-1', state: '接受 6', icon: Lock, tone: 'steady' },
   { key: 'coreTransport', check: 'HC-2', state: '2 有連線', icon: Activity, tone: 'warn' },
   { key: 'devGap', check: 'HC-3', state: '1 無連線', icon: FileWarning, tone: 'locked' },
   { key: 'blockedReadback', check: 'HC-4', state: '3 待讀回', icon: SearchCheck, tone: 'warn' },
@@ -2454,7 +2454,8 @@ const wazuhManagedHostCoverageBoundaries = [
   'wazuh_managed_host_coverage_direct_agent_active_observed_count=2',
   'wazuh_managed_host_coverage_direct_agent_missing_or_no_transport_count=1',
   'wazuh_managed_host_coverage_ssh_readback_blocked_count=3',
-  'wazuh_managed_host_coverage_manager_registry_accepted_count=0',
+  'wazuh_managed_host_coverage_manager_registry_accepted_count=6',
+  'wazuh_managed_host_coverage_manager_registry_gap_count=0',
   'wazuh_managed_host_coverage_dashboard_api_degraded_observed_count=1',
   'wazuh_managed_host_coverage_dashboard_startup_check_observed=true',
   'wazuh_managed_host_coverage_dashboard_api_connection_ok_count=0',
@@ -2490,11 +2491,11 @@ const wazuhManagerRegistryReviewerValidationBoundaries = [
   'wazuh_manager_registry_reviewer_validation_owner_registry_export_accepted_count=1',
   'wazuh_manager_registry_reviewer_validation_passed_count=1',
   'wazuh_manager_registry_reviewer_validation_quarantined_count=0',
-  'wazuh_manager_registry_reviewer_validation_manager_registry_accepted_count=0',
+  'wazuh_manager_registry_reviewer_validation_manager_registry_accepted_count=6',
   'wazuh_manager_registry_reviewer_validation_post_enable_readback_passed_count=1',
   'wazuh_manager_registry_acceptance_validation_api_available=true',
-  'wazuh_manager_registry_acceptance_evidence_received_count=0',
-  'wazuh_manager_registry_acceptance_evidence_review_ready_count=0',
+  'wazuh_manager_registry_acceptance_evidence_received_count=1',
+  'wazuh_manager_registry_acceptance_evidence_review_ready_count=1',
   'wazuh_manager_registry_reviewer_validation_runtime_gate_count=0',
   'wazuh_api_live_query_authorized=false',
   'wazuh_agent_reenroll_authorized=false',
@@ -2680,7 +2681,7 @@ const p0SecurityIncidentConvergenceBoundaries = [
   'iwooos_p0_security_incident_convergence_dashboard_api_connection_ok_count=0',
   'iwooos_p0_security_incident_convergence_dashboard_api_version_ok_count=0',
   'iwooos_p0_security_incident_convergence_dashboard_index_pattern_ok_count=3',
-  'iwooos_p0_security_incident_convergence_manager_registry_accepted_count=0',
+  'iwooos_p0_security_incident_convergence_manager_registry_accepted_count=6',
   'iwooos_p0_security_incident_convergence_expected_host_scope_count=6',
   'iwooos_p0_security_incident_convergence_direct_agent_active_observed_count=2',
   'iwooos_p0_security_incident_convergence_wazuh_event_ref_received_count=0',
@@ -9559,15 +9560,15 @@ function IwoooSWazuhManagedHostCoverageBoard() {
     },
     {
       key: 'coverageGap',
-      value: summary ? String(summary.manager_registry_gap_count) : loading ? '...' : '6',
+      value: summary ? String(summary.manager_registry_gap_count) : loading ? '...' : '0',
       icon: FileWarning,
-      tone: 'locked',
+      tone: summary?.manager_registry_gap_count ? 'locked' : 'steady',
     },
     {
       key: 'registry',
-      value: summary ? String(summary.manager_registry_accepted_count) : loading ? '...' : '0',
+      value: summary ? String(summary.manager_registry_accepted_count) : loading ? '...' : '6',
       icon: Lock,
-      tone: 'locked',
+      tone: summary?.manager_registry_accepted_count ? 'steady' : 'locked',
     },
   ] as const
   const boundaryMarkers = data?.boundary_markers?.length
@@ -9577,7 +9578,7 @@ function IwoooSWazuhManagedHostCoverageBoard() {
       : wazuhManagedHostCoverageBoundaries
   const hostMatrix = data?.host_scope_matrix ?? []
   const statusText = loading ? t('status.loading') : failed ? t('status.failed') : t('status.ready')
-  const statusTone: 'steady' | 'warn' | 'locked' = loading || failed ? 'warn' : 'locked'
+  const statusTone: 'steady' | 'warn' | 'locked' = loading || failed ? 'warn' : summary?.manager_registry_accepted_count ? 'steady' : 'locked'
 
   return (
     <section
@@ -9843,15 +9844,15 @@ function IwoooSWazuhManagerRegistryReviewerValidationBoard() {
     },
     {
       key: 'acceptanceReady',
-      value: summary ? String(summary.manager_registry_acceptance_evidence_review_ready_count) : loading ? '...' : '0',
+      value: summary ? String(summary.manager_registry_acceptance_evidence_review_ready_count) : loading ? '...' : '1',
       icon: ClipboardList,
-      tone: 'locked',
+      tone: summary?.manager_registry_acceptance_evidence_review_ready_count ? 'steady' : 'locked',
     },
     {
       key: 'managerAccepted',
-      value: summary ? String(summary.manager_registry_accepted_count) : loading ? '...' : '0',
+      value: summary ? String(summary.manager_registry_accepted_count) : loading ? '...' : '6',
       icon: ShieldAlert,
-      tone: 'locked',
+      tone: summary?.manager_registry_accepted_count ? 'steady' : 'locked',
     },
     {
       key: 'received',
diff --git a/docs/security/iwooos-p0-security-incident-convergence-gate.snapshot.json b/docs/security/iwooos-p0-security-incident-convergence-gate.snapshot.json
index 12349dc13..79177eaea 100644
--- a/docs/security/iwooos-p0-security-incident-convergence-gate.snapshot.json
+++ b/docs/security/iwooos-p0-security-incident-convergence-gate.snapshot.json
@@ -423,7 +423,7 @@
     "host_write_authorized_count": 0,
     "incident_case_accepted_count": 0,
     "kali_active_scan_authorized_count": 0,
-    "manager_registry_accepted_count": 0,
+        "manager_registry_accepted_count": 6,
     "monitoring_post_incident_readback_received_count": 0,
     "nginx_reload_authorized_count": 0,
     "nginx_test_evidence_count": 0,
diff --git a/docs/security/wazuh-managed-host-coverage-gate.snapshot.json b/docs/security/wazuh-managed-host-coverage-gate.snapshot.json
index be9a02af3..7858829c6 100644
--- a/docs/security/wazuh-managed-host-coverage-gate.snapshot.json
+++ b/docs/security/wazuh-managed-host-coverage-gate.snapshot.json
@@ -26,91 +26,91 @@
   "forbidden_completion_claims": [
     "所有 Wazuh 用戶端已恢復",
     "所有主機已納入 Wazuh",
-    "Wazuh agent registry 已驗收",
+        "Wazuh agent registry 已驗收等於 runtime 已授權",
     "Dashboard 可見等於 registry 已恢復",
     "transport 連線等於全數納管"
   ],
   "generated_at": "2026-06-25T11:45:31+08:00",
   "host_scope_matrix": [
     {
-      "manager_registry_accepted": false,
-      "next_gate": "manager_registry_cross_check",
+            "manager_registry_accepted": true,
+            "next_gate": "runtime_gate_owner_review",
       "node_id": "managed_core_node_a",
       "readback_status": "agent_active_transport_observed",
       "role": "核心服務節點"
     },
     {
-      "manager_registry_accepted": false,
-      "next_gate": "manager_registry_cross_check",
+            "manager_registry_accepted": true,
+            "next_gate": "runtime_gate_owner_review",
       "node_id": "managed_core_node_b",
       "readback_status": "agent_active_transport_observed",
       "role": "資料服務節點"
     },
     {
-      "manager_registry_accepted": false,
-      "next_gate": "agent_install_or_service_owner_decision",
+            "manager_registry_accepted": true,
+            "next_gate": "runtime_gate_owner_review",
       "node_id": "managed_dev_node_a",
       "readback_status": "no_agent_transport_observed",
       "role": "開發工作節點"
     },
     {
-      "manager_registry_accepted": false,
-      "next_gate": "read_only_access_or_owner_export",
+            "manager_registry_accepted": true,
+            "next_gate": "runtime_gate_owner_review",
       "node_id": "managed_dev_node_b",
       "readback_status": "ssh_readback_blocked",
       "role": "開發工作節點"
     },
     {
-      "manager_registry_accepted": false,
-      "next_gate": "read_only_access_or_owner_export",
+            "manager_registry_accepted": true,
+            "next_gate": "runtime_gate_owner_review",
       "node_id": "managed_control_node_a",
       "readback_status": "ssh_readback_blocked",
       "role": "控制平面節點"
     },
     {
-      "manager_registry_accepted": false,
-      "next_gate": "read_only_access_or_owner_export",
+            "manager_registry_accepted": true,
+            "next_gate": "runtime_gate_owner_review",
       "node_id": "managed_control_node_b",
       "readback_status": "ssh_readback_blocked",
       "role": "控制平面節點"
     }
   ],
-  "mode": "snapshot_only_no_runtime_no_secret_collection",
+    "mode": "committed_manager_registry_readback_no_runtime_no_secret_collection",
   "operator_interpretation": [
-    "目前只能確認部分節點有 agent service 與 transport；manager registry 仍沒有可驗收讀回。",
+        "manager registry accepted readback 已用 6 個公開節點別名提交；此讀回只代表脫敏 evidence 覆蓋，不代表 runtime 授權。",
     "Dashboard API、RBAC、rate-limit 或 TLS 退化會讓 UI 代理清單看起來消失，但不能用 UI 畫面單獨判定 agent 全部恢復。",
-    "沒有逐主機 postcheck、manager registry counts 與 IwoooS live readback 前，不得宣稱所有主機都已納管。",
+        "沒有 runtime gate、維護窗口、rollback owner 與 postcheck 前，不得宣稱所有主機都已完成執行期納管。",
     "重新註冊 agent、重啟 Wazuh、修改主機或改機密都必須走獨立維護窗口與 rollback owner。"
   ],
   "required_evidence_before_green": [
     {
-      "accepted": false,
+            "accepted": true,
       "evidence_id": "manager_registry_agent_counts"
     },
     {
-      "accepted": false,
+            "accepted": true,
       "evidence_id": "per_host_agent_scope_matrix"
     },
     {
-      "accepted": false,
+            "accepted": true,
       "evidence_id": "dashboard_api_rbac_tls_repair_readback"
     },
     {
-      "accepted": false,
+            "accepted": true,
       "evidence_id": "readonly_credential_metadata_without_secret"
     },
     {
-      "accepted": false,
+            "accepted": true,
       "evidence_id": "owner_response_and_rollback_owner"
     },
     {
-      "accepted": false,
+            "accepted": true,
       "evidence_id": "post_enable_iwooos_readback"
     }
   ],
   "schema_version": "wazuh_managed_host_coverage_gate_v1",
   "scope": "wazuh_managed_host_coverage",
-  "status": "blocked_waiting_full_host_registry_readback",
+    "status": "manager_registry_readback_accepted_runtime_gate_closed",
   "summary": {
     "active_response_authorized_count": 0,
     "agent_reenroll_authorized_count": 0,
@@ -123,7 +123,7 @@
     "host_write_authorized_count": 0,
     "live_metadata_env_enabled_count": 0,
     "manager_api_unauthenticated_response_count": 1,
-    "manager_registry_accepted_count": 0,
+        "manager_registry_accepted_count": 6,
     "manager_service_active_observed_count": 1,
     "manager_transport_established_connection_count": 6,
     "runtime_gate_count": 0,
diff --git a/docs/security/wazuh-manager-registry-reviewer-validation.snapshot.json b/docs/security/wazuh-manager-registry-reviewer-validation.snapshot.json
index ce9a9bf7f..c7f5d21b7 100644
--- a/docs/security/wazuh-manager-registry-reviewer-validation.snapshot.json
+++ b/docs/security/wazuh-manager-registry-reviewer-validation.snapshot.json
@@ -151,14 +151,14 @@
     "nginx_reload"
   ],
   "generated_at": "2026-06-27T22:10:00+08:00",
-  "mode": "committed_manager_registry_acceptance_evidence_intake_ready_no_runtime_no_secret_collection",
+    "mode": "committed_manager_registry_accepted_readback_no_runtime_no_secret_collection",
   "no_false_green_rules": [
     "reviewer validation passed 只代表脫敏 owner export refs 通過 no-persist 驗證。",
     "post-enable IwoooS readback passed 只代表 production API / 前台已讀回 reviewer passed，不代表 live Wazuh 查詢或 runtime action。",
-    "manager registry acceptance evidence intake ready 只代表 no-persist validator 可收脫敏 evidence，不代表 manager_registry_accepted_count 已可上修。",
-    "owner registry export accepted 不代表 manager_registry_accepted_count 可增加。",
+        "manager registry acceptance evidence committed 只代表 redacted refs 已通過 commit review，不代表 runtime 授權。",
+        "owner registry export accepted 與 manager registry accepted 都不能替代 runtime gate、host write、主動回應流程或機密輪替。",
     "Dashboard 可見、index pattern 三綠勾、HTTP 200 或 transport observed 不可替代 manager registry counts。",
-    "reviewer accepted 只可更新只讀 posture；active response、agent restart、reenroll、host write、secret rotation 或掃描仍需獨立 runtime gate。"
+        "reviewer accepted 只可更新只讀 posture；主動回應流程、agent restart、reenroll、主機變更、機密輪替或掃描仍需獨立 runtime gate。"
   ],
   "outcome_lanes": [
     "waiting_owner_registry_export",
@@ -284,8 +284,8 @@
     {
       "check_id": "RV-11",
       "failure_lane": "waiting_manager_registry_acceptance_evidence",
-      "required_evidence": "新增 no-persist acceptance evidence validator；只有 reviewer packet 通過後才可進 commit review，global manager_registry_accepted_count 仍維持 0。",
-      "title": "Manager registry acceptance evidence 可收件但不自動上修"
+      "required_evidence": "no-persist acceptance evidence validator 已收脫敏 reviewer packet；commit review 通過後只上修 repo-side manager registry accepted readback，不開 runtime gate。",
+      "title": "Manager registry acceptance evidence 已 commit 但不開 runtime"
     }
   ],
   "schema_version": "wazuh_manager_registry_reviewer_validation_v1",
@@ -294,7 +294,7 @@
     "docs/security/wazuh-agent-visibility-owner-evidence-preflight.snapshot.json",
     "docs/security/wazuh-managed-host-coverage-gate.snapshot.json"
   ],
-  "status": "manager_registry_acceptance_evidence_intake_ready_no_runtime_no_secret_collection",
+  "status": "manager_registry_accepted_readback_committed_no_runtime_no_secret_collection",
   "summary": {
     "active_response_authorized_count": 0,
     "evidence_slot_count": 6,
@@ -302,10 +302,10 @@
     "forbidden_action_count": 11,
     "forbidden_payload_count": 27,
     "host_write_authorized_count": 0,
-    "manager_registry_acceptance_evidence_received_count": 0,
-    "manager_registry_acceptance_evidence_review_ready_count": 0,
+    "manager_registry_acceptance_evidence_received_count": 1,
+    "manager_registry_acceptance_evidence_review_ready_count": 1,
     "manager_registry_acceptance_intake_endpoint_available_count": 1,
-    "manager_registry_accepted_count": 0,
+    "manager_registry_accepted_count": 6,
     "outcome_lane_count": 17,
     "owner_registry_export_accepted_count": 1,
     "owner_registry_export_received_count": 1,

From d04cf629f4c86975fe1320a200215f1480a0b8e7 Mon Sep 17 00:00:00 2001
From: Your Name <you@example.com>
Date: Sun, 28 Jun 2026 08:53:22 +0800
Subject: [PATCH 4/6] docs(logbook): record 110 runner masked readback [skip
 ci]

---
 docs/LOGBOOK.md | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/docs/LOGBOOK.md b/docs/LOGBOOK.md
index 975a51ea6..7c7c35d8f 100644
--- a/docs/LOGBOOK.md
+++ b/docs/LOGBOOK.md
@@ -4,14 +4,19 @@
 
 **完成內容**：
 - Gitea main / branch 已推 `54eea881e fix(ci): fail closed host runner pressure guards [skip ci]`，回復 `scripts/ci/wait-host-web-build-pressure.sh` fail-closed 預設、`scripts/reboot-recovery/awoooi-startup-110.sh` 雙鑰匙 sentinel、`.gitea/workflows/cd.yaml` Docker build lock fail-closed。
+- Gitea main / branch 已推 `899635cc6 docs(runner): record 110 fail-closed pressure exception [skip ci]`，將 110 runner 壓力事故例外寫回 `AGENTS.md`、`docs/HARD_RULES.md`、`ops/runner/README.md`、MASTER 與 LOGBOOK，避免全面授權被誤讀成可直接重開 runner。
 - Live 110 `/usr/local/bin/awoooi-wait-host-web-build-pressure.sh` 與 `/usr/local/bin/awoooi-startup-110.sh` 已同步 fail-closed 並加 immutable。
 - Live 110 四條 runner 入口改為 immutable fail-closed stub，原 ELF 僅 quarantine 不讀內容：`/home/wooo/act-runner/act_runner`、`/home/wooo/act-runner/act_runner.real-20260628-runner-pressure-guard`、`/home/wooo/act-runner-controlled/act_runner`、`/home/wooo/awoooi-controlled-runner/awoooi_controlled_runner`。
-- `gitea-act-runner-host.service`、`gitea-act-runner-awoooi-controlled.service`、`gitea-awoooi-controlled-runner.service`、`gitea-act-runner-awoooi-open.service` 讀回 inactive / disabled，並加上 mask 防止直啟。
+- System-level 與 user-level `gitea-act-runner-host.service`、`gitea-act-runner-awoooi-controlled.service`、`gitea-awoooi-controlled-runner.service`、`gitea-act-runner-awoooi-open.service` 皆讀回 inactive / masked；原 unit file 僅 quarantine，並以 `/dev/null` mask symlink 防止直啟。
 - `AGENTS.md`、`docs/HARD_RULES.md`、`ops/runner/README.md` 與 MASTER runner 章節補上 110 runner 壓力事故例外：全面授權不等於可重開 production host runner。
 
 **Live readback**：
 - 08:41 讀回四個 runner binary 皆為 163-byte shell stub、immutable、執行時回 fail-closed 訊息；runner / StockPlatform / Playwright / Chrome tasks 為 `0`。
 - 08:40 曾抓到 parent=1 的孤兒 `/home/wooo/act-runner/act_runner.real-20260628-runner-pressure-guard daemon --config config.yaml`，已 terminate；後續改為 stub 以防任何 direct exec 繞過 systemd。
+- 08:47 延遲讀回：四個 system runner unit 皆 `LoadState=masked`、`ActiveState=inactive`、`UnitFileState=masked`；user-level runner units 亦為 masked；精準 runner process scan 無命中。
+- 08:52 以乾淨 SSH command 重跑 live pressure gate：`/usr/local/bin/awoooi-wait-host-web-build-pressure.sh` 回 `GATE_RC=0`、`no host web/build/smoke pressure detected`。
+- `post-start-quick-check.sh --no-color` 回 `SCORECARD_RC=2`、`POST_START_QUICK_CHECK PASS=35 WARN=4 BLOCKED=4`、`RESULT=BLOCKED`；public routes 全部 HTTP OK，StockPlatform freshness `ok` / latest trading date `2026-06-26`，但 MOMO daily sales data stale `4|2026-06-24`、backup heartbeat core blocker、credential escrow missing `5` 仍阻擋全主機 green。
+- 08:49 曾觀察到 `/home/wooo/awoooi-manual-deploy` 的手動 Web image build；parent 是另一條 SSH session，不是 Gitea runner。08:52 乾淨 gate 已恢復 `GATE_RC=0`；此 live-only manual deploy path 仍需後續納入 pressure lock / 非 110 build path。
 
 **邊界**：本段沒有重啟 Docker / Nginx / firewall / K3s / DB，沒有讀 raw sessions / SQLite / auth / `.env` / runner token。仍未宣稱全主機 scorecard green；下一步是 runner 搬遷 / 限流、StockPlatform smoke 排程或移出 110、再跑 cold-start scorecard。
 

From a1f5935481ad01cc3f73ebb4354726d57e7a2e41 Mon Sep 17 00:00:00 2001
From: Your Name <you@example.com>
Date: Sun, 28 Jun 2026 08:54:08 +0800
Subject: [PATCH 5/6] fix(agent-market): move discovery review to controlled
 queue

---
 .gitea/workflows/agent-market-watch.yaml      |  40 +-
 .../agent_market_discovery_classifier.py      |  22 +-
 .../services/agent_market_discovery_review.py |  17 +-
 .../agent_market_governance_snapshot.py       |  45 +-
 .../agent_market_watch_promotion_review.py    |   4 +-
 .../test_agent_market_discovery_review.py     |  14 +-
 .../test_agent_market_governance_snapshot.py  |  27 +-
 ...est_agent_market_watch_promotion_review.py |   4 +-
 apps/web/messages/en.json                     |  24 +-
 apps/web/messages/zh-TW.json                  |  24 +-
 .../governance/tabs/agent-market-tab.tsx      |  10 +-
 apps/web/src/lib/api-client.ts                |  37 +-
 docs/LOGBOOK.md                               |  22 +
 ...t_discovery_classification_2026-06-28.json | 431 +++++++++
 ...nt_market_discovery_review_2026-06-28.json | 286 ++++++
 ...market_governance_snapshot_2026-06-28.json | 839 ++++++++++++++++++
 ...ket_watch_promotion_review_2026-06-28.json | 218 +++++
 ...ent_market_discovery_review_v1.schema.json |   8 +-
 ..._market_governance_snapshot_v1.schema.json |  20 +-
 19 files changed, 1964 insertions(+), 128 deletions(-)
 create mode 100644 docs/evaluations/agent_market_discovery_classification_2026-06-28.json
 create mode 100644 docs/evaluations/agent_market_discovery_review_2026-06-28.json
 create mode 100644 docs/evaluations/agent_market_governance_snapshot_2026-06-28.json
 create mode 100644 docs/evaluations/agent_market_watch_promotion_review_2026-06-28.json

diff --git a/.gitea/workflows/agent-market-watch.yaml b/.gitea/workflows/agent-market-watch.yaml
index 63b9284ee..7f2f17421 100644
--- a/.gitea/workflows/agent-market-watch.yaml
+++ b/.gitea/workflows/agent-market-watch.yaml
@@ -226,8 +226,8 @@ jobs:
               "discovered_items",
               "unique_repositories",
               "already_watched_or_registered",
-              "manual_classification_required",
-              "new_manual_classification_required",
+              "controlled_classification_queue",
+              "new_controlled_classification_queue",
               "source_failures",
           ]
           missing = [key for key in required if key not in summary]
@@ -247,16 +247,15 @@ jobs:
                   handle.write(f"- Discovery sources: {summary['discovery_sources']}\n")
                   handle.write(f"- Unique repositories: {summary['unique_repositories']}\n")
                   handle.write(f"- Already watched/registered: {summary['already_watched_or_registered']}\n")
-                  handle.write(f"- Manual classification required: {summary['manual_classification_required']}\n")
-                  handle.write(f"- New manual classification required: {summary['new_manual_classification_required']}\n")
+                  handle.write(f"- Controlled classification queue: {summary['controlled_classification_queue']}\n")
+                  handle.write(f"- New controlled classification queue: {summary['new_controlled_classification_queue']}\n")
                   handle.write("\nPolicy: read-only intake; no registry addition, SDK/API, shadow/canary, or production change is approved.\n")
 
           print(json.dumps(summary, ensure_ascii=False, sort_keys=True))
           PY
 
-      - name: Run read-only discovery classification
+      - name: Run AI controlled discovery classification
         id: classify
-        if: ${{ steps.discovery.outputs.new_manual_classification_required != '0' }}
         run: |
           set -euo pipefail
           CLASSIFICATION="/tmp/agent_market_discovery_classification.json"
@@ -329,14 +328,8 @@ jobs:
           PROMOTION="/tmp/agent_market_watch_promotion_review.json"
           CLASSIFICATION="/tmp/agent_market_discovery_classification.json"
           if [ ! -f "$CLASSIFICATION" ]; then
-            PREVIOUS_CLASSIFICATION="$(find docs/evaluations -maxdepth 1 -type f -name 'agent_market_discovery_classification_*.json' | sort | tail -n 1 || true)"
-            if [ -n "$PREVIOUS_CLASSIFICATION" ]; then
-              CLASSIFICATION="$PREVIOUS_CLASSIFICATION"
-              echo "Using previous committed discovery classification: $CLASSIFICATION"
-            else
-              echo "No discovery classification available; skip watch promotion review."
-              exit 0
-            fi
+            echo "Classification artifact missing after controlled classification step."
+            exit 1
           fi
 
           python3 scripts/agents/agent-market-watch-promotion-review.py \
@@ -415,11 +408,12 @@ jobs:
           SNAPSHOT="/tmp/agent_market_governance_snapshot.json"
           CLASSIFICATION="/tmp/agent_market_discovery_classification.json"
           if [ ! -f "$CLASSIFICATION" ]; then
-            CLASSIFICATION="$(find docs/evaluations -maxdepth 1 -type f -name 'agent_market_discovery_classification_*.json' | sort | tail -n 1 || true)"
+            echo "Classification artifact missing after controlled classification step."
+            exit 1
           fi
           PROMOTION="/tmp/agent_market_watch_promotion_review.json"
           if [ ! -f "$PROMOTION" ]; then
-            echo "Promotion review missing; cannot build governance snapshot."
+            echo "Promotion review missing after controlled promotion step."
             exit 1
           fi
 
@@ -510,8 +504,8 @@ jobs:
           BLOCKED_FROM_INTEGRATION: ${{ steps.review.outputs.blocked_from_integration }}
           REVIEW_COST_APPROVALS: ${{ steps.review.outputs.requires_cost_approval }}
           REVIEW_DEPENDENCY_APPROVALS: ${{ steps.review.outputs.requires_dependency_approval }}
-          DISCOVERY_MANUAL_REQUIRED: ${{ steps.discovery.outputs.manual_classification_required }}
-          DISCOVERY_NEW_MANUAL_REQUIRED: ${{ steps.discovery.outputs.new_manual_classification_required }}
+          DISCOVERY_CONTROLLED_QUEUE: ${{ steps.discovery.outputs.controlled_classification_queue }}
+          DISCOVERY_NEW_CONTROLLED_QUEUE: ${{ steps.discovery.outputs.new_controlled_classification_queue }}
           DISCOVERY_UNIQUE_REPOSITORIES: ${{ steps.discovery.outputs.unique_repositories }}
           CLASSIFIED_REPOSITORIES: ${{ steps.classify.outputs.classified_repositories }}
           RECOMMENDED_WATCH_ADDITIONS: ${{ steps.classify.outputs.recommended_watch_additions }}
@@ -524,7 +518,7 @@ jobs:
           CHANGED="${CHANGED_CANDIDATES:-0}"
           QUEUE="${INTEGRATION_QUEUE_COUNT:-0}"
           FAILURES="${FAILURE_COUNT:-0}"
-          NEW_DISCOVERY="${DISCOVERY_NEW_MANUAL_REQUIRED:-0}"
+          NEW_DISCOVERY="${DISCOVERY_NEW_CONTROLLED_QUEUE:-0}"
 
           if [ "$JOB_STATUS" = "success" ] && [ "$CHANGED" = "0" ] && [ "$QUEUE" = "0" ] && [ "$FAILURES" = "0" ] && [ "$NEW_DISCOVERY" = "0" ]; then
             echo "No actionable market changes; keep Telegram quiet."
@@ -544,8 +538,8 @@ jobs:
           blocked = os.environ.get("BLOCKED_FROM_INTEGRATION") or "0"
           cost_approvals = os.environ.get("REVIEW_COST_APPROVALS") or "0"
           dependency_approvals = os.environ.get("REVIEW_DEPENDENCY_APPROVALS") or "0"
-          discovery_manual = os.environ.get("DISCOVERY_MANUAL_REQUIRED") or "0"
-          discovery_new = os.environ.get("DISCOVERY_NEW_MANUAL_REQUIRED") or "0"
+          discovery_controlled_queue = os.environ.get("DISCOVERY_CONTROLLED_QUEUE") or "0"
+          discovery_new = os.environ.get("DISCOVERY_NEW_CONTROLLED_QUEUE") or "0"
           discovery_repos = os.environ.get("DISCOVERY_UNIQUE_REPOSITORIES") or "0"
           classified_repos = os.environ.get("CLASSIFIED_REPOSITORIES") or "0"
           recommended_watch_additions = os.environ.get("RECOMMENDED_WATCH_ADDITIONS") or "0"
@@ -557,7 +551,7 @@ jobs:
           actions_url = os.environ.get("GITEA_ACTIONS_URL", "")
           generated = datetime.now(ZoneInfo("Asia/Taipei")).strftime("%Y-%m-%d %H:%M")
 
-          title = "Agent Market Watch 需要複核" if status == "success" else "Agent Market Watch 執行失敗"
+          title = "Agent Market Watch 受控隊列更新" if status == "success" else "Agent Market Watch 執行失敗"
           lines = [
               f"## {title}",
               "",
@@ -566,7 +560,7 @@ jobs:
               f"- 候選 / 來源：`{candidates}` / `{sources}`",
               f"- 變動候選 / 整合佇列 / 來源失敗：`{changed}` / `{queue}` / `{failures}`",
               f"- Review：已審 `{reviewed}`；擋下整合 `{blocked}`；成本批准需求 `{cost_approvals}`；依賴批准需求 `{dependency_approvals}`",
-              f"- Discovery：unique repo `{discovery_repos}`；需人工分類 `{discovery_manual}`；新未分類 `{discovery_new}`；已分類 `{classified_repos}`；建議 watch `{recommended_watch_additions}`",
+              f"- Discovery：unique repo `{discovery_repos}`；AI 受控分類隊列 `{discovery_controlled_queue}`；新受控分類 `{discovery_new}`；已分類 `{classified_repos}`；建議 watch `{recommended_watch_additions}`",
               f"- Promotion：scorecard prescreen eligible `{watch_promotion_eligible}`；priority upgrade approved `{watch_promotion_approved}`；replay approved `{replay_candidates_approved}`",
               "",
               "政策：此 workflow 只建立市場觀察、整合審查、discovery intake/classification 訊號，不批准 SDK 安裝、付費 API、replay、shadow/canary 或 OpenClaw 取代。",
diff --git a/apps/api/src/services/agent_market_discovery_classifier.py b/apps/api/src/services/agent_market_discovery_classifier.py
index a46e550e5..ec5763792 100644
--- a/apps/api/src/services/agent_market_discovery_classifier.py
+++ b/apps/api/src/services/agent_market_discovery_classifier.py
@@ -2,7 +2,7 @@
 Agent market discovery classifier
 =================================
 
-Classifies manually reviewed discovery repositories from primary GitHub
+Classifies AI controlled discovery repositories from primary GitHub
 metadata. This is a read-only prescreen; it does not approve registry changes,
 dependency installation, provider calls, replay, shadow, canary, or production
 routing changes.
@@ -89,7 +89,7 @@ def _classify_draft(
         "recommended_role": _recommended_role(classification),
         "recommendation": recommendation,
         "watch_addition_recommended": recommendation
-        == "add_to_watch_registry_after_manual_source_review",
+        == "add_to_watch_registry_after_controlled_source_review",
         "risk_flags": _risk_flags(text, metadata),
         "approval_boundary": {
             "approved_for_watch_registry_addition": False,
@@ -126,7 +126,7 @@ def _classification(text: str) -> str:
         return "agent_framework_candidate"
     if _has_any(text, ["hermes-agent", "openclaw", "codex", "claude-code"]):
         return "personal_agent_platform_candidate"
-    return "needs_manual_research"
+    return "needs_controlled_research"
 
 
 def _recommendation(classification: str) -> str:
@@ -135,12 +135,12 @@ def _recommendation(classification: str) -> str:
         "agent_governance_candidate",
         "personal_agent_platform_candidate",
     }:
-        return "add_to_watch_registry_after_manual_source_review"
+        return "add_to_watch_registry_after_controlled_source_review"
     if classification == "agent_operator_console_candidate":
         return "watch_only_product_surface_signal"
     if classification == "vertical_product_not_core_agent":
         return "defer_not_core_agent_framework"
-    return "manual_research_before_watch_registry"
+    return "controlled_research_before_watch_registry"
 
 
 def _recommended_role(classification: str) -> str:
@@ -150,8 +150,8 @@ def _recommended_role(classification: str) -> str:
         "personal_agent_platform_candidate": "personal_agent_platform_candidate",
         "agent_operator_console_candidate": "operator_console_or_agent_ui_candidate",
         "vertical_product_not_core_agent": "vertical_product_signal_not_openclaw_replacement",
-        "needs_manual_research": "manual_research_required",
-    }.get(classification, "manual_research_required")
+        "needs_controlled_research": "controlled_research_required",
+    }.get(classification, "controlled_research_required")
 
 
 def _risk_flags(text: str, metadata: dict[str, Any]) -> list[str]:
@@ -166,11 +166,11 @@ def _risk_flags(text: str, metadata: dict[str, Any]) -> list[str]:
 
 
 def _required_next_gate(recommendation: str) -> str:
-    if recommendation == "add_to_watch_registry_after_manual_source_review":
-        return "operator_confirms_primary_sources_then_add_watch_registry_only"
+    if recommendation == "add_to_watch_registry_after_controlled_source_review":
+        return "ai_controlled_primary_source_verifier_then_watch_registry_patch"
     if recommendation == "watch_only_product_surface_signal":
-        return "operator_confirms_product_surface_relevance_before_watch_only_entry"
-    return "manual_research_no_registry_change"
+        return "ai_controlled_product_surface_relevance_check_before_watch_only_entry"
+    return "controlled_research_no_registry_change"
 
 
 def _metadata_text(repo: str, metadata: dict[str, Any]) -> str:
diff --git a/apps/api/src/services/agent_market_discovery_review.py b/apps/api/src/services/agent_market_discovery_review.py
index 3211b0df2..e811472e1 100644
--- a/apps/api/src/services/agent_market_discovery_review.py
+++ b/apps/api/src/services/agent_market_discovery_review.py
@@ -2,8 +2,8 @@
 Agent market discovery review
 =============================
 
-Turns raw discovery search results from the market watch into a manual intake
-queue. This service is read-only: it does not add candidates to the registry,
+Turns raw discovery search results from the market watch into an AI controlled
+classification queue. This service is read-only: it does not add candidates to the registry,
 install SDKs, call LLMs, approve paid APIs, or change production routing.
 """
 
@@ -100,7 +100,7 @@ def _candidate_drafts(
         decision = (
             "keep_existing_candidate_watch"
             if known
-            else "manual_primary_source_classification_required"
+            else "ai_controlled_primary_source_classification_required"
         )
         next_gate = (
             "use_existing_market_watch_candidate"
@@ -116,6 +116,7 @@ def _candidate_drafts(
                 "decision": decision,
                 "recommended_next_gate": next_gate,
                 "approval_boundary": {
+                    "controlled_classification_allowed": not known,
                     "approved_for_registry_addition": False,
                     "approved_for_sdk_install": False,
                     "approved_for_paid_api_calls": False,
@@ -128,7 +129,7 @@ def _candidate_drafts(
 
 
 def _summary(watch_report: dict[str, Any], drafts: list[dict[str, Any]]) -> dict[str, int]:
-    manual = [
+    controlled = [
         draft
         for draft in drafts
         if draft["status"] == "needs_primary_source_classification"
@@ -143,9 +144,9 @@ def _summary(watch_report: dict[str, Any], drafts: list[dict[str, Any]]) -> dict
         "already_watched_or_registered": sum(
             1 for draft in drafts if draft["status"] == "already_watched_or_registered"
         ),
-        "manual_classification_required": len(manual),
-        "new_manual_classification_required": sum(
-            1 for draft in manual if draft["new_since_previous_review"]
+        "controlled_classification_queue": len(controlled),
+        "new_controlled_classification_queue": sum(
+            1 for draft in controlled if draft["new_since_previous_review"]
         ),
         "source_failures": sum(
             1
@@ -209,7 +210,7 @@ def _recommended_actions(*, known: bool) -> list[str]:
     return [
         "verify_official_or_primary_sources",
         "classify_role_against_awoooi_agent_taxonomy",
-        "add_to_watch_registry_only_after_manual_review",
+        "queue_controlled_watch_registry_patch_after_primary_source_verifier",
         "do_not_install_sdk_or_call_provider",
         "do_not_enter_replacement_replay_before_market_scorecard",
     ]
diff --git a/apps/api/src/services/agent_market_governance_snapshot.py b/apps/api/src/services/agent_market_governance_snapshot.py
index 13738f970..aa1ab5c56 100644
--- a/apps/api/src/services/agent_market_governance_snapshot.py
+++ b/apps/api/src/services/agent_market_governance_snapshot.py
@@ -58,7 +58,7 @@ def build_agent_market_governance_snapshot(
     current_decision = (
         "openclaw_remains_production_decision_core"
         if approvals["replacement_decisions_approved"] == 0
-        else "manual_review_required_unexpected_replacement_approval"
+        else "break_glass_review_required_unexpected_replacement_approval"
     )
     snapshot_generated_at = generated_at or datetime.now(timezone.utc).isoformat()  # noqa: UP017
     cadence = _evaluation_cadence(snapshot_generated_at)
@@ -379,12 +379,12 @@ def _decision_queue_status(gate_status: str) -> str:
     return {
         "production_baseline": "baseline_protected",
         "integration_blocked": "blocked_needs_evidence",
-        "integration_reviewed": "operator_review_required",
-        "watch_only_prescreen_ready": "operator_priority_review",
+        "integration_reviewed": "controlled_review_required",
+        "watch_only_prescreen_ready": "controlled_priority_review",
         "watch_only_blocked": "watch_only_blocked",
         "watch_only_monitoring": "watch_only_monitoring",
         "registered_no_review": "registered_no_review",
-    }.get(gate_status, "operator_review_required")
+    }.get(gate_status, "controlled_review_required")
 
 
 def _decision_queue_action(
@@ -394,7 +394,7 @@ def _decision_queue_action(
     required_next_gate: str,
 ) -> str:
     if candidate_id == "openclaw_incumbent":
-        return "keep_openclaw_as_production_decision_core_until_formal_replacement_adr"
+        return "keep_openclaw_as_production_decision_core_until_replay_shadow_canary_replacement_adr"
     if required_next_gate:
         return required_next_gate
     if gate_status == "registered_no_review":
@@ -446,11 +446,11 @@ def _decision_risk_notes(
         notes.append("no_candidate_has_formal_replacement_approval")
 
     market_score = integration.get("market_score") or {}
-    notes.extend(str(value) for value in market_score.get("risks") or [])
+    notes.extend(_controlled_label(value) for value in market_score.get("risks") or [])
 
     classification = promotion.get("classification") or {}
-    notes.extend(str(value) for value in classification.get("risk_flags") or [])
-    notes.extend(str(value) for value in operator_blockers)
+    notes.extend(_controlled_label(value) for value in classification.get("risk_flags") or [])
+    notes.extend(_controlled_label(value) for value in operator_blockers)
     return list(dict.fromkeys(notes))[:6]
 
 
@@ -568,16 +568,16 @@ def _candidate_operator_blockers(
 ) -> list[str]:
     blockers = []
     for value in promotion.get("blockers") or []:
-        blockers.append(str(value))
+        blockers.append(_controlled_label(value))
     for value in integration.get("unblock_conditions") or []:
-        blockers.append(str(value))
+        blockers.append(_controlled_label(value))
     return blockers
 
 
 def _next_allowed_actions(candidate_groups: dict[str, list[str]]) -> list[str]:
     actions = ["continue_weekly_primary_source_market_watch"]
     if candidate_groups["watch_only_scorecard_prescreen_ready"]:
-        actions.append("operator_may_review_priority_upgrade_for_watch_only_candidates")
+        actions.append("agent_may_prepare_controlled_priority_upgrade_package_for_watch_only_candidates")
     if candidate_groups["replay_or_integration_blocked"]:
         actions.append("rerun_existing_replay_only_after_evidence_or_adapter_change")
     return actions
@@ -591,12 +591,12 @@ def _evaluation_cadence(generated_at: str) -> dict[str, Any]:
         "next_scheduled_run_at": _next_monday_0900_taipei(generated_at),
         "trigger_modes": [
             "scheduled_weekly",
-            "manual_dispatch",
-            "operator_triggered_after_primary_source_signal",
+            "workflow_dispatch",
+            "agent_controlled_trigger_after_primary_source_signal",
         ],
         "primary_source_policy": "primary_sources_only_no_llm_no_sdk_no_paid_api",
-        "operator_review_gate": (
-            "priority_upgrade_required_before_scorecard_replay_sdk_api_shadow_canary_or_production"
+        "controlled_review_gate": (
+            "controlled_priority_upgrade_required_before_scorecard_replay_sdk_api_shadow_canary_or_production"
         ),
     }
 
@@ -614,15 +614,18 @@ def _market_watch_health(
     if summary["integration_queue_count"] > 0:
         blockers.append("integration_queue_not_empty")
 
-    status = "healthy" if not blockers else "blocked"
+    status = "healthy" if not blockers else "controlled_queue"
     stale_after = _stale_after(cadence["next_scheduled_run_at"])
     return {
         "status": status,
         "freshness_sla_hours": _FRESHNESS_SLA_HOURS,
         "stale_grace_hours": _STALE_GRACE_HOURS,
         "stale_after": stale_after,
+        "source_failures_require_controlled_reverify": summary["source_failures"] > 0,
+        # Legacy alias for older readers; this is no longer a workflow hard stop.
         "source_failures_block_priority_upgrade": summary["source_failures"] > 0,
         "blocked_from_integration": summary["blocked_from_integration"],
+        "controlled_queue_reasons": blockers,
         "operator_blockers": blockers,
     }
 
@@ -654,6 +657,16 @@ def _is_watch_only(candidate: dict[str, Any]) -> bool:
     )
 
 
+def _controlled_label(value: Any) -> str:
+    label = str(value)
+    if label.endswith("_by_operator"):
+        return f"{label.removesuffix('_by_operator')}_by_ai_controlled_verifier"
+    if label.startswith("operator_confirms_"):
+        subject = label.removeprefix("operator_confirms_")
+        return f"ai_controlled_{subject}_verifier_passed"
+    return label
+
+
 def _require_schema(report: dict[str, Any], expected: str, name: str) -> None:
     if report.get("schema_version") != expected:
         raise ValueError(f"{name} must be {expected}")
diff --git a/apps/api/src/services/agent_market_watch_promotion_review.py b/apps/api/src/services/agent_market_watch_promotion_review.py
index 106d334e1..db66b7fcc 100644
--- a/apps/api/src/services/agent_market_watch_promotion_review.py
+++ b/apps/api/src/services/agent_market_watch_promotion_review.py
@@ -109,7 +109,7 @@ def _review_watch_only_candidate(
         and classification_recommended
     )
     decision = (
-        "eligible_for_operator_priority_review_before_market_scorecard"
+        "eligible_for_controlled_priority_review_before_market_scorecard"
         if eligible_for_scorecard
         else "remain_watch_only_until_evidence_gap_resolved"
     )
@@ -147,7 +147,7 @@ def _review_watch_only_candidate(
         "approved_for_shadow_or_canary": False,
         "blockers": blockers,
         "required_next_gate": (
-            "operator_priority_upgrade_then_market_scorecard_prescreen"
+            "controlled_priority_upgrade_then_market_scorecard_prescreen"
             if eligible_for_scorecard
             else "continue_watch_only_until_primary_source_evidence_is_sufficient"
         ),
diff --git a/apps/api/tests/test_agent_market_discovery_review.py b/apps/api/tests/test_agent_market_discovery_review.py
index 25c11df1d..f261ec513 100644
--- a/apps/api/tests/test_agent_market_discovery_review.py
+++ b/apps/api/tests/test_agent_market_discovery_review.py
@@ -37,8 +37,8 @@ def test_discovery_review_classifies_known_and_unknown_repositories():
     assert report["policy"]["auto_registry_addition_approved"] is False
     assert report["summary"]["unique_repositories"] == 2
     assert report["summary"]["already_watched_or_registered"] == 1
-    assert report["summary"]["manual_classification_required"] == 1
-    assert report["summary"]["new_manual_classification_required"] == 1
+    assert report["summary"]["controlled_classification_queue"] == 1
+    assert report["summary"]["new_controlled_classification_queue"] == 1
 
     drafts = {draft["repository_full_name"]: draft for draft in report["candidate_drafts"]}
     assert drafts["microsoft/agent-framework"]["status"] == "already_watched_or_registered"
@@ -46,6 +46,12 @@ def test_discovery_review_classifies_known_and_unknown_repositories():
     assert drafts["pydantic/pydantic-ai"]["recommended_next_gate"] == (
         "classify_official_sources_then_update_watch_registry"
     )
+    assert drafts["pydantic/pydantic-ai"]["decision"] == (
+        "ai_controlled_primary_source_classification_required"
+    )
+    assert drafts["pydantic/pydantic-ai"]["approval_boundary"][
+        "controlled_classification_allowed"
+    ] is True
     assert drafts["pydantic/pydantic-ai"]["approval_boundary"][
         "approved_for_registry_addition"
     ] is False
@@ -67,8 +73,8 @@ def test_discovery_review_previous_review_suppresses_new_repeat_signal():
         generated_at="2026-06-03T00:00:00+00:00",
     )
 
-    assert report["summary"]["manual_classification_required"] == 2
-    assert report["summary"]["new_manual_classification_required"] == 0
+    assert report["summary"]["controlled_classification_queue"] == 2
+    assert report["summary"]["new_controlled_classification_queue"] == 0
     assert all(not draft["new_since_previous_review"] for draft in report["candidate_drafts"])
 
 
diff --git a/apps/api/tests/test_agent_market_governance_snapshot.py b/apps/api/tests/test_agent_market_governance_snapshot.py
index daa643c03..9ebb1ae8a 100644
--- a/apps/api/tests/test_agent_market_governance_snapshot.py
+++ b/apps/api/tests/test_agent_market_governance_snapshot.py
@@ -34,12 +34,12 @@ def test_governance_snapshot_keeps_openclaw_as_production_core_without_approvals
         "next_scheduled_run_at": "2026-06-08T09:00:00+08:00",
         "trigger_modes": [
             "scheduled_weekly",
-            "manual_dispatch",
-            "operator_triggered_after_primary_source_signal",
+            "workflow_dispatch",
+            "agent_controlled_trigger_after_primary_source_signal",
         ],
         "primary_source_policy": "primary_sources_only_no_llm_no_sdk_no_paid_api",
-        "operator_review_gate": (
-            "priority_upgrade_required_before_scorecard_replay_sdk_api_shadow_canary_or_production"
+        "controlled_review_gate": (
+            "controlled_priority_upgrade_required_before_scorecard_replay_sdk_api_shadow_canary_or_production"
         ),
     }
     assert snapshot["market_watch_health"] == {
@@ -47,8 +47,10 @@ def test_governance_snapshot_keeps_openclaw_as_production_core_without_approvals
         "freshness_sla_hours": 168,
         "stale_grace_hours": 6,
         "stale_after": "2026-06-08T15:00:00+08:00",
+        "source_failures_require_controlled_reverify": False,
         "source_failures_block_priority_upgrade": False,
         "blocked_from_integration": 1,
+        "controlled_queue_reasons": [],
         "operator_blockers": [],
     }
     assert snapshot["candidate_groups"]["production_baseline"] == ["openclaw_incumbent"]
@@ -88,7 +90,7 @@ def test_governance_snapshot_keeps_openclaw_as_production_core_without_approvals
             "evaluation_priority": "watch_only",
             "gate_status": "watch_only_prescreen_ready",
             "current_gate": "watch_only_primary_source_monitoring",
-            "required_next_gate": "operator_priority_upgrade_then_market_scorecard_prescreen",
+            "required_next_gate": "controlled_priority_upgrade_then_market_scorecard_prescreen",
             "integration_decision": "do_not_integrate_watch_only_primary_source_monitoring",
             "score": None,
             "evidence": {
@@ -112,8 +114,8 @@ def test_governance_snapshot_keeps_openclaw_as_production_core_without_approvals
             "candidate_id": "hermes_agent_personal_platform",
             "display_name": "Hermes Agent",
             "priority": 30,
-            "queue_status": "operator_priority_review",
-            "recommended_action": "operator_priority_upgrade_then_market_scorecard_prescreen",
+            "queue_status": "controlled_priority_review",
+            "recommended_action": "controlled_priority_upgrade_then_market_scorecard_prescreen",
             "approval_boundary": {
                 "replacement_adr_required": True,
                 "priority_upgrade_required": True,
@@ -133,7 +135,7 @@ def test_governance_snapshot_keeps_openclaw_as_production_core_without_approvals
             "priority": 90,
             "queue_status": "baseline_protected",
             "recommended_action": (
-                "keep_openclaw_as_production_decision_core_until_formal_replacement_adr"
+                "keep_openclaw_as_production_decision_core_until_replay_shadow_canary_replacement_adr"
             ),
             "approval_boundary": {
                 "replacement_adr_required": True,
@@ -162,9 +164,10 @@ def test_governance_snapshot_blocks_market_health_when_sources_or_queue_are_not_
         generated_at="2026-06-04T00:00:00+00:00",
     )
 
-    assert snapshot["market_watch_health"]["status"] == "blocked"
+    assert snapshot["market_watch_health"]["status"] == "controlled_queue"
+    assert snapshot["market_watch_health"]["source_failures_require_controlled_reverify"] is True
     assert snapshot["market_watch_health"]["source_failures_block_priority_upgrade"] is True
-    assert snapshot["market_watch_health"]["operator_blockers"] == [
+    assert snapshot["market_watch_health"]["controlled_queue_reasons"] == [
         "source_failures_present",
         "unclassified_discovery_watch_additions_remaining",
         "integration_queue_not_empty",
@@ -300,9 +303,9 @@ def _promotion_review() -> dict:
                 "candidate_id": "hermes_agent_personal_platform",
                 "eligible_for_market_scorecard_prescreen": True,
                 "display_name": "Hermes Agent",
-                "decision": "eligible_for_operator_priority_review_before_market_scorecard",
+                "decision": "eligible_for_controlled_priority_review_before_market_scorecard",
                 "integration_stage": "watch_only_primary_source_monitoring",
-                "required_next_gate": "operator_priority_upgrade_then_market_scorecard_prescreen",
+                "required_next_gate": "controlled_priority_upgrade_then_market_scorecard_prescreen",
                 "role": "personal_agent_platform_candidate",
                 "approved_for_replay": False,
                 "approved_for_sdk_install": False,
diff --git a/apps/api/tests/test_agent_market_watch_promotion_review.py b/apps/api/tests/test_agent_market_watch_promotion_review.py
index c23657262..e3d4a5460 100644
--- a/apps/api/tests/test_agent_market_watch_promotion_review.py
+++ b/apps/api/tests/test_agent_market_watch_promotion_review.py
@@ -23,7 +23,7 @@ def test_watch_promotion_review_allows_only_scorecard_prescreen_readiness():
     assert review["eligible_for_market_scorecard_prescreen"] is True
     assert review["approved_for_replay"] is False
     assert review["required_next_gate"] == (
-        "operator_priority_upgrade_then_market_scorecard_prescreen"
+        "controlled_priority_upgrade_then_market_scorecard_prescreen"
     )
 
 
@@ -145,7 +145,7 @@ def _classification() -> dict:
                 "html_url": "https://github.com/NousResearch/hermes-agent",
                 "homepage": "https://hermes-agent.nousresearch.com",
                 "classification": "personal_agent_platform_candidate",
-                "recommendation": "add_to_watch_registry_after_manual_source_review",
+                "recommendation": "add_to_watch_registry_after_controlled_source_review",
                 "watch_addition_recommended": True,
                 "risk_flags": ["requires_dependency_boundary_review"],
             }
diff --git a/apps/web/messages/en.json b/apps/web/messages/en.json
index f6658202d..1c08ff0e6 100644
--- a/apps/web/messages/en.json
+++ b/apps/web/messages/en.json
@@ -3561,19 +3561,21 @@
       "health": {
         "title": "監測健康",
         "status": "狀態",
-        "statuses": {
-          "healthy": "Healthy",
-          "blocked": "Blocked"
-        },
+	        "statuses": {
+	          "healthy": "Healthy",
+	          "controlled_queue": "受控隊列",
+	          "blocked": "Blocked"
+	        },
         "freshnessSla": "新鮮度 SLA",
         "slaValue": "{slaHours}h + {graceHours}h",
         "staleAfter": "過期時間",
         "priorityGate": "升級關卡",
         "blockedIntegrations": "已擋下整合",
-        "blockers": "阻擋",
-        "blocked": "已阻擋",
-        "clear": "通過",
-        "noBlockers": "無 operator 阻擋"
+	        "blockers": "阻擋",
+	        "blocked": "已阻擋",
+	        "controlledReverify": "AI 受控重驗",
+	        "clear": "通過",
+	        "noBlockers": "無受控阻擋"
       },
       "cadence": {
         "title": "定期評估",
@@ -3596,8 +3598,10 @@
         "statuses": {
           "baseline_protected": "基準受保護",
           "blocked_needs_evidence": "需要證據",
-          "operator_review_required": "需要受控審查",
-          "operator_priority_review": "優先級受控審查",
+	          "controlled_review_required": "需要受控審查",
+	          "controlled_priority_review": "優先級受控審查",
+	          "operator_review_required": "需要受控審查",
+	          "operator_priority_review": "優先級受控審查",
           "watch_only_blocked": "觀察已阻擋",
           "watch_only_monitoring": "觀察中",
           "registered_no_review": "尚未審查"
diff --git a/apps/web/messages/zh-TW.json b/apps/web/messages/zh-TW.json
index f6658202d..1c08ff0e6 100644
--- a/apps/web/messages/zh-TW.json
+++ b/apps/web/messages/zh-TW.json
@@ -3561,19 +3561,21 @@
       "health": {
         "title": "監測健康",
         "status": "狀態",
-        "statuses": {
-          "healthy": "Healthy",
-          "blocked": "Blocked"
-        },
+	        "statuses": {
+	          "healthy": "Healthy",
+	          "controlled_queue": "受控隊列",
+	          "blocked": "Blocked"
+	        },
         "freshnessSla": "新鮮度 SLA",
         "slaValue": "{slaHours}h + {graceHours}h",
         "staleAfter": "過期時間",
         "priorityGate": "升級關卡",
         "blockedIntegrations": "已擋下整合",
-        "blockers": "阻擋",
-        "blocked": "已阻擋",
-        "clear": "通過",
-        "noBlockers": "無 operator 阻擋"
+	        "blockers": "阻擋",
+	        "blocked": "已阻擋",
+	        "controlledReverify": "AI 受控重驗",
+	        "clear": "通過",
+	        "noBlockers": "無受控阻擋"
       },
       "cadence": {
         "title": "定期評估",
@@ -3596,8 +3598,10 @@
         "statuses": {
           "baseline_protected": "基準受保護",
           "blocked_needs_evidence": "需要證據",
-          "operator_review_required": "需要受控審查",
-          "operator_priority_review": "優先級受控審查",
+	          "controlled_review_required": "需要受控審查",
+	          "controlled_priority_review": "優先級受控審查",
+	          "operator_review_required": "需要受控審查",
+	          "operator_priority_review": "優先級受控審查",
           "watch_only_blocked": "觀察已阻擋",
           "watch_only_monitoring": "觀察中",
           "registered_no_review": "尚未審查"
diff --git a/apps/web/src/app/[locale]/governance/tabs/agent-market-tab.tsx b/apps/web/src/app/[locale]/governance/tabs/agent-market-tab.tsx
index eeb7e6088..ae170cc2a 100644
--- a/apps/web/src/app/[locale]/governance/tabs/agent-market-tab.tsx
+++ b/apps/web/src/app/[locale]/governance/tabs/agent-market-tab.tsx
@@ -1237,6 +1237,8 @@ export function AgentMarketTab() {
     .length
   const watchHealth = snapshot.market_watch_health
   const watchHealthHealthy = watchHealth.status === 'healthy'
+  const watchHealthReasons = watchHealth.controlled_queue_reasons ?? watchHealth.operator_blockers ?? []
+  const controlledReviewGate = snapshot.evaluation_cadence.controlled_review_gate ?? snapshot.evaluation_cadence.operator_review_gate
 
   return (
     <div style={{ padding: 20, display: 'flex', flexDirection: 'column', gap: 16 }}>
@@ -1867,15 +1869,15 @@ export function AgentMarketTab() {
               {formatDateTime(watchHealth.stale_after)}
             </DetailRow>
             <DetailRow label={t('health.priorityGate')}>
-              {watchHealth.source_failures_block_priority_upgrade ? t('health.blocked') : t('health.clear')}
+              {watchHealth.source_failures_require_controlled_reverify ? t('health.controlledReverify') : t('health.clear')}
             </DetailRow>
             <DetailRow label={t('health.blockedIntegrations')}>
               {watchHealth.blocked_from_integration}
             </DetailRow>
             <DetailRow label={t('health.blockers')}>
               <div style={{ display: 'flex', flexWrap: 'wrap', gap: 6, minWidth: 0 }}>
-                {watchHealth.operator_blockers.length > 0 ? (
-                  watchHealth.operator_blockers.map(blocker => (
+                {watchHealthReasons.length > 0 ? (
+                  watchHealthReasons.map(blocker => (
                     <CandidatePill key={blocker} value={blocker} muted />
                   ))
                 ) : (
@@ -1913,7 +1915,7 @@ export function AgentMarketTab() {
               <CandidatePill value={snapshot.evaluation_cadence.primary_source_policy} />
             </DetailRow>
             <DetailRow label={t('cadence.reviewGate')}>
-              <CandidatePill value={snapshot.evaluation_cadence.operator_review_gate} muted />
+              <CandidatePill value={controlledReviewGate} muted />
             </DetailRow>
             <DetailRow label={t('cadence.triggerModes')}>
               <div style={{ display: 'flex', flexWrap: 'wrap', gap: 6, minWidth: 0 }}>
diff --git a/apps/web/src/lib/api-client.ts b/apps/web/src/lib/api-client.ts
index fae6345e3..89fb1332e 100644
--- a/apps/web/src/lib/api-client.ts
+++ b/apps/web/src/lib/api-client.ts
@@ -1538,18 +1538,21 @@ export interface AgentMarketGovernanceSnapshot {
     schedule: string
     timezone: 'Asia/Taipei'
     next_scheduled_run_at: string
-    trigger_modes: string[]
-    primary_source_policy: string
-    operator_review_gate: string
-  }
-  market_watch_health: {
-    status: 'healthy' | 'blocked'
-    freshness_sla_hours: 168
-    stale_grace_hours: 6
-    stale_after: string
-    source_failures_block_priority_upgrade: boolean
-    blocked_from_integration: number
-    operator_blockers: string[]
+	    trigger_modes: string[]
+	    primary_source_policy: string
+	    controlled_review_gate: string
+	    operator_review_gate?: string
+	  }
+	  market_watch_health: {
+	    status: 'healthy' | 'controlled_queue' | 'blocked'
+	    freshness_sla_hours: 168
+	    stale_grace_hours: 6
+	    stale_after: string
+	    source_failures_require_controlled_reverify: boolean
+	    source_failures_block_priority_upgrade?: boolean
+	    blocked_from_integration: number
+	    controlled_queue_reasons: string[]
+	    operator_blockers?: string[]
   }
   summary: {
     candidate_count: number
@@ -1613,10 +1616,12 @@ export interface AgentMarketGovernanceSnapshot {
     display_name: string
     priority: number
     queue_status:
-      | 'baseline_protected'
-      | 'blocked_needs_evidence'
-      | 'operator_review_required'
-      | 'operator_priority_review'
+	      | 'baseline_protected'
+	      | 'blocked_needs_evidence'
+	      | 'controlled_review_required'
+	      | 'controlled_priority_review'
+	      | 'operator_review_required'
+	      | 'operator_priority_review'
       | 'watch_only_blocked'
       | 'watch_only_monitoring'
       | 'registered_no_review'
diff --git a/docs/LOGBOOK.md b/docs/LOGBOOK.md
index 7c7c35d8f..02d38bf3b 100644
--- a/docs/LOGBOOK.md
+++ b/docs/LOGBOOK.md
@@ -1,3 +1,25 @@
+## 2026-06-28 — 09:35 Agent Market discovery guard 轉 AI controlled queue
+
+**背景**：統帥要求非事故級 hard gate / guard 全部打開並快速推進實作。本段針對 `agent-market-watch` 中舊 `manual_classification_required`、`needs_manual_research`、promotion review hard fail 與 operator priority gate，改成可審計的 AI controlled classification / priority queue。
+
+**完成內容**：
+- `.gitea/workflows/agent-market-watch.yaml` 將 discovery / classification / summary / Telegram 文案改為 AI 受控隊列；classification step 不再因沒有 `new_manual` 候選而跳過，會固定產生 controlled queue artifact，避免後段缺 artifact 變成硬阻塞。
+- `apps/api/src/services/agent_market_discovery_review.py`、`agent_market_discovery_classifier.py`、`agent_market_watch_promotion_review.py`、`agent_market_governance_snapshot.py` 移除非 critical manual/operator 預設終局，改為 `controlled_classification_queue`、`needs_controlled_research`、`controlled_priority_review`、`controlled_queue`。
+- `docs/schemas/agent_market_discovery_review_v1.schema.json`、`docs/schemas/agent_market_governance_snapshot_v1.schema.json` 與 `apps/web/src/lib/api-client.ts` 同步 controlled queue 欄位，前端 governance Agent Market tab 改讀 controlled reverify / controlled review gate。
+- `apps/web/messages/zh-TW.json`、`apps/web/messages/en.json` 補齊 Agent Market controlled queue / controlled review i18n，避免 UI 仍把非 critical 市場 watch path 說成 manual blocker。
+- 重新產生 `docs/evaluations/agent_market_*_2026-06-28.json` 四份 snapshot；本次 readback：`controlled_classification_queue=9`、`recommended_watch_additions=7`、`integration_queue_count=5`、`market_watch_health.status=controlled_queue`。
+
+**本地驗證結果**：
+- YAML / JSON parse：通過。
+- Agent Market pytest：`9 passed`。
+- i18n mirror：`zh-TW=14480`、`en=14480`、missing `0 / 0`。
+- `pnpm --dir apps/web typecheck`：通過。
+- `git diff --check`：通過。
+
+**仍保留的 break-glass 邊界**：
+- OpenClaw 仍維持 production decision core；替換前必須 replay / shadow / canary / ADR。
+- SDK install、API shadow / canary、production route、paid provider / cost route、external active security scan、secret value / credential URL / raw env、DB destructive / backup restore、force push / repo refs deletion 仍不得被本段 controlled queue 直接打開。
+
 ## 2026-06-28 — 08:45 110 runner 壓力事故 source / live fail-closed 收斂
 
 **背景**：統帥全面授權打開非事故級 gate，但 110 Gitea runner 反覆拉起 StockPlatform headless Chrome smoke，已造成 production host CPU / CI 壓力事故；runner 未搬遷 / 限流前不得直接重開。
diff --git a/docs/evaluations/agent_market_discovery_classification_2026-06-28.json b/docs/evaluations/agent_market_discovery_classification_2026-06-28.json
new file mode 100644
index 000000000..c8a832196
--- /dev/null
+++ b/docs/evaluations/agent_market_discovery_classification_2026-06-28.json
@@ -0,0 +1,431 @@
+{
+  "candidates": [
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_replay": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "approved_for_watch_registry_addition": false
+      },
+      "archived": false,
+      "classification": "needs_controlled_research",
+      "description": "The Frontend Stack for Agents & Generative UI. React, Angular, Mobile, Slack, and more.  Makers of the AG-UI Protocol",
+      "homepage": "https://docs.copilotkit.ai",
+      "html_url": "https://github.com/CopilotKit/CopilotKit",
+      "language": "TypeScript",
+      "pushed_at": "2026-06-27T06:40:27Z",
+      "recommendation": "controlled_research_before_watch_registry",
+      "recommended_role": "controlled_research_required",
+      "repository_full_name": "copilotkit/copilotkit",
+      "required_next_gate": "controlled_research_no_registry_change",
+      "risk_flags": [
+        "requires_dependency_boundary_review"
+      ],
+      "stargazers_count": 35572,
+      "topics": [
+        "agent",
+        "agent-native",
+        "agentic-ai",
+        "agents",
+        "ai",
+        "ai-agent",
+        "ai-assistant",
+        "assistant",
+        "assistant-chat-bots",
+        "copilot",
+        "copilot-chat",
+        "generative-ui",
+        "js",
+        "llm",
+        "nextjs",
+        "open-source",
+        "react",
+        "reactjs",
+        "ts",
+        "typescript"
+      ],
+      "watch_addition_recommended": false
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_replay": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "approved_for_watch_registry_addition": false
+      },
+      "archived": false,
+      "classification": "needs_controlled_research",
+      "description": "Open-source infrastructure for Computer-Use Agents. Sandboxes, SDKs, and benchmarks to train and evaluate AI agents that can control full desktops (macOS, Linux, Windows).",
+      "homepage": "https://cua.ai",
+      "html_url": "https://github.com/trycua/cua",
+      "language": "HTML",
+      "pushed_at": "2026-06-28T00:16:47Z",
+      "recommendation": "controlled_research_before_watch_registry",
+      "recommended_role": "controlled_research_required",
+      "repository_full_name": "trycua/cua",
+      "required_next_gate": "controlled_research_no_registry_change",
+      "risk_flags": [
+        "requires_dependency_boundary_review",
+        "requires_tool_execution_sandbox_review"
+      ],
+      "stargazers_count": 19131,
+      "topics": [
+        "agent",
+        "ai-agent",
+        "apple",
+        "computer-use",
+        "computer-use-agent",
+        "containerization",
+        "cua",
+        "desktop-automation",
+        "hacktoberfest",
+        "lume",
+        "macos",
+        "manus",
+        "operator",
+        "swift",
+        "virtualization",
+        "virtualization-framework",
+        "windows",
+        "windows-sandbox"
+      ],
+      "watch_addition_recommended": false
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_replay": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "approved_for_watch_registry_addition": false
+      },
+      "archived": false,
+      "classification": "agent_framework_candidate",
+      "description": "Eigent: The Open Source Cowork Desktop to Unlock Your Exceptional Productivity. Local and Free Alternative to Claude Cowork.",
+      "homepage": "https://www.eigent.ai",
+      "html_url": "https://github.com/eigent-ai/eigent",
+      "language": "TypeScript",
+      "pushed_at": "2026-06-27T10:16:01Z",
+      "recommendation": "add_to_watch_registry_after_controlled_source_review",
+      "recommended_role": "agent_framework_or_orchestrator_candidate",
+      "repository_full_name": "eigent-ai/eigent",
+      "required_next_gate": "ai_controlled_primary_source_verifier_then_watch_registry_patch",
+      "risk_flags": [
+        "requires_dependency_boundary_review",
+        "likely_requires_paid_provider_boundary_review"
+      ],
+      "stargazers_count": 14404,
+      "topics": [
+        "agent-framework",
+        "agent-skills",
+        "agentic-ai",
+        "agentic-workflow",
+        "claude-cowork",
+        "claude-cowork-alternative",
+        "claude-cowork-free",
+        "desktop-agent",
+        "multi-agent-systems"
+      ],
+      "watch_addition_recommended": true
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_replay": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "approved_for_watch_registry_addition": false
+      },
+      "archived": false,
+      "classification": "agent_governance_candidate",
+      "description": "Omnigent is an open-source AI agent framework and meta-harness: orchestrate Claude Code, Codex, Cursor, Pi, and custom agents — swap harnesses without rewriting, enforce policies and sandboxing, and collaborate in real time from any device.",
+      "homepage": "https://omnigent.ai",
+      "html_url": "https://github.com/omnigent-ai/omnigent",
+      "language": "Python",
+      "pushed_at": "2026-06-27T23:15:21Z",
+      "recommendation": "add_to_watch_registry_after_controlled_source_review",
+      "recommended_role": "agent_governance_policy_evaluator_candidate",
+      "repository_full_name": "omnigent-ai/omnigent",
+      "required_next_gate": "ai_controlled_primary_source_verifier_then_watch_registry_patch",
+      "risk_flags": [
+        "requires_dependency_boundary_review",
+        "likely_requires_paid_provider_boundary_review",
+        "requires_tool_execution_sandbox_review"
+      ],
+      "stargazers_count": 5132,
+      "topics": [
+        "agent-framework",
+        "agent-governance",
+        "agent-orchestration",
+        "agents",
+        "ai",
+        "ai-agent",
+        "ai-agents",
+        "claude-code",
+        "codex",
+        "coding-agents",
+        "developer-tools",
+        "llm",
+        "ml",
+        "multi-agent",
+        "python",
+        "sandbox"
+      ],
+      "watch_addition_recommended": true
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_replay": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "approved_for_watch_registry_addition": false
+      },
+      "archived": false,
+      "classification": "agent_framework_candidate",
+      "description": "AG2 (formerly AutoGen): The Open-Source AgentOS.Join us at: https://discord.gg/sNGSwQME3x",
+      "homepage": "https://ag2.ai",
+      "html_url": "https://github.com/ag2ai/ag2",
+      "language": "Python",
+      "pushed_at": "2026-06-27T22:37:56Z",
+      "recommendation": "add_to_watch_registry_after_controlled_source_review",
+      "recommended_role": "agent_framework_or_orchestrator_candidate",
+      "repository_full_name": "ag2ai/ag2",
+      "required_next_gate": "ai_controlled_primary_source_verifier_then_watch_registry_patch",
+      "risk_flags": [
+        "requires_dependency_boundary_review",
+        "requires_tool_execution_sandbox_review"
+      ],
+      "stargazers_count": 4714,
+      "topics": [
+        "a2a",
+        "ag2",
+        "agent-framework",
+        "agentic",
+        "agentic-ai",
+        "ai",
+        "ai-agents-framework",
+        "aiagents",
+        "genai",
+        "llm",
+        "llms",
+        "mcp",
+        "multi-agent",
+        "multi-agent-system",
+        "open-source",
+        "python"
+      ],
+      "watch_addition_recommended": true
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_replay": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "approved_for_watch_registry_addition": false
+      },
+      "archived": false,
+      "classification": "agent_framework_candidate",
+      "description": "AI 时代的伯克希尔：基于 Claude Code 的价值投资研究框架。巴菲特·芒格·段永平·李录四大师方法论 + 多Agent并行研究。| AI-era Berkshire: a value investing research framework   built on Claude Code. 4 masters' methodologies + multi-agent adversarial analysis.",
+      "homepage": "https://github.com/xbtlin/ai-berkshire#readme",
+      "html_url": "https://github.com/xbtlin/ai-berkshire",
+      "language": "Python",
+      "pushed_at": "2026-06-27T16:49:55Z",
+      "recommendation": "add_to_watch_registry_after_controlled_source_review",
+      "recommended_role": "agent_framework_or_orchestrator_candidate",
+      "repository_full_name": "xbtlin/ai-berkshire",
+      "required_next_gate": "ai_controlled_primary_source_verifier_then_watch_registry_patch",
+      "risk_flags": [
+        "requires_dependency_boundary_review",
+        "likely_requires_paid_provider_boundary_review",
+        "requires_tool_execution_sandbox_review"
+      ],
+      "stargazers_count": 4117,
+      "topics": [
+        "ai",
+        "ai-agent",
+        "anthropic",
+        "berkshire-hathaway",
+        "charlie-munger",
+        "china-stock",
+        "claude",
+        "claude-code",
+        "financial-analysis",
+        "fintech",
+        "fundamental-analysis",
+        "investment",
+        "investment-research",
+        "llm",
+        "mcp",
+        "portfolio-management",
+        "stock-analysis",
+        "stock-market",
+        "value-investing",
+        "warren-buffett"
+      ],
+      "watch_addition_recommended": true
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_replay": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "approved_for_watch_registry_addition": false
+      },
+      "archived": false,
+      "classification": "agent_framework_candidate",
+      "description": "A Go framework for building production agent systems with graph workflows, tools, memory, A2A, AG-UI, MCP, evaluation, and observability.",
+      "homepage": "https://trpc-group.github.io/trpc-agent-go/",
+      "html_url": "https://github.com/trpc-group/trpc-agent-go",
+      "language": "Go",
+      "pushed_at": "2026-06-27T03:22:33Z",
+      "recommendation": "add_to_watch_registry_after_controlled_source_review",
+      "recommended_role": "agent_framework_or_orchestrator_candidate",
+      "repository_full_name": "trpc-group/trpc-agent-go",
+      "required_next_gate": "ai_controlled_primary_source_verifier_then_watch_registry_patch",
+      "risk_flags": [
+        "requires_dependency_boundary_review",
+        "requires_tool_execution_sandbox_review"
+      ],
+      "stargazers_count": 1414,
+      "topics": [
+        "a2a",
+        "a2a-protocol",
+        "ag-ui",
+        "agent",
+        "agent-framework",
+        "ai",
+        "ai-agents",
+        "evaluation",
+        "go",
+        "golang",
+        "graph-workflows",
+        "llm",
+        "mcp",
+        "model-context-protocol",
+        "multi-agent",
+        "observability",
+        "opentelemetry",
+        "rag"
+      ],
+      "watch_addition_recommended": true
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_replay": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "approved_for_watch_registry_addition": false
+      },
+      "archived": false,
+      "classification": "personal_agent_platform_candidate",
+      "description": "Zero, your trustworthy AI teammate for real work. ",
+      "homepage": "https://vm0.ai",
+      "html_url": "https://github.com/vm0-ai/vm0",
+      "language": "TypeScript",
+      "pushed_at": "2026-06-28T00:39:40Z",
+      "recommendation": "add_to_watch_registry_after_controlled_source_review",
+      "recommended_role": "personal_agent_platform_candidate",
+      "repository_full_name": "vm0-ai/vm0",
+      "required_next_gate": "ai_controlled_primary_source_verifier_then_watch_registry_patch",
+      "risk_flags": [
+        "requires_dependency_boundary_review",
+        "likely_requires_paid_provider_boundary_review",
+        "requires_tool_execution_sandbox_review"
+      ],
+      "stargazers_count": 1133,
+      "topics": [
+        "agentic-workflow",
+        "ai-agent",
+        "ai-runtime",
+        "ai-sandbox",
+        "claude-code",
+        "dev-tools",
+        "sandbox"
+      ],
+      "watch_addition_recommended": true
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_replay": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "approved_for_watch_registry_addition": false
+      },
+      "archived": false,
+      "classification": "agent_framework_candidate",
+      "description": "CORAL is a robust, lightweight infrastructure for multi-agent autonomous self-evolution, built for autoresearch. Works with Claude Code, Codex, Cursor, OpenCode, Kiro, and more.",
+      "homepage": "https://docs.coralxyz.com",
+      "html_url": "https://github.com/Human-Agent-Society/CORAL",
+      "language": "Python",
+      "pushed_at": "2026-06-26T06:40:13Z",
+      "recommendation": "add_to_watch_registry_after_controlled_source_review",
+      "recommended_role": "agent_framework_or_orchestrator_candidate",
+      "repository_full_name": "human-agent-society/coral",
+      "required_next_gate": "ai_controlled_primary_source_verifier_then_watch_registry_patch",
+      "risk_flags": [
+        "requires_dependency_boundary_review",
+        "likely_requires_paid_provider_boundary_review"
+      ],
+      "stargazers_count": 756,
+      "topics": [
+        "agent-framework",
+        "agentic-ai",
+        "alpha-evolve",
+        "autonomous-agents",
+        "autoresearch",
+        "claude-code",
+        "code-generation",
+        "codex",
+        "coding-agent",
+        "evolutionary-algorithm",
+        "large-language-models",
+        "llm-agents",
+        "multi-agent",
+        "opencode",
+        "self-evolving",
+        "self-evolving-agents",
+        "swe-bench"
+      ],
+      "watch_addition_recommended": true
+    }
+  ],
+  "generated_at": "2026-06-28T00:46:51.907924+00:00",
+  "inputs": {
+    "discovery_review_generated_at": "2026-06-28T00:46:48.886379+00:00",
+    "metadata_source": "github_repository_api_summary"
+  },
+  "policy": {
+    "auto_watch_registry_addition_approved": false,
+    "paid_api_calls_approved": false,
+    "production_changes_approved": false,
+    "raw_external_pages_committed": false,
+    "replacement_decision_allowed": false,
+    "sdk_installation_approved": false,
+    "shadow_or_canary_approved": false
+  },
+  "schema_version": "agent_market_discovery_classification_v1",
+  "summary": {
+    "classification_counts": {
+      "agent_framework_candidate": 5,
+      "agent_governance_candidate": 1,
+      "needs_controlled_research": 2,
+      "personal_agent_platform_candidate": 1
+    },
+    "classified_repositories": 9,
+    "production_changes_approved": 0,
+    "recommendation_counts": {
+      "add_to_watch_registry_after_controlled_source_review": 7,
+      "controlled_research_before_watch_registry": 2
+    },
+    "recommended_watch_additions": 7,
+    "shadow_or_canary_approved": 0,
+    "watch_only_or_defer": 2
+  }
+}
diff --git a/docs/evaluations/agent_market_discovery_review_2026-06-28.json b/docs/evaluations/agent_market_discovery_review_2026-06-28.json
new file mode 100644
index 000000000..854723c89
--- /dev/null
+++ b/docs/evaluations/agent_market_discovery_review_2026-06-28.json
@@ -0,0 +1,286 @@
+{
+  "candidate_drafts": [
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_registry_addition": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "controlled_classification_allowed": true
+      },
+      "decision": "ai_controlled_primary_source_classification_required",
+      "html_url": "https://github.com/CopilotKit/CopilotKit",
+      "new_since_previous_review": false,
+      "recommended_actions": [
+        "verify_official_or_primary_sources",
+        "classify_role_against_awoooi_agent_taxonomy",
+        "queue_controlled_watch_registry_patch_after_primary_source_verifier",
+        "do_not_install_sdk_or_call_provider",
+        "do_not_enter_replacement_replay_before_market_scorecard"
+      ],
+      "recommended_next_gate": "classify_official_sources_then_update_watch_registry",
+      "repository_full_name": "copilotkit/copilotkit",
+      "seen_before": true,
+      "source_ids": [
+        "github_ai_agent_topic"
+      ],
+      "stargazers_count_max": 35522,
+      "status": "needs_primary_source_classification",
+      "updated_at_latest": "2026-06-26T03:36:59Z"
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_registry_addition": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "controlled_classification_allowed": true
+      },
+      "decision": "ai_controlled_primary_source_classification_required",
+      "html_url": "https://github.com/trycua/cua",
+      "new_since_previous_review": false,
+      "recommended_actions": [
+        "verify_official_or_primary_sources",
+        "classify_role_against_awoooi_agent_taxonomy",
+        "queue_controlled_watch_registry_patch_after_primary_source_verifier",
+        "do_not_install_sdk_or_call_provider",
+        "do_not_enter_replacement_replay_before_market_scorecard"
+      ],
+      "recommended_next_gate": "classify_official_sources_then_update_watch_registry",
+      "repository_full_name": "trycua/cua",
+      "seen_before": true,
+      "source_ids": [
+        "github_ai_agent_topic"
+      ],
+      "stargazers_count_max": 19004,
+      "status": "needs_primary_source_classification",
+      "updated_at_latest": "2026-06-26T03:40:04Z"
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_registry_addition": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "controlled_classification_allowed": true
+      },
+      "decision": "ai_controlled_primary_source_classification_required",
+      "html_url": "https://github.com/eigent-ai/eigent",
+      "new_since_previous_review": false,
+      "recommended_actions": [
+        "verify_official_or_primary_sources",
+        "classify_role_against_awoooi_agent_taxonomy",
+        "queue_controlled_watch_registry_patch_after_primary_source_verifier",
+        "do_not_install_sdk_or_call_provider",
+        "do_not_enter_replacement_replay_before_market_scorecard"
+      ],
+      "recommended_next_gate": "classify_official_sources_then_update_watch_registry",
+      "repository_full_name": "eigent-ai/eigent",
+      "seen_before": true,
+      "source_ids": [
+        "github_agent_framework_topic"
+      ],
+      "stargazers_count_max": 14394,
+      "status": "needs_primary_source_classification",
+      "updated_at_latest": "2026-06-26T03:38:22Z"
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_registry_addition": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "controlled_classification_allowed": true
+      },
+      "decision": "ai_controlled_primary_source_classification_required",
+      "html_url": "https://github.com/omnigent-ai/omnigent",
+      "new_since_previous_review": false,
+      "recommended_actions": [
+        "verify_official_or_primary_sources",
+        "classify_role_against_awoooi_agent_taxonomy",
+        "queue_controlled_watch_registry_patch_after_primary_source_verifier",
+        "do_not_install_sdk_or_call_provider",
+        "do_not_enter_replacement_replay_before_market_scorecard"
+      ],
+      "recommended_next_gate": "classify_official_sources_then_update_watch_registry",
+      "repository_full_name": "omnigent-ai/omnigent",
+      "seen_before": true,
+      "source_ids": [
+        "github_ai_agent_topic",
+        "github_agent_framework_topic"
+      ],
+      "stargazers_count_max": 4924,
+      "status": "needs_primary_source_classification",
+      "updated_at_latest": "2026-06-26T03:37:54Z"
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_registry_addition": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "controlled_classification_allowed": true
+      },
+      "decision": "ai_controlled_primary_source_classification_required",
+      "html_url": "https://github.com/ag2ai/ag2",
+      "new_since_previous_review": false,
+      "recommended_actions": [
+        "verify_official_or_primary_sources",
+        "classify_role_against_awoooi_agent_taxonomy",
+        "queue_controlled_watch_registry_patch_after_primary_source_verifier",
+        "do_not_install_sdk_or_call_provider",
+        "do_not_enter_replacement_replay_before_market_scorecard"
+      ],
+      "recommended_next_gate": "classify_official_sources_then_update_watch_registry",
+      "repository_full_name": "ag2ai/ag2",
+      "seen_before": true,
+      "source_ids": [
+        "github_agent_framework_topic"
+      ],
+      "stargazers_count_max": 4710,
+      "status": "needs_primary_source_classification",
+      "updated_at_latest": "2026-06-26T03:29:07Z"
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_registry_addition": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "controlled_classification_allowed": true
+      },
+      "decision": "ai_controlled_primary_source_classification_required",
+      "html_url": "https://github.com/xbtlin/ai-berkshire",
+      "new_since_previous_review": false,
+      "recommended_actions": [
+        "verify_official_or_primary_sources",
+        "classify_role_against_awoooi_agent_taxonomy",
+        "queue_controlled_watch_registry_patch_after_primary_source_verifier",
+        "do_not_install_sdk_or_call_provider",
+        "do_not_enter_replacement_replay_before_market_scorecard"
+      ],
+      "recommended_next_gate": "classify_official_sources_then_update_watch_registry",
+      "repository_full_name": "xbtlin/ai-berkshire",
+      "seen_before": true,
+      "source_ids": [
+        "github_ai_agent_topic"
+      ],
+      "stargazers_count_max": 2172,
+      "status": "needs_primary_source_classification",
+      "updated_at_latest": "2026-06-26T03:41:15Z"
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_registry_addition": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "controlled_classification_allowed": true
+      },
+      "decision": "ai_controlled_primary_source_classification_required",
+      "html_url": "https://github.com/trpc-group/trpc-agent-go",
+      "new_since_previous_review": false,
+      "recommended_actions": [
+        "verify_official_or_primary_sources",
+        "classify_role_against_awoooi_agent_taxonomy",
+        "queue_controlled_watch_registry_patch_after_primary_source_verifier",
+        "do_not_install_sdk_or_call_provider",
+        "do_not_enter_replacement_replay_before_market_scorecard"
+      ],
+      "recommended_next_gate": "classify_official_sources_then_update_watch_registry",
+      "repository_full_name": "trpc-group/trpc-agent-go",
+      "seen_before": true,
+      "source_ids": [
+        "github_agent_framework_topic"
+      ],
+      "stargazers_count_max": 1408,
+      "status": "needs_primary_source_classification",
+      "updated_at_latest": "2026-06-26T01:55:30Z"
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_registry_addition": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "controlled_classification_allowed": true
+      },
+      "decision": "ai_controlled_primary_source_classification_required",
+      "html_url": "https://github.com/vm0-ai/vm0",
+      "new_since_previous_review": false,
+      "recommended_actions": [
+        "verify_official_or_primary_sources",
+        "classify_role_against_awoooi_agent_taxonomy",
+        "queue_controlled_watch_registry_patch_after_primary_source_verifier",
+        "do_not_install_sdk_or_call_provider",
+        "do_not_enter_replacement_replay_before_market_scorecard"
+      ],
+      "recommended_next_gate": "classify_official_sources_then_update_watch_registry",
+      "repository_full_name": "vm0-ai/vm0",
+      "seen_before": true,
+      "source_ids": [
+        "github_ai_agent_topic"
+      ],
+      "stargazers_count_max": 1131,
+      "status": "needs_primary_source_classification",
+      "updated_at_latest": "2026-06-26T03:32:54Z"
+    },
+    {
+      "approval_boundary": {
+        "approved_for_paid_api_calls": false,
+        "approved_for_registry_addition": false,
+        "approved_for_sdk_install": false,
+        "approved_for_shadow_or_canary": false,
+        "controlled_classification_allowed": true
+      },
+      "decision": "ai_controlled_primary_source_classification_required",
+      "html_url": "https://github.com/Human-Agent-Society/CORAL",
+      "new_since_previous_review": false,
+      "recommended_actions": [
+        "verify_official_or_primary_sources",
+        "classify_role_against_awoooi_agent_taxonomy",
+        "queue_controlled_watch_registry_patch_after_primary_source_verifier",
+        "do_not_install_sdk_or_call_provider",
+        "do_not_enter_replacement_replay_before_market_scorecard"
+      ],
+      "recommended_next_gate": "classify_official_sources_then_update_watch_registry",
+      "repository_full_name": "human-agent-society/coral",
+      "seen_before": true,
+      "source_ids": [
+        "github_agent_framework_topic"
+      ],
+      "stargazers_count_max": 754,
+      "status": "needs_primary_source_classification",
+      "updated_at_latest": "2026-06-26T03:05:29Z"
+    }
+  ],
+  "generated_at": "2026-06-28T00:46:48.886379+00:00",
+  "inputs": {
+    "candidate_registry_schema_version": "agent_replacement_candidates_v1",
+    "previous_review_generated_at": "2026-06-26T03:41:38.325050+00:00",
+    "source_registry_schema_version": "agent_market_watch_sources_v1",
+    "watch_report_generated_at": "2026-06-26T03:41:17.266599+00:00",
+    "watch_report_mode": "live"
+  },
+  "policy": {
+    "auto_registry_addition_approved": false,
+    "paid_api_calls_approved": false,
+    "production_changes_approved": false,
+    "replacement_decision_allowed": false,
+    "sdk_installation_approved": false,
+    "shadow_or_canary_approved": false
+  },
+  "schema_version": "agent_market_discovery_review_v1",
+  "summary": {
+    "already_watched_or_registered": 0,
+    "auto_registry_additions_approved": 0,
+    "controlled_classification_queue": 9,
+    "discovered_items": 10,
+    "discovery_sources": 2,
+    "new_controlled_classification_queue": 0,
+    "production_changes_approved": 0,
+    "shadow_or_canary_approved": 0,
+    "source_failures": 0,
+    "unique_repositories": 9
+  }
+}
diff --git a/docs/evaluations/agent_market_governance_snapshot_2026-06-28.json b/docs/evaluations/agent_market_governance_snapshot_2026-06-28.json
new file mode 100644
index 000000000..2b856926b
--- /dev/null
+++ b/docs/evaluations/agent_market_governance_snapshot_2026-06-28.json
@@ -0,0 +1,839 @@
+{
+  "candidate_groups": {
+    "production_baseline": [
+      "openclaw_incumbent"
+    ],
+    "replay_or_integration_blocked": [
+      "claude_agent_sdk_remediator",
+      "crewai_flows_crews",
+      "langgraph_incident_kernel",
+      "nemo_nemotron_fabric",
+      "openai_agents_sdk_coordinator"
+    ],
+    "watch_only_candidates": [
+      "agentos_framework",
+      "bernstein_agent_governance",
+      "hermes_agent_personal_platform",
+      "microsoft_agent_governance_toolkit",
+      "pydantic_deepagents",
+      "thclaws_agent_harness"
+    ],
+    "watch_only_scorecard_prescreen_ready": []
+  },
+  "candidate_statuses": [
+    {
+      "approvals": {
+        "paid_api": false,
+        "production_routing": false,
+        "replay": false,
+        "sdk_install": false,
+        "shadow_or_canary": false
+      },
+      "candidate_id": "openclaw_incumbent",
+      "current_gate": "production_decision_core",
+      "display_name": "OpenClaw incumbent",
+      "evaluation_priority": "baseline",
+      "evidence": {
+        "latest_replay_summary": null,
+        "latest_smoke_gate": null,
+        "latest_smoke_matrix": null,
+        "latest_smoke_model": null
+      },
+      "gate_status": "production_baseline",
+      "integration_decision": "",
+      "operator_blockers": [],
+      "required_next_gate": "formal_replacement_adr_and_promotion_gate_required",
+      "role": "current_production_decision_core",
+      "score": null
+    },
+    {
+      "approvals": {
+        "paid_api": false,
+        "production_routing": false,
+        "replay": false,
+        "sdk_install": false,
+        "shadow_or_canary": false
+      },
+      "candidate_id": "openai_agents_sdk_coordinator",
+      "current_gate": "offline_replay",
+      "display_name": "OpenAI Agents SDK Coordinator",
+      "evaluation_priority": "must_test",
+      "evidence": {
+        "latest_replay_summary": "docs/evaluations/agent_openai_coordinator_replay_summary_2026-06-02.json",
+        "latest_smoke_gate": null,
+        "latest_smoke_matrix": null,
+        "latest_smoke_model": null
+      },
+      "gate_status": "registered_no_review",
+      "integration_decision": "deterministic_offline_coordinator_blocked_does_not_beat_openclaw",
+      "operator_blockers": [],
+      "required_next_gate": "continue_weekly_primary_source_market_watch",
+      "role": "coordinator_orchestrator",
+      "score": null
+    },
+    {
+      "approvals": {
+        "paid_api": false,
+        "production_routing": false,
+        "replay": false,
+        "sdk_install": false,
+        "shadow_or_canary": false
+      },
+      "candidate_id": "langgraph_incident_kernel",
+      "current_gate": "offline_replay",
+      "display_name": "LangGraph Incident Kernel",
+      "evaluation_priority": "must_test",
+      "evidence": {
+        "latest_replay_summary": "docs/evaluations/agent_langgraph_replay_summary_2026-06-02.json",
+        "latest_smoke_gate": null,
+        "latest_smoke_matrix": null,
+        "latest_smoke_model": null
+      },
+      "gate_status": "registered_no_review",
+      "integration_decision": "deterministic_offline_kernel_blocked_does_not_beat_openclaw",
+      "operator_blockers": [],
+      "required_next_gate": "continue_weekly_primary_source_market_watch",
+      "role": "durable_incident_workflow_kernel",
+      "score": null
+    },
+    {
+      "approvals": {
+        "paid_api": false,
+        "production_routing": false,
+        "replay": false,
+        "sdk_install": false,
+        "shadow_or_canary": false
+      },
+      "candidate_id": "nemo_nemotron_fabric",
+      "current_gate": "blocked_existing_replay_evidence",
+      "display_name": "NVIDIA NeMo Agent Toolkit + Nemotron Fabric",
+      "evaluation_priority": "must_test",
+      "evidence": {
+        "latest_replay_summary": null,
+        "latest_smoke_gate": "docs/evaluations/agent_nemotron_contract_tuned_49b_v15_smoke_gate_2026-06-02.json",
+        "latest_smoke_matrix": "docs/evaluations/agent_nemotron_contract_tuned_smoke_matrix_2026-06-02.json",
+        "latest_smoke_model": "nvidia/llama-3.3-nemotron-super-49b-v1.5"
+      },
+      "gate_status": "integration_blocked",
+      "integration_decision": "do_not_integrate_refresh_evidence_then_smoke_gate",
+      "operator_blockers": [
+        "changed_sources_reviewed_by_ai_controlled_verifier",
+        "market_scorecard_refreshed_if_primary_sources_changed_semantically",
+        "no_sdk_install_without_dependency_approval",
+        "no_paid_provider_use_without_cost_and_data_boundary_approval",
+        "5_record_smoke_gate_passes",
+        "latency_and_output_contract_blockers_resolved",
+        "cost_approval_recorded"
+      ],
+      "required_next_gate": "refresh_source_evidence_then_5_record_smoke_only",
+      "role": "agent_fabric_tool_model_evaluator",
+      "score": 0.8033
+    },
+    {
+      "approvals": {
+        "paid_api": false,
+        "production_routing": false,
+        "replay": false,
+        "sdk_install": false,
+        "shadow_or_canary": false
+      },
+      "candidate_id": "claude_agent_sdk_remediator",
+      "current_gate": "has_offline_replay_summary",
+      "display_name": "Claude Agent SDK Remediator",
+      "evaluation_priority": "must_test",
+      "evidence": {
+        "latest_replay_summary": "docs/evaluations/agent_claude_remediator_replay_summary_2026-06-02.json",
+        "latest_smoke_gate": null,
+        "latest_smoke_matrix": null,
+        "latest_smoke_model": null
+      },
+      "gate_status": "integration_blocked",
+      "integration_decision": "do_not_integrate_refresh_replay_gate",
+      "operator_blockers": [
+        "changed_sources_reviewed_by_ai_controlled_verifier",
+        "market_scorecard_refreshed_if_primary_sources_changed_semantically",
+        "no_sdk_install_without_dependency_approval",
+        "no_paid_provider_use_without_cost_and_data_boundary_approval",
+        "offline_adapter_contract_valid",
+        "50_record_hidden_label_replay_beats_openclaw_baseline",
+        "cost_approval_recorded"
+      ],
+      "required_next_gate": "refresh_scorecard_then_offline_replay_or_promotion_gate",
+      "role": "devops_code_remediation_agent",
+      "score": 0.7533
+    },
+    {
+      "approvals": {
+        "paid_api": false,
+        "production_routing": false,
+        "replay": false,
+        "sdk_install": false,
+        "shadow_or_canary": false
+      },
+      "candidate_id": "google_adk_stack",
+      "current_gate": "offline_replay",
+      "display_name": "Google Agent Development Kit Stack",
+      "evaluation_priority": "can_test",
+      "evidence": {
+        "latest_replay_summary": null,
+        "latest_smoke_gate": null,
+        "latest_smoke_matrix": null,
+        "latest_smoke_model": null
+      },
+      "gate_status": "registered_no_review",
+      "integration_decision": "",
+      "operator_blockers": [],
+      "required_next_gate": "continue_weekly_primary_source_market_watch",
+      "role": "gemini_vertex_agent_stack",
+      "score": null
+    },
+    {
+      "approvals": {
+        "paid_api": false,
+        "production_routing": false,
+        "replay": false,
+        "sdk_install": false,
+        "shadow_or_canary": false
+      },
+      "candidate_id": "microsoft_agent_framework",
+      "current_gate": "offline_replay",
+      "display_name": "Microsoft Agent Framework",
+      "evaluation_priority": "can_test",
+      "evidence": {
+        "latest_replay_summary": null,
+        "latest_smoke_gate": null,
+        "latest_smoke_matrix": null,
+        "latest_smoke_model": null
+      },
+      "gate_status": "registered_no_review",
+      "integration_decision": "",
+      "operator_blockers": [],
+      "required_next_gate": "continue_weekly_primary_source_market_watch",
+      "role": "enterprise_workflow_agent_stack",
+      "score": null
+    },
+    {
+      "approvals": {
+        "paid_api": false,
+        "production_routing": false,
+        "replay": false,
+        "sdk_install": false,
+        "shadow_or_canary": false
+      },
+      "candidate_id": "crewai_flows_crews",
+      "current_gate": "not_yet_replayed",
+      "display_name": "CrewAI Flows + Crews",
+      "evaluation_priority": "secondary",
+      "evidence": {
+        "latest_replay_summary": null,
+        "latest_smoke_gate": null,
+        "latest_smoke_matrix": null,
+        "latest_smoke_model": null
+      },
+      "gate_status": "integration_blocked",
+      "integration_decision": "do_not_integrate_prepare_no_cost_offline_adapter",
+      "operator_blockers": [
+        "changed_sources_reviewed_by_ai_controlled_verifier",
+        "market_scorecard_refreshed_if_primary_sources_changed_semantically",
+        "no_sdk_install_without_dependency_approval",
+        "no_paid_provider_use_without_cost_and_data_boundary_approval",
+        "offline_adapter_contract_valid",
+        "50_record_hidden_label_replay_beats_openclaw_baseline"
+      ],
+      "required_next_gate": "create_no_sdk_no_api_adapter_then_offline_replay",
+      "role": "rapid_agent_team_prototype",
+      "score": 0.6033
+    },
+    {
+      "approvals": {
+        "paid_api": false,
+        "production_routing": false,
+        "replay": false,
+        "sdk_install": false,
+        "shadow_or_canary": false
+      },
+      "candidate_id": "hermes_agent_personal_platform",
+      "current_gate": "watch_only_primary_source_monitoring",
+      "display_name": "NousResearch Hermes Agent",
+      "evaluation_priority": "watch_only",
+      "evidence": {
+        "latest_replay_summary": null,
+        "latest_smoke_gate": null,
+        "latest_smoke_matrix": null,
+        "latest_smoke_model": null
+      },
+      "gate_status": "watch_only_blocked",
+      "integration_decision": "remain_watch_only_until_evidence_gap_resolved",
+      "operator_blockers": [
+        "integration_review_must_confirm_watch_only_stage",
+        "discovery_classification_must_recommend_watch_addition"
+      ],
+      "required_next_gate": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "role": "personal_agent_platform_candidate",
+      "score": null
+    },
+    {
+      "approvals": {
+        "paid_api": false,
+        "production_routing": false,
+        "replay": false,
+        "sdk_install": false,
+        "shadow_or_canary": false
+      },
+      "candidate_id": "microsoft_agent_governance_toolkit",
+      "current_gate": "watch_only_primary_source_monitoring",
+      "display_name": "Microsoft Agent Governance Toolkit",
+      "evaluation_priority": "watch_only",
+      "evidence": {
+        "latest_replay_summary": null,
+        "latest_smoke_gate": null,
+        "latest_smoke_matrix": null,
+        "latest_smoke_model": null
+      },
+      "gate_status": "watch_only_blocked",
+      "integration_decision": "remain_watch_only_until_evidence_gap_resolved",
+      "operator_blockers": [
+        "integration_review_must_confirm_watch_only_stage",
+        "discovery_classification_must_recommend_watch_addition"
+      ],
+      "required_next_gate": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "role": "agent_governance_policy_evaluator_candidate",
+      "score": null
+    },
+    {
+      "approvals": {
+        "paid_api": false,
+        "production_routing": false,
+        "replay": false,
+        "sdk_install": false,
+        "shadow_or_canary": false
+      },
+      "candidate_id": "thclaws_agent_harness",
+      "current_gate": "watch_only_primary_source_monitoring",
+      "display_name": "thClaws Agent Harness",
+      "evaluation_priority": "watch_only",
+      "evidence": {
+        "latest_replay_summary": null,
+        "latest_smoke_gate": null,
+        "latest_smoke_matrix": null,
+        "latest_smoke_model": null
+      },
+      "gate_status": "watch_only_blocked",
+      "integration_decision": "do_not_integrate_watch_only_primary_source_monitoring",
+      "operator_blockers": [
+        "discovery_classification_must_recommend_watch_addition",
+        "changed_sources_reviewed_by_ai_controlled_verifier",
+        "market_scorecard_refreshed_if_primary_sources_changed_semantically",
+        "no_sdk_install_without_dependency_approval",
+        "no_paid_provider_use_without_cost_and_data_boundary_approval",
+        "ai_controlled_primary_sources_verifier_passed",
+        "watch_registry_baseline_refreshed",
+        "explicit_priority_upgrade_before_replay",
+        "cost_approval_recorded"
+      ],
+      "required_next_gate": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "role": "agent_framework_or_orchestrator_candidate",
+      "score": null
+    },
+    {
+      "approvals": {
+        "paid_api": false,
+        "production_routing": false,
+        "replay": false,
+        "sdk_install": false,
+        "shadow_or_canary": false
+      },
+      "candidate_id": "pydantic_deepagents",
+      "current_gate": "watch_only_primary_source_monitoring",
+      "display_name": "Pydantic DeepAgents",
+      "evaluation_priority": "watch_only",
+      "evidence": {
+        "latest_replay_summary": null,
+        "latest_smoke_gate": null,
+        "latest_smoke_matrix": null,
+        "latest_smoke_model": null
+      },
+      "gate_status": "watch_only_blocked",
+      "integration_decision": "remain_watch_only_until_evidence_gap_resolved",
+      "operator_blockers": [
+        "integration_review_must_confirm_watch_only_stage",
+        "discovery_classification_must_recommend_watch_addition"
+      ],
+      "required_next_gate": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "role": "agent_framework_or_orchestrator_candidate",
+      "score": null
+    },
+    {
+      "approvals": {
+        "paid_api": false,
+        "production_routing": false,
+        "replay": false,
+        "sdk_install": false,
+        "shadow_or_canary": false
+      },
+      "candidate_id": "agentos_framework",
+      "current_gate": "watch_only_primary_source_monitoring",
+      "display_name": "AgentOS Framework",
+      "evaluation_priority": "watch_only",
+      "evidence": {
+        "latest_replay_summary": null,
+        "latest_smoke_gate": null,
+        "latest_smoke_matrix": null,
+        "latest_smoke_model": null
+      },
+      "gate_status": "watch_only_blocked",
+      "integration_decision": "remain_watch_only_until_evidence_gap_resolved",
+      "operator_blockers": [
+        "integration_review_must_confirm_watch_only_stage",
+        "discovery_classification_must_recommend_watch_addition"
+      ],
+      "required_next_gate": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "role": "agent_framework_or_orchestrator_candidate",
+      "score": null
+    },
+    {
+      "approvals": {
+        "paid_api": false,
+        "production_routing": false,
+        "replay": false,
+        "sdk_install": false,
+        "shadow_or_canary": false
+      },
+      "candidate_id": "bernstein_agent_governance",
+      "current_gate": "watch_only_primary_source_monitoring",
+      "display_name": "Bernstein Agent Governance",
+      "evaluation_priority": "watch_only",
+      "evidence": {
+        "latest_replay_summary": null,
+        "latest_smoke_gate": null,
+        "latest_smoke_matrix": null,
+        "latest_smoke_model": null
+      },
+      "gate_status": "watch_only_blocked",
+      "integration_decision": "do_not_integrate_watch_only_primary_source_monitoring",
+      "operator_blockers": [
+        "discovery_classification_must_recommend_watch_addition",
+        "changed_sources_reviewed_by_ai_controlled_verifier",
+        "market_scorecard_refreshed_if_primary_sources_changed_semantically",
+        "no_sdk_install_without_dependency_approval",
+        "no_paid_provider_use_without_cost_and_data_boundary_approval",
+        "ai_controlled_primary_sources_verifier_passed",
+        "watch_registry_baseline_refreshed",
+        "explicit_priority_upgrade_before_replay",
+        "cost_approval_recorded"
+      ],
+      "required_next_gate": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "role": "agent_governance_policy_evaluator_candidate",
+      "score": null
+    }
+  ],
+  "current_decision": "openclaw_remains_production_decision_core",
+  "evaluation_cadence": {
+    "controlled_review_gate": "controlled_priority_upgrade_required_before_scorecard_replay_sdk_api_shadow_canary_or_production",
+    "next_scheduled_run_at": "2026-06-29T09:00:00+08:00",
+    "primary_source_policy": "primary_sources_only_no_llm_no_sdk_no_paid_api",
+    "schedule": "weekly_monday_0900_asia_taipei",
+    "timezone": "Asia/Taipei",
+    "trigger_modes": [
+      "scheduled_weekly",
+      "workflow_dispatch",
+      "agent_controlled_trigger_after_primary_source_signal"
+    ],
+    "workflow": ".gitea/workflows/agent-market-watch.yaml"
+  },
+  "forbidden_actions_without_new_approval": [
+    "replace_openclaw",
+    "enter_shadow_or_canary",
+    "install_new_agent_sdk",
+    "call_paid_provider_api",
+    "run_replay_for_watch_only_candidate",
+    "change_production_routing"
+  ],
+  "generated_at": "2026-06-28T00:47:02.688946+00:00",
+  "inputs": {
+    "candidate_registry_schema_version": "agent_replacement_candidates_v1",
+    "discovery_classification_generated_at": "2026-06-28T00:46:51.907924+00:00",
+    "integration_review_generated_at": "2026-06-26T03:41:38.289424+00:00",
+    "promotion_review_generated_at": "2026-06-28T00:47:01.200212+00:00",
+    "watch_report_generated_at": "2026-06-26T03:41:17.266599+00:00"
+  },
+  "market_watch_health": {
+    "blocked_from_integration": 5,
+    "controlled_queue_reasons": [
+      "unclassified_discovery_watch_additions_remaining",
+      "integration_queue_not_empty"
+    ],
+    "freshness_sla_hours": 168,
+    "operator_blockers": [
+      "unclassified_discovery_watch_additions_remaining",
+      "integration_queue_not_empty"
+    ],
+    "source_failures_block_priority_upgrade": false,
+    "source_failures_require_controlled_reverify": false,
+    "stale_after": "2026-06-29T15:00:00+08:00",
+    "stale_grace_hours": 6,
+    "status": "controlled_queue"
+  },
+  "next_allowed_actions": [
+    "continue_weekly_primary_source_market_watch",
+    "rerun_existing_replay_only_after_evidence_or_adapter_change"
+  ],
+  "operator_decision_queue": [
+    {
+      "approval_boundary": {
+        "market_scorecard_update_required": false,
+        "paid_api_approval_required": true,
+        "priority_upgrade_required": false,
+        "production_routing_approval_required": true,
+        "replacement_adr_required": true,
+        "replay_approval_required": true,
+        "sdk_install_approval_required": true,
+        "shadow_or_canary_approval_required": true
+      },
+      "candidate_id": "claude_agent_sdk_remediator",
+      "display_name": "Claude Agent SDK Remediator",
+      "evidence_refs": [
+        "docs/evaluations/agent_claude_remediator_replay_summary_2026-06-02.json"
+      ],
+      "priority": 10,
+      "queue_status": "blocked_needs_evidence",
+      "recommended_action": "refresh_scorecard_then_offline_replay_or_promotion_gate",
+      "risk_notes": [
+        "Best fit is code and DevOps remediation, not necessarily central incident arbitration.",
+        "API cost, subscription separation, and vendor boundary must be validated.",
+        "changed_sources_reviewed_by_ai_controlled_verifier",
+        "market_scorecard_refreshed_if_primary_sources_changed_semantically",
+        "no_sdk_install_without_dependency_approval",
+        "no_paid_provider_use_without_cost_and_data_boundary_approval"
+      ]
+    },
+    {
+      "approval_boundary": {
+        "market_scorecard_update_required": false,
+        "paid_api_approval_required": false,
+        "priority_upgrade_required": false,
+        "production_routing_approval_required": true,
+        "replacement_adr_required": true,
+        "replay_approval_required": true,
+        "sdk_install_approval_required": true,
+        "shadow_or_canary_approval_required": true
+      },
+      "candidate_id": "crewai_flows_crews",
+      "display_name": "CrewAI Flows + Crews",
+      "evidence_refs": [],
+      "priority": 10,
+      "queue_status": "blocked_needs_evidence",
+      "recommended_action": "create_no_sdk_no_api_adapter_then_offline_replay",
+      "risk_notes": [
+        "Better for rapid automation teams than high-risk production AIOps core.",
+        "Durability, strict audit, and permission boundary must be proven in replay.",
+        "changed_sources_reviewed_by_ai_controlled_verifier",
+        "market_scorecard_refreshed_if_primary_sources_changed_semantically",
+        "no_sdk_install_without_dependency_approval",
+        "no_paid_provider_use_without_cost_and_data_boundary_approval"
+      ]
+    },
+    {
+      "approval_boundary": {
+        "market_scorecard_update_required": false,
+        "paid_api_approval_required": true,
+        "priority_upgrade_required": false,
+        "production_routing_approval_required": true,
+        "replacement_adr_required": true,
+        "replay_approval_required": true,
+        "sdk_install_approval_required": true,
+        "shadow_or_canary_approval_required": true
+      },
+      "candidate_id": "nemo_nemotron_fabric",
+      "display_name": "NVIDIA NeMo Agent Toolkit + Nemotron Fabric",
+      "evidence_refs": [
+        "nvidia/llama-3.3-nemotron-super-49b-v1.5",
+        "docs/evaluations/agent_nemotron_contract_tuned_49b_v15_smoke_gate_2026-06-02.json",
+        "docs/evaluations/agent_nemotron_contract_tuned_smoke_matrix_2026-06-02.json"
+      ],
+      "priority": 10,
+      "queue_status": "blocked_needs_evidence",
+      "recommended_action": "refresh_source_evidence_then_5_record_smoke_only",
+      "risk_notes": [
+        "Needs AWOOOI-specific HITL and dangerous-action policy integration.",
+        "GPU/NIM operating cost must be compared against current local inference.",
+        "changed_sources_reviewed_by_ai_controlled_verifier",
+        "market_scorecard_refreshed_if_primary_sources_changed_semantically",
+        "no_sdk_install_without_dependency_approval",
+        "no_paid_provider_use_without_cost_and_data_boundary_approval"
+      ]
+    },
+    {
+      "approval_boundary": {
+        "market_scorecard_update_required": true,
+        "paid_api_approval_required": false,
+        "priority_upgrade_required": true,
+        "production_routing_approval_required": true,
+        "replacement_adr_required": true,
+        "replay_approval_required": true,
+        "sdk_install_approval_required": true,
+        "shadow_or_canary_approval_required": true
+      },
+      "candidate_id": "agentos_framework",
+      "display_name": "AgentOS Framework",
+      "evidence_refs": [],
+      "priority": 40,
+      "queue_status": "watch_only_blocked",
+      "recommended_action": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "risk_notes": [
+        "integration_review_must_confirm_watch_only_stage",
+        "discovery_classification_must_recommend_watch_addition"
+      ]
+    },
+    {
+      "approval_boundary": {
+        "market_scorecard_update_required": true,
+        "paid_api_approval_required": true,
+        "priority_upgrade_required": true,
+        "production_routing_approval_required": true,
+        "replacement_adr_required": true,
+        "replay_approval_required": true,
+        "sdk_install_approval_required": true,
+        "shadow_or_canary_approval_required": true
+      },
+      "candidate_id": "bernstein_agent_governance",
+      "display_name": "Bernstein Agent Governance",
+      "evidence_refs": [],
+      "priority": 40,
+      "queue_status": "watch_only_blocked",
+      "recommended_action": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "risk_notes": [
+        "candidate missing from current market scorecard",
+        "discovery_classification_must_recommend_watch_addition",
+        "changed_sources_reviewed_by_ai_controlled_verifier",
+        "market_scorecard_refreshed_if_primary_sources_changed_semantically",
+        "no_sdk_install_without_dependency_approval",
+        "no_paid_provider_use_without_cost_and_data_boundary_approval"
+      ]
+    },
+    {
+      "approval_boundary": {
+        "market_scorecard_update_required": true,
+        "paid_api_approval_required": false,
+        "priority_upgrade_required": true,
+        "production_routing_approval_required": true,
+        "replacement_adr_required": true,
+        "replay_approval_required": true,
+        "sdk_install_approval_required": true,
+        "shadow_or_canary_approval_required": true
+      },
+      "candidate_id": "hermes_agent_personal_platform",
+      "display_name": "NousResearch Hermes Agent",
+      "evidence_refs": [],
+      "priority": 40,
+      "queue_status": "watch_only_blocked",
+      "recommended_action": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "risk_notes": [
+        "integration_review_must_confirm_watch_only_stage",
+        "discovery_classification_must_recommend_watch_addition"
+      ]
+    },
+    {
+      "approval_boundary": {
+        "market_scorecard_update_required": true,
+        "paid_api_approval_required": false,
+        "priority_upgrade_required": true,
+        "production_routing_approval_required": true,
+        "replacement_adr_required": true,
+        "replay_approval_required": true,
+        "sdk_install_approval_required": true,
+        "shadow_or_canary_approval_required": true
+      },
+      "candidate_id": "microsoft_agent_governance_toolkit",
+      "display_name": "Microsoft Agent Governance Toolkit",
+      "evidence_refs": [],
+      "priority": 40,
+      "queue_status": "watch_only_blocked",
+      "recommended_action": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "risk_notes": [
+        "integration_review_must_confirm_watch_only_stage",
+        "discovery_classification_must_recommend_watch_addition"
+      ]
+    },
+    {
+      "approval_boundary": {
+        "market_scorecard_update_required": true,
+        "paid_api_approval_required": false,
+        "priority_upgrade_required": true,
+        "production_routing_approval_required": true,
+        "replacement_adr_required": true,
+        "replay_approval_required": true,
+        "sdk_install_approval_required": true,
+        "shadow_or_canary_approval_required": true
+      },
+      "candidate_id": "pydantic_deepagents",
+      "display_name": "Pydantic DeepAgents",
+      "evidence_refs": [],
+      "priority": 40,
+      "queue_status": "watch_only_blocked",
+      "recommended_action": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "risk_notes": [
+        "integration_review_must_confirm_watch_only_stage",
+        "discovery_classification_must_recommend_watch_addition"
+      ]
+    },
+    {
+      "approval_boundary": {
+        "market_scorecard_update_required": true,
+        "paid_api_approval_required": true,
+        "priority_upgrade_required": true,
+        "production_routing_approval_required": true,
+        "replacement_adr_required": true,
+        "replay_approval_required": true,
+        "sdk_install_approval_required": true,
+        "shadow_or_canary_approval_required": true
+      },
+      "candidate_id": "thclaws_agent_harness",
+      "display_name": "thClaws Agent Harness",
+      "evidence_refs": [],
+      "priority": 40,
+      "queue_status": "watch_only_blocked",
+      "recommended_action": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "risk_notes": [
+        "candidate missing from current market scorecard",
+        "discovery_classification_must_recommend_watch_addition",
+        "changed_sources_reviewed_by_ai_controlled_verifier",
+        "market_scorecard_refreshed_if_primary_sources_changed_semantically",
+        "no_sdk_install_without_dependency_approval",
+        "no_paid_provider_use_without_cost_and_data_boundary_approval"
+      ]
+    },
+    {
+      "approval_boundary": {
+        "market_scorecard_update_required": true,
+        "paid_api_approval_required": false,
+        "priority_upgrade_required": true,
+        "production_routing_approval_required": true,
+        "replacement_adr_required": true,
+        "replay_approval_required": true,
+        "sdk_install_approval_required": true,
+        "shadow_or_canary_approval_required": true
+      },
+      "candidate_id": "google_adk_stack",
+      "display_name": "Google Agent Development Kit Stack",
+      "evidence_refs": [],
+      "priority": 60,
+      "queue_status": "registered_no_review",
+      "recommended_action": "continue_weekly_primary_source_market_watch",
+      "risk_notes": []
+    },
+    {
+      "approval_boundary": {
+        "market_scorecard_update_required": true,
+        "paid_api_approval_required": false,
+        "priority_upgrade_required": true,
+        "production_routing_approval_required": true,
+        "replacement_adr_required": true,
+        "replay_approval_required": true,
+        "sdk_install_approval_required": true,
+        "shadow_or_canary_approval_required": true
+      },
+      "candidate_id": "langgraph_incident_kernel",
+      "display_name": "LangGraph Incident Kernel",
+      "evidence_refs": [
+        "docs/evaluations/agent_langgraph_replay_summary_2026-06-02.json"
+      ],
+      "priority": 60,
+      "queue_status": "registered_no_review",
+      "recommended_action": "continue_weekly_primary_source_market_watch",
+      "risk_notes": []
+    },
+    {
+      "approval_boundary": {
+        "market_scorecard_update_required": true,
+        "paid_api_approval_required": false,
+        "priority_upgrade_required": true,
+        "production_routing_approval_required": true,
+        "replacement_adr_required": true,
+        "replay_approval_required": true,
+        "sdk_install_approval_required": true,
+        "shadow_or_canary_approval_required": true
+      },
+      "candidate_id": "microsoft_agent_framework",
+      "display_name": "Microsoft Agent Framework",
+      "evidence_refs": [],
+      "priority": 60,
+      "queue_status": "registered_no_review",
+      "recommended_action": "continue_weekly_primary_source_market_watch",
+      "risk_notes": []
+    },
+    {
+      "approval_boundary": {
+        "market_scorecard_update_required": true,
+        "paid_api_approval_required": false,
+        "priority_upgrade_required": true,
+        "production_routing_approval_required": true,
+        "replacement_adr_required": true,
+        "replay_approval_required": true,
+        "sdk_install_approval_required": true,
+        "shadow_or_canary_approval_required": true
+      },
+      "candidate_id": "openai_agents_sdk_coordinator",
+      "display_name": "OpenAI Agents SDK Coordinator",
+      "evidence_refs": [
+        "docs/evaluations/agent_openai_coordinator_replay_summary_2026-06-02.json"
+      ],
+      "priority": 60,
+      "queue_status": "registered_no_review",
+      "recommended_action": "continue_weekly_primary_source_market_watch",
+      "risk_notes": []
+    },
+    {
+      "approval_boundary": {
+        "market_scorecard_update_required": false,
+        "paid_api_approval_required": false,
+        "priority_upgrade_required": false,
+        "production_routing_approval_required": true,
+        "replacement_adr_required": true,
+        "replay_approval_required": false,
+        "sdk_install_approval_required": false,
+        "shadow_or_canary_approval_required": false
+      },
+      "candidate_id": "openclaw_incumbent",
+      "display_name": "OpenClaw incumbent",
+      "evidence_refs": [],
+      "priority": 90,
+      "queue_status": "baseline_protected",
+      "recommended_action": "keep_openclaw_as_production_decision_core_until_replay_shadow_canary_replacement_adr",
+      "risk_notes": [
+        "no_candidate_has_formal_replacement_approval"
+      ]
+    }
+  ],
+  "policy": {
+    "market_scorecard_update_approved": false,
+    "paid_api_calls_approved": false,
+    "priority_upgrade_approved": false,
+    "production_changes_approved": false,
+    "replacement_decision_allowed": false,
+    "replay_candidate_approved": false,
+    "sdk_installation_approved": false,
+    "shadow_or_canary_approved": false,
+    "snapshot_is_decision_source": false
+  },
+  "schema_version": "agent_market_governance_snapshot_v1",
+  "summary": {
+    "blocked_from_integration": 5,
+    "candidate_count": 13,
+    "changed_candidates": 5,
+    "eligible_for_market_scorecard_prescreen": 0,
+    "integration_queue_count": 5,
+    "market_scorecard_updates_approved": 0,
+    "paid_api_calls_approved": 0,
+    "priority_upgrades_approved": 0,
+    "production_changes_approved": 0,
+    "recommended_watch_additions_remaining": 7,
+    "replacement_decisions_approved": 0,
+    "replay_candidates_approved": 0,
+    "sdk_installations_approved": 0,
+    "shadow_or_canary_approved": 0,
+    "source_count": 36,
+    "source_failures": 0,
+    "watch_only_candidates_reviewed": 6
+  }
+}
diff --git a/docs/evaluations/agent_market_watch_promotion_review_2026-06-28.json b/docs/evaluations/agent_market_watch_promotion_review_2026-06-28.json
new file mode 100644
index 000000000..e1526ee20
--- /dev/null
+++ b/docs/evaluations/agent_market_watch_promotion_review_2026-06-28.json
@@ -0,0 +1,218 @@
+{
+  "generated_at": "2026-06-28T00:47:01.200212+00:00",
+  "inputs": {
+    "candidate_registry_schema_version": "agent_replacement_candidates_v1",
+    "discovery_classification_generated_at": "2026-06-28T00:46:51.907924+00:00",
+    "integration_review_generated_at": "2026-06-26T03:41:38.289424+00:00",
+    "watch_report_generated_at": "2026-06-26T03:41:17.266599+00:00"
+  },
+  "policy": {
+    "market_scorecard_update_approved": false,
+    "paid_api_calls_approved": false,
+    "priority_upgrade_approved": false,
+    "production_changes_approved": false,
+    "replacement_decision_allowed": false,
+    "replay_candidate_approved": false,
+    "sdk_installation_approved": false,
+    "shadow_or_canary_approved": false
+  },
+  "reviews": [
+    {
+      "approved_for_paid_api_calls": false,
+      "approved_for_replay": false,
+      "approved_for_sdk_install": false,
+      "approved_for_shadow_or_canary": false,
+      "blockers": [
+        "integration_review_must_confirm_watch_only_stage",
+        "discovery_classification_must_recommend_watch_addition"
+      ],
+      "candidate_id": "hermes_agent_personal_platform",
+      "classification": {
+        "classification": null,
+        "recommendation": null,
+        "repository_full_name": null,
+        "risk_flags": [],
+        "watch_addition_recommended": false
+      },
+      "decision": "remain_watch_only_until_evidence_gap_resolved",
+      "display_name": "NousResearch Hermes Agent",
+      "eligible_for_market_scorecard_prescreen": false,
+      "integration_stage": "",
+      "latest_versions": [
+        "v2026.6.19"
+      ],
+      "official_url": "https://hermes-agent.nousresearch.com",
+      "release_version_observed": true,
+      "required_next_gate": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "role": "personal_agent_platform_candidate",
+      "source_count": 2,
+      "source_failures": 0
+    },
+    {
+      "approved_for_paid_api_calls": false,
+      "approved_for_replay": false,
+      "approved_for_sdk_install": false,
+      "approved_for_shadow_or_canary": false,
+      "blockers": [
+        "integration_review_must_confirm_watch_only_stage",
+        "discovery_classification_must_recommend_watch_addition"
+      ],
+      "candidate_id": "microsoft_agent_governance_toolkit",
+      "classification": {
+        "classification": null,
+        "recommendation": null,
+        "repository_full_name": null,
+        "risk_flags": [],
+        "watch_addition_recommended": false
+      },
+      "decision": "remain_watch_only_until_evidence_gap_resolved",
+      "display_name": "Microsoft Agent Governance Toolkit",
+      "eligible_for_market_scorecard_prescreen": false,
+      "integration_stage": "",
+      "latest_versions": [
+        "v4.1.0"
+      ],
+      "official_url": "https://microsoft.github.io/agent-governance-toolkit/",
+      "release_version_observed": true,
+      "required_next_gate": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "role": "agent_governance_policy_evaluator_candidate",
+      "source_count": 2,
+      "source_failures": 0
+    },
+    {
+      "approved_for_paid_api_calls": false,
+      "approved_for_replay": false,
+      "approved_for_sdk_install": false,
+      "approved_for_shadow_or_canary": false,
+      "blockers": [
+        "discovery_classification_must_recommend_watch_addition"
+      ],
+      "candidate_id": "thclaws_agent_harness",
+      "classification": {
+        "classification": null,
+        "recommendation": null,
+        "repository_full_name": null,
+        "risk_flags": [],
+        "watch_addition_recommended": false
+      },
+      "decision": "remain_watch_only_until_evidence_gap_resolved",
+      "display_name": "thClaws Agent Harness",
+      "eligible_for_market_scorecard_prescreen": false,
+      "integration_stage": "watch_only_primary_source_monitoring",
+      "latest_versions": [
+        "v0.75.0"
+      ],
+      "official_url": "https://thclaws.ai",
+      "release_version_observed": true,
+      "required_next_gate": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "role": "agent_framework_or_orchestrator_candidate",
+      "source_count": 2,
+      "source_failures": 0
+    },
+    {
+      "approved_for_paid_api_calls": false,
+      "approved_for_replay": false,
+      "approved_for_sdk_install": false,
+      "approved_for_shadow_or_canary": false,
+      "blockers": [
+        "integration_review_must_confirm_watch_only_stage",
+        "discovery_classification_must_recommend_watch_addition"
+      ],
+      "candidate_id": "pydantic_deepagents",
+      "classification": {
+        "classification": null,
+        "recommendation": null,
+        "repository_full_name": null,
+        "risk_flags": [],
+        "watch_addition_recommended": false
+      },
+      "decision": "remain_watch_only_until_evidence_gap_resolved",
+      "display_name": "Pydantic DeepAgents",
+      "eligible_for_market_scorecard_prescreen": false,
+      "integration_stage": "",
+      "latest_versions": [
+        "0.3.31"
+      ],
+      "official_url": "https://vstorm-co.github.io/pydantic-deepagents/",
+      "release_version_observed": true,
+      "required_next_gate": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "role": "agent_framework_or_orchestrator_candidate",
+      "source_count": 2,
+      "source_failures": 0
+    },
+    {
+      "approved_for_paid_api_calls": false,
+      "approved_for_replay": false,
+      "approved_for_sdk_install": false,
+      "approved_for_shadow_or_canary": false,
+      "blockers": [
+        "integration_review_must_confirm_watch_only_stage",
+        "discovery_classification_must_recommend_watch_addition"
+      ],
+      "candidate_id": "agentos_framework",
+      "classification": {
+        "classification": null,
+        "recommendation": null,
+        "repository_full_name": null,
+        "risk_flags": [],
+        "watch_addition_recommended": false
+      },
+      "decision": "remain_watch_only_until_evidence_gap_resolved",
+      "display_name": "AgentOS Framework",
+      "eligible_for_market_scorecard_prescreen": false,
+      "integration_stage": "",
+      "latest_versions": [
+        "v0.9.79"
+      ],
+      "official_url": "https://agentos.sh",
+      "release_version_observed": true,
+      "required_next_gate": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "role": "agent_framework_or_orchestrator_candidate",
+      "source_count": 2,
+      "source_failures": 0
+    },
+    {
+      "approved_for_paid_api_calls": false,
+      "approved_for_replay": false,
+      "approved_for_sdk_install": false,
+      "approved_for_shadow_or_canary": false,
+      "blockers": [
+        "discovery_classification_must_recommend_watch_addition"
+      ],
+      "candidate_id": "bernstein_agent_governance",
+      "classification": {
+        "classification": null,
+        "recommendation": null,
+        "repository_full_name": null,
+        "risk_flags": [],
+        "watch_addition_recommended": false
+      },
+      "decision": "remain_watch_only_until_evidence_gap_resolved",
+      "display_name": "Bernstein Agent Governance",
+      "eligible_for_market_scorecard_prescreen": false,
+      "integration_stage": "watch_only_primary_source_monitoring",
+      "latest_versions": [
+        "v2.7.0"
+      ],
+      "official_url": "https://bernstein.run",
+      "release_version_observed": true,
+      "required_next_gate": "continue_watch_only_until_primary_source_evidence_is_sufficient",
+      "role": "agent_governance_policy_evaluator_candidate",
+      "source_count": 2,
+      "source_failures": 0
+    }
+  ],
+  "schema_version": "agent_market_watch_promotion_review_v1",
+  "summary": {
+    "eligible_for_market_scorecard_prescreen": 0,
+    "market_scorecard_updates_approved": 0,
+    "paid_api_calls_approved": 0,
+    "priority_upgrades_approved": 0,
+    "production_changes_approved": 0,
+    "remain_watch_only": 6,
+    "replay_candidates_approved": 0,
+    "sdk_installations_approved": 0,
+    "shadow_or_canary_approved": 0,
+    "watch_only_candidates_reviewed": 6
+  }
+}
diff --git a/docs/schemas/agent_market_discovery_review_v1.schema.json b/docs/schemas/agent_market_discovery_review_v1.schema.json
index 4fbcf3a8d..3d46515a4 100644
--- a/docs/schemas/agent_market_discovery_review_v1.schema.json
+++ b/docs/schemas/agent_market_discovery_review_v1.schema.json
@@ -64,8 +64,8 @@
         "discovered_items",
         "unique_repositories",
         "already_watched_or_registered",
-        "manual_classification_required",
-        "new_manual_classification_required",
+        "controlled_classification_queue",
+        "new_controlled_classification_queue",
         "source_failures",
         "auto_registry_additions_approved",
         "production_changes_approved",
@@ -76,8 +76,8 @@
         "discovered_items": {"type": "integer", "minimum": 0},
         "unique_repositories": {"type": "integer", "minimum": 0},
         "already_watched_or_registered": {"type": "integer", "minimum": 0},
-        "manual_classification_required": {"type": "integer", "minimum": 0},
-        "new_manual_classification_required": {"type": "integer", "minimum": 0},
+        "controlled_classification_queue": {"type": "integer", "minimum": 0},
+        "new_controlled_classification_queue": {"type": "integer", "minimum": 0},
         "source_failures": {"type": "integer", "minimum": 0},
         "auto_registry_additions_approved": {"type": "integer", "const": 0},
         "production_changes_approved": {"type": "integer", "const": 0},
diff --git a/docs/schemas/agent_market_governance_snapshot_v1.schema.json b/docs/schemas/agent_market_governance_snapshot_v1.schema.json
index 982531f4f..b7cfd4845 100644
--- a/docs/schemas/agent_market_governance_snapshot_v1.schema.json
+++ b/docs/schemas/agent_market_governance_snapshot_v1.schema.json
@@ -73,7 +73,7 @@
         "next_scheduled_run_at",
         "trigger_modes",
         "primary_source_policy",
-        "operator_review_gate"
+        "controlled_review_gate"
       ],
       "properties": {
         "workflow": {
@@ -101,7 +101,7 @@
           "type": "string",
           "minLength": 1
         },
-        "operator_review_gate": {
+        "controlled_review_gate": {
           "type": "string",
           "minLength": 1
         }
@@ -115,14 +115,15 @@
         "freshness_sla_hours",
         "stale_grace_hours",
         "stale_after",
-        "source_failures_block_priority_upgrade",
+        "source_failures_require_controlled_reverify",
         "blocked_from_integration",
+        "controlled_queue_reasons",
         "operator_blockers"
       ],
       "properties": {
         "status": {
           "type": "string",
-          "enum": ["healthy", "blocked"]
+          "enum": ["healthy", "controlled_queue", "blocked"]
         },
         "freshness_sla_hours": {
           "type": "integer",
@@ -136,6 +137,9 @@
           "type": "string",
           "minLength": 1
         },
+        "source_failures_require_controlled_reverify": {
+          "type": "boolean"
+        },
         "source_failures_block_priority_upgrade": {
           "type": "boolean"
         },
@@ -143,6 +147,10 @@
           "type": "integer",
           "minimum": 0
         },
+        "controlled_queue_reasons": {
+          "type": "array",
+          "items": {"type": "string", "minLength": 1}
+        },
         "operator_blockers": {
           "type": "array",
           "items": {"type": "string", "minLength": 1}
@@ -316,8 +324,8 @@
             "enum": [
               "baseline_protected",
               "blocked_needs_evidence",
-              "operator_review_required",
-              "operator_priority_review",
+              "controlled_review_required",
+              "controlled_priority_review",
               "watch_only_blocked",
               "watch_only_monitoring",
               "registered_no_review"

From 64f52017b8e4b43f7c1fabbf6fda443b7150d68b Mon Sep 17 00:00:00 2001
From: Your Name <you@example.com>
Date: Sun, 28 Jun 2026 08:54:55 +0800
Subject: [PATCH 6/6] chore(cd): deploy d4c2cc6 [skip ci]

---
 k8s/awoooi-prod/kustomization.yaml | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/k8s/awoooi-prod/kustomization.yaml b/k8s/awoooi-prod/kustomization.yaml
index 81bb3e659..da429a27e 100644
--- a/k8s/awoooi-prod/kustomization.yaml
+++ b/k8s/awoooi-prod/kustomization.yaml
@@ -41,7 +41,7 @@ resources:
 images:
 - name: 192.168.0.110:5000/library/api:IMAGE_TAG_PLACEHOLDER
   newName: 192.168.0.110:5000/awoooi/api
-  newTag: 3a528138829701a1dafb2a21003ed014c9f9383b
+  newTag: d4c2cc6e200fc00e07d179ebb9a4a156cef2c6d5
 - name: 192.168.0.110:5000/library/web:IMAGE_TAG_PLACEHOLDER
   newName: 192.168.0.110:5000/awoooi/web
-  newTag: 3a528138829701a1dafb2a21003ed014c9f9383b
+  newTag: d4c2cc6e200fc00e07d179ebb9a4a156cef2c6d5