diff --git a/apps/api/src/api/v1/agents.py b/apps/api/src/api/v1/agents.py index 553d2a71..c6f51abc 100644 --- a/apps/api/src/api/v1/agents.py +++ b/apps/api/src/api/v1/agents.py @@ -85,6 +85,9 @@ from src.services.ai_agent_live_read_model_gate import ( from src.services.ai_agent_12_agent_war_room import ( load_latest_ai_agent_12_agent_war_room, ) +from src.services.ai_agent_professional_task_expansion import ( + load_latest_ai_agent_professional_task_expansion, +) from src.services.ai_agent_matched_playbook_learning_gap import ( load_latest_ai_agent_matched_playbook_learning_gap, ) @@ -760,6 +763,36 @@ async def get_agent_12_agent_war_room() -> dict[str, Any]: ) from exc +@router.get( + "/agent-professional-task-expansion", + response_model=dict[str, Any], + summary="取得 AI Agent 專業任務擴展與 Telegram Runtime Bridge 快照", + description=( + "讀取最新已提交的 P2-405A AI Agent 專業任務擴展與 Telegram Runtime Bridge 只讀快照；" + "此端點只呈現 OpenClaw、Hermes、NemoTron 與專責 Agent 可承接的專業任務、MCP/RAG、" + "風險分層、Telegram no-send preview 與後續 canary gate，" + "不寫 Gateway queue、不送 Telegram、不呼叫 Bot API、不讀 secret、不執行 production write、" + "不改主機、不執行 kubectl。" + ), +) +async def get_agent_professional_task_expansion() -> dict[str, Any]: + """回傳最新 AI Agent 專業任務擴展與 Telegram Runtime Bridge 只讀快照。""" + try: + payload = await asyncio.to_thread(load_latest_ai_agent_professional_task_expansion) + return redact_public_lan_topology(payload) + except FileNotFoundError as exc: + raise HTTPException( + status_code=status.HTTP_404_NOT_FOUND, + detail=str(exc), + ) from exc + except (json.JSONDecodeError, ValueError) as exc: + logger.error("ai_agent_professional_task_expansion_invalid", error=str(exc)) + raise HTTPException( + status_code=status.HTTP_500_INTERNAL_SERVER_ERROR, + detail="AI Agent 專業任務擴展與 Telegram Runtime Bridge 快照無效", + ) from exc + + @router.get( "/agent-communication-learning-contract", response_model=dict[str, Any], diff --git a/apps/api/src/services/ai_agent_professional_task_expansion.py b/apps/api/src/services/ai_agent_professional_task_expansion.py new file mode 100644 index 00000000..da658fcc --- /dev/null +++ b/apps/api/src/services/ai_agent_professional_task_expansion.py @@ -0,0 +1,241 @@ +""" +AI Agent professional task expansion and Telegram runtime bridge snapshot. + +Loads the latest committed P2-405A read-only contract. The contract expands +professional AI Agent work and defines Telegram bridge stages, but it does not +write Telegram Gateway queues, send Telegram messages, call the Bot API, read +secrets, or execute production changes. +""" + +from __future__ import annotations + +import copy +import json +from pathlib import Path +from typing import Any + +from src.services.snapshot_paths import default_evaluations_dir + +_DEFAULT_EVALUATIONS_DIR = default_evaluations_dir(Path(__file__)) +_SNAPSHOT_PATTERN = "ai_agent_professional_task_expansion_*.json" +_SCHEMA_VERSION = "ai_agent_professional_task_expansion_v1" +_RUNTIME_AUTHORITY = "professional_task_expansion_and_telegram_bridge_read_only_no_send" +_EXPECTED_TASK_COUNT = 24 +_EXPECTED_DOMAIN_COUNT = 8 +_EXPECTED_STAGE_COUNT = 5 +_EXPECTED_MESSAGE_TYPE_COUNT = 6 +_ZERO_ROLLUP_FIELDS = { + "current_live_count", + "gateway_queue_write_count", + "telegram_send_count", + "bot_api_call_count", + "delivery_receipt_write_count", + "production_write_count", + "secret_read_count", + "paid_api_call_count", + "host_write_count", + "kubectl_action_count", +} +_FORBIDDEN_PUBLIC_TERMS = { + "work_window_transcript", + "raw prompt", + "private reasoning", + "chain-of-thought", + "telegram token", + "authorization header", + "secret value", +} + + +def load_latest_ai_agent_professional_task_expansion( + evaluations_dir: Path | None = None, +) -> dict[str, Any]: + """Load the newest committed AI Agent professional task expansion snapshot.""" + directory = evaluations_dir or _DEFAULT_EVALUATIONS_DIR + candidates = sorted(directory.glob(_SNAPSHOT_PATTERN)) + if not candidates: + raise FileNotFoundError( + f"no AI Agent professional task expansion snapshots found in {directory}" + ) + + latest = candidates[-1] + with latest.open(encoding="utf-8") as handle: + payload = json.load(handle) + + if not isinstance(payload, dict): + raise ValueError(f"{latest}: expected JSON object") + + label = str(latest) + _require_schema(payload, label) + _require_telegram_bridge(payload, label) + _require_professional_tasks(payload, label) + _require_reporting_and_redaction(payload, label) + _require_rollups(payload, label) + _require_no_forbidden_public_terms(payload, label) + return payload + + +def _require_schema(payload: dict[str, Any], label: str) -> None: + if payload.get("schema_version") != _SCHEMA_VERSION: + raise ValueError(f"{label}: expected schema_version={_SCHEMA_VERSION}") + + status = payload.get("program_status") or {} + expected = { + "current_priority": "P2", + "current_task_id": "P2-405A", + "next_task_id": "P2-405B", + "read_only_mode": True, + "runtime_authority": _RUNTIME_AUTHORITY, + "overall_completion_percent": 82, + } + mismatches = _mismatches(status, expected) + if mismatches: + raise ValueError(f"{label}: program_status mismatch: {mismatches}") + if not status.get("status_note"): + raise ValueError(f"{label}: program_status.status_note is required") + + +def _require_telegram_bridge(payload: dict[str, Any], label: str) -> None: + bridge = payload.get("telegram_runtime_bridge") or {} + expected = { + "canonical_room": "AwoooI SRE 戰情室", + "canonical_room_env": "SRE_GROUP_CHAT_ID", + "gateway_required": True, + "no_send_preview_ready": True, + "queue_preview_readback_ready": True, + "approved_canary_required": True, + "direct_bot_api_allowed": False, + "bot_api_call_enabled": False, + "gateway_queue_write_enabled": False, + "telegram_send_enabled": False, + "delivery_receipt_write_enabled": False, + } + mismatches = _mismatches(bridge, expected) + if mismatches: + raise ValueError(f"{label}: telegram_runtime_bridge mismatch: {mismatches}") + + stages = bridge.get("stages") or [] + if len(stages) != _EXPECTED_STAGE_COUNT: + raise ValueError(f"{label}: expected {_EXPECTED_STAGE_COUNT} Telegram stages") + if any(stage.get("live_send_enabled") is not False for stage in stages): + raise ValueError(f"{label}: Telegram stages must keep live_send_enabled false") + + message_types = bridge.get("message_types") or [] + if len(message_types) != _EXPECTED_MESSAGE_TYPE_COUNT: + raise ValueError(f"{label}: expected {_EXPECTED_MESSAGE_TYPE_COUNT} message types") + + +def _require_professional_tasks(payload: dict[str, Any], label: str) -> None: + domains = payload.get("professional_task_domains") or [] + if len(domains) != _EXPECTED_DOMAIN_COUNT: + raise ValueError(f"{label}: expected {_EXPECTED_DOMAIN_COUNT} professional task domains") + domain_ids = {domain.get("domain_id") for domain in domains} + + tasks = payload.get("professional_tasks") or [] + if len(tasks) != _EXPECTED_TASK_COUNT: + raise ValueError(f"{label}: expected {_EXPECTED_TASK_COUNT} professional tasks") + + task_ids = [task.get("task_id") for task in tasks] + if len(set(task_ids)) != len(task_ids): + raise ValueError(f"{label}: task_id values must be unique") + + owners = {task.get("owner_agent") for task in tasks} + required_owners = { + "openclaw", + "hermes", + "nemotron", + "telegram_ops_liaison", + "security_sentinel", + "sre_sentinel", + "devops_commander", + } + if not required_owners.issubset(owners): + raise ValueError(f"{label}: professional tasks must include owners {sorted(required_owners)}") + + for task in tasks: + task_id = task.get("task_id") + if task.get("domain_id") not in domain_ids: + raise ValueError(f"{label}: {task_id}.domain_id must reference a known domain") + if task.get("current_live_count_24h") != 0: + raise ValueError(f"{label}: {task_id}.current_live_count_24h must remain zero") + if not task.get("required_mcp"): + raise ValueError(f"{label}: {task_id}.required_mcp must not be empty") + if not task.get("required_rag"): + raise ValueError(f"{label}: {task_id}.required_rag must not be empty") + if not task.get("blocked_actions"): + raise ValueError(f"{label}: {task_id}.blocked_actions must not be empty") + + risk = task.get("risk_tier") + if risk in {"high", "critical"} and task.get("approval_required") is not True: + raise ValueError(f"{label}: {task_id} high/critical tasks must require approval") + if risk == "critical" and task.get("automation_mode") not in { + "approval_required_before_execution", + "blocked_until_owner_response", + }: + raise ValueError(f"{label}: {task_id} critical tasks must stay approval/blocker gated") + + +def _require_reporting_and_redaction(payload: dict[str, Any], label: str) -> None: + reporting = payload.get("reporting_contract") or {} + for cadence in ("daily", "weekly", "monthly", "action_required"): + if (reporting.get(cadence) or {}).get("required") is not True: + raise ValueError(f"{label}: reporting_contract.{cadence}.required must be true") + + redaction = payload.get("redaction_contract") or {} + expected = { + "redaction_required": True, + "conversation_transcript_display_allowed": False, + "raw_prompt_display_allowed": False, + "private_reasoning_display_allowed": False, + "secret_value_display_allowed": False, + "raw_runtime_payload_display_allowed": False, + "telegram_message_must_be_sanitized": True, + } + mismatches = _mismatches(redaction, expected) + if mismatches: + raise ValueError(f"{label}: redaction_contract mismatch: {mismatches}") + + +def _require_rollups(payload: dict[str, Any], label: str) -> None: + rollups = payload.get("rollups") or {} + tasks = payload.get("professional_tasks") or [] + domains = payload.get("professional_task_domains") or [] + bridge = payload.get("telegram_runtime_bridge") or {} + + expected = { + "professional_task_count": len(tasks), + "domain_count": len(domains), + "telegram_stage_count": len(bridge.get("stages") or []), + "telegram_message_type_count": len(bridge.get("message_types") or []), + "approval_required_count": sum(1 for task in tasks if task.get("approval_required") is True), + "low_risk_task_count": sum(1 for task in tasks if task.get("risk_tier") == "low"), + "medium_risk_task_count": sum(1 for task in tasks if task.get("risk_tier") == "medium"), + "high_risk_task_count": sum(1 for task in tasks if task.get("risk_tier") == "high"), + "critical_risk_task_count": sum(1 for task in tasks if task.get("risk_tier") == "critical"), + } + mismatches = _mismatches(rollups, expected) + if mismatches: + raise ValueError(f"{label}: rollups mismatch: {mismatches}") + + for field in _ZERO_ROLLUP_FIELDS: + if rollups.get(field) != 0: + raise ValueError(f"{label}: rollups.{field} must remain zero") + + +def _require_no_forbidden_public_terms(payload: dict[str, Any], label: str) -> None: + scrubbed = copy.deepcopy(payload) + redaction = scrubbed.get("redaction_contract") + if isinstance(redaction, dict): + redaction["forbidden_terms"] = [] + public_text = json.dumps(scrubbed, ensure_ascii=False).lower() + leaked = sorted(term for term in _FORBIDDEN_PUBLIC_TERMS if term.lower() in public_text) + if leaked: + raise ValueError(f"{label}: forbidden public terms leaked: {leaked}") + + +def _mismatches(payload: dict[str, Any], expected: dict[str, Any]) -> dict[str, dict[str, Any]]: + return { + key: {"expected": expected_value, "actual": payload.get(key)} + for key, expected_value in expected.items() + if payload.get(key) != expected_value + } diff --git a/apps/api/tests/test_ai_agent_professional_task_expansion.py b/apps/api/tests/test_ai_agent_professional_task_expansion.py new file mode 100644 index 00000000..d20c6430 --- /dev/null +++ b/apps/api/tests/test_ai_agent_professional_task_expansion.py @@ -0,0 +1,126 @@ +from __future__ import annotations + +import copy +import json +import os +from pathlib import Path + +import pytest + +os.environ.setdefault("DATABASE_URL", "postgresql+asyncpg://test:test@localhost/test") + +from src.services.ai_agent_professional_task_expansion import ( + load_latest_ai_agent_professional_task_expansion, +) + + +def test_load_latest_ai_agent_professional_task_expansion_snapshot() -> None: + snapshot = load_latest_ai_agent_professional_task_expansion() + + assert snapshot["schema_version"] == "ai_agent_professional_task_expansion_v1" + assert snapshot["program_status"]["current_task_id"] == "P2-405A" + assert snapshot["program_status"]["next_task_id"] == "P2-405B" + assert snapshot["program_status"]["overall_completion_percent"] == 82 + assert snapshot["program_status"]["runtime_authority"] == ( + "professional_task_expansion_and_telegram_bridge_read_only_no_send" + ) + + bridge = snapshot["telegram_runtime_bridge"] + assert bridge["canonical_room"] == "AwoooI SRE 戰情室" + assert bridge["canonical_room_env"] == "SRE_GROUP_CHAT_ID" + assert bridge["no_send_preview_ready"] is True + assert bridge["queue_preview_readback_ready"] is True + assert bridge["direct_bot_api_allowed"] is False + assert bridge["bot_api_call_enabled"] is False + assert bridge["gateway_queue_write_enabled"] is False + assert bridge["telegram_send_enabled"] is False + assert len(bridge["stages"]) == 5 + assert len(bridge["message_types"]) == 6 + + rollups = snapshot["rollups"] + assert rollups["professional_task_count"] == 24 + assert rollups["domain_count"] == 8 + assert rollups["telegram_stage_count"] == 5 + assert rollups["telegram_message_type_count"] == 6 + assert rollups["approval_required_count"] == 19 + assert rollups["low_risk_task_count"] == 3 + assert rollups["medium_risk_task_count"] == 10 + assert rollups["high_risk_task_count"] == 6 + assert rollups["critical_risk_task_count"] == 5 + assert rollups["current_live_count"] == 0 + assert rollups["gateway_queue_write_count"] == 0 + assert rollups["telegram_send_count"] == 0 + assert rollups["bot_api_call_count"] == 0 + assert rollups["delivery_receipt_write_count"] == 0 + assert rollups["production_write_count"] == 0 + assert rollups["secret_read_count"] == 0 + assert rollups["paid_api_call_count"] == 0 + assert rollups["host_write_count"] == 0 + assert rollups["kubectl_action_count"] == 0 + + +def test_professional_tasks_cover_required_agents_and_reporting() -> None: + snapshot = load_latest_ai_agent_professional_task_expansion() + + owners = {task["owner_agent"] for task in snapshot["professional_tasks"]} + assert { + "openclaw", + "hermes", + "nemotron", + "telegram_ops_liaison", + "security_sentinel", + "sre_sentinel", + "devops_commander", + }.issubset(owners) + + assert snapshot["reporting_contract"]["daily"]["required"] is True + assert snapshot["reporting_contract"]["weekly"]["required"] is True + assert snapshot["reporting_contract"]["monthly"]["required"] is True + assert snapshot["reporting_contract"]["action_required"]["required"] is True + assert snapshot["redaction_contract"]["conversation_transcript_display_allowed"] is False + assert snapshot["redaction_contract"]["raw_prompt_display_allowed"] is False + assert snapshot["redaction_contract"]["private_reasoning_display_allowed"] is False + assert snapshot["redaction_contract"]["secret_value_display_allowed"] is False + + +def test_rejects_telegram_send_enabled(tmp_path: Path) -> None: + snapshot = copy.deepcopy(load_latest_ai_agent_professional_task_expansion()) + snapshot["telegram_runtime_bridge"]["telegram_send_enabled"] = True + _write_snapshot(tmp_path, snapshot) + + with pytest.raises(ValueError, match="telegram_runtime_bridge mismatch"): + load_latest_ai_agent_professional_task_expansion(tmp_path) + + +def test_rejects_gateway_queue_write_count(tmp_path: Path) -> None: + snapshot = copy.deepcopy(load_latest_ai_agent_professional_task_expansion()) + snapshot["rollups"]["gateway_queue_write_count"] = 1 + _write_snapshot(tmp_path, snapshot) + + with pytest.raises(ValueError, match="must remain zero"): + load_latest_ai_agent_professional_task_expansion(tmp_path) + + +def test_rejects_high_risk_without_approval(tmp_path: Path) -> None: + snapshot = copy.deepcopy(load_latest_ai_agent_professional_task_expansion()) + high_task = next(task for task in snapshot["professional_tasks"] if task["risk_tier"] == "high") + high_task["approval_required"] = False + snapshot["rollups"]["approval_required_count"] -= 1 + _write_snapshot(tmp_path, snapshot) + + with pytest.raises(ValueError, match="high/critical tasks must require approval"): + load_latest_ai_agent_professional_task_expansion(tmp_path) + + +def test_rejects_forbidden_public_terms_outside_policy_list(tmp_path: Path) -> None: + snapshot = copy.deepcopy(load_latest_ai_agent_professional_task_expansion()) + snapshot["professional_tasks"][0]["title"] = "raw prompt leakage candidate" + _write_snapshot(tmp_path, snapshot) + + with pytest.raises(ValueError, match="forbidden public terms leaked"): + load_latest_ai_agent_professional_task_expansion(tmp_path) + + +def _write_snapshot(directory: Path, payload: dict) -> None: + path = directory / "ai_agent_professional_task_expansion_2099-01-01.json" + path.write_text(json.dumps(payload, ensure_ascii=False), encoding="utf-8") diff --git a/apps/api/tests/test_ai_agent_professional_task_expansion_api.py b/apps/api/tests/test_ai_agent_professional_task_expansion_api.py new file mode 100644 index 00000000..28d1fa2f --- /dev/null +++ b/apps/api/tests/test_ai_agent_professional_task_expansion_api.py @@ -0,0 +1,39 @@ +from __future__ import annotations + +import os + +from fastapi.testclient import TestClient + +os.environ.setdefault("DATABASE_URL", "postgresql+asyncpg://test:test@localhost/test") + +from src.main import app + + +def test_ai_agent_professional_task_expansion_endpoint() -> None: + client = TestClient(app) + + response = client.get("/api/v1/agents/agent-professional-task-expansion") + + assert response.status_code == 200 + payload = response.json() + assert payload["schema_version"] == "ai_agent_professional_task_expansion_v1" + assert payload["program_status"]["current_task_id"] == "P2-405A" + assert payload["program_status"]["next_task_id"] == "P2-405B" + assert payload["program_status"]["overall_completion_percent"] == 82 + assert payload["program_status"]["runtime_authority"] == ( + "professional_task_expansion_and_telegram_bridge_read_only_no_send" + ) + assert payload["telegram_runtime_bridge"]["canonical_room"] == "AwoooI SRE 戰情室" + assert payload["telegram_runtime_bridge"]["direct_bot_api_allowed"] is False + assert payload["telegram_runtime_bridge"]["gateway_queue_write_enabled"] is False + assert payload["telegram_runtime_bridge"]["telegram_send_enabled"] is False + assert payload["telegram_runtime_bridge"]["bot_api_call_enabled"] is False + assert payload["rollups"]["professional_task_count"] == 24 + assert payload["rollups"]["domain_count"] == 8 + assert payload["rollups"]["approval_required_count"] == 19 + assert payload["rollups"]["high_risk_task_count"] == 6 + assert payload["rollups"]["critical_risk_task_count"] == 5 + assert payload["rollups"]["gateway_queue_write_count"] == 0 + assert payload["rollups"]["telegram_send_count"] == 0 + assert payload["rollups"]["bot_api_call_count"] == 0 + assert payload["rollups"]["production_write_count"] == 0 diff --git a/apps/web/messages/en.json b/apps/web/messages/en.json index e6da4568..48c5bf1e 100644 --- a/apps/web/messages/en.json +++ b/apps/web/messages/en.json @@ -6447,6 +6447,43 @@ "monthly": "月報" } }, + "professionalTaskExpansion": { + "title": "P2-405A AI Agent 專業任務擴展", + "source": "產生 {generated}；目前 {current}；下一步 {next}", + "runtime": "runtime={value}", + "telegramTitle": "Telegram Runtime Bridge", + "redactionTitle": "脫敏與實發邊界", + "tasksTitle": "首批專業任務", + "metrics": { + "overall": "契約進度", + "tasks": "專業任務", + "domains": "任務領域", + "telegramStages": "TG bridge", + "messageTypes": "訊息類型", + "approvals": "需批准", + "highCritical": "高風險 / critical", + "liveWrites": "live / send / write" + }, + "labels": { + "gateway": "Gateway required={value}", + "noSend": "no-send preview={value}", + "queuePreview": "queue preview={value}", + "directBot": "direct Bot API={value}", + "send": "Telegram send={value}", + "redaction": "脫敏鎖定={value}", + "gatewayWrites": "Gateway 寫入={value}", + "botCalls": "Bot API={value}", + "productionWrites": "正式寫入={value}", + "approval": "需批准={value}", + "liveCount": "live={value}" + }, + "riskTiers": { + "low": "低風險", + "medium": "中風險", + "high": "高風險", + "critical": "Critical" + } + }, "resultCaptureReleaseVerifierOwnerReviewPacket": { "title": "P2-137 釋出驗證器負責人審查包", "source": "產生 {generated}；目前 {current}；下一步 {next}", diff --git a/apps/web/messages/zh-TW.json b/apps/web/messages/zh-TW.json index e6da4568..48c5bf1e 100644 --- a/apps/web/messages/zh-TW.json +++ b/apps/web/messages/zh-TW.json @@ -6447,6 +6447,43 @@ "monthly": "月報" } }, + "professionalTaskExpansion": { + "title": "P2-405A AI Agent 專業任務擴展", + "source": "產生 {generated}；目前 {current}；下一步 {next}", + "runtime": "runtime={value}", + "telegramTitle": "Telegram Runtime Bridge", + "redactionTitle": "脫敏與實發邊界", + "tasksTitle": "首批專業任務", + "metrics": { + "overall": "契約進度", + "tasks": "專業任務", + "domains": "任務領域", + "telegramStages": "TG bridge", + "messageTypes": "訊息類型", + "approvals": "需批准", + "highCritical": "高風險 / critical", + "liveWrites": "live / send / write" + }, + "labels": { + "gateway": "Gateway required={value}", + "noSend": "no-send preview={value}", + "queuePreview": "queue preview={value}", + "directBot": "direct Bot API={value}", + "send": "Telegram send={value}", + "redaction": "脫敏鎖定={value}", + "gatewayWrites": "Gateway 寫入={value}", + "botCalls": "Bot API={value}", + "productionWrites": "正式寫入={value}", + "approval": "需批准={value}", + "liveCount": "live={value}" + }, + "riskTiers": { + "low": "低風險", + "medium": "中風險", + "high": "高風險", + "critical": "Critical" + } + }, "resultCaptureReleaseVerifierOwnerReviewPacket": { "title": "P2-137 釋出驗證器負責人審查包", "source": "產生 {generated}；目前 {current}；下一步 {next}", diff --git a/apps/web/src/app/[locale]/governance/tabs/automation-inventory-tab.tsx b/apps/web/src/app/[locale]/governance/tabs/automation-inventory-tab.tsx index c6ab801b..5d0e9649 100644 --- a/apps/web/src/app/[locale]/governance/tabs/automation-inventory-tab.tsx +++ b/apps/web/src/app/[locale]/governance/tabs/automation-inventory-tab.tsx @@ -41,6 +41,7 @@ import { redactPublicIdentifier } from '@/lib/public-security-redaction' import { apiClient, type AiAgent12AgentWarRoomSnapshot, + type AiAgentProfessionalTaskExpansionSnapshot, type AiAgentCandidateOperationDryRunEvidenceSnapshot, type AiAgentCriticReviewerResultCaptureSnapshot, type AiAgentDeploymentLayoutSnapshot, @@ -463,6 +464,7 @@ export function AutomationInventoryTab() { const [providerRouteMatrix, setProviderRouteMatrix] = useState(null) const [deploymentLayout, setDeploymentLayout] = useState(null) const [warRoom, setWarRoom] = useState(null) + const [professionalTaskExpansion, setProfessionalTaskExpansion] = useState(null) const [proactiveOperations, setProactiveOperations] = useState(null) const [interactionLearningProof, setInteractionLearningProof] = useState(null) const [liveReadModelGate, setLiveReadModelGate] = useState(null) @@ -545,6 +547,7 @@ export function AutomationInventoryTab() { apiClient.getAiProviderRouteMatrix(), apiClient.getAiAgentDeploymentLayout(), apiClient.getAiAgent12AgentWarRoom(), + apiClient.getAiAgentProfessionalTaskExpansion(), apiClient.getAiAgentProactiveOperationsContract(), apiClient.getAiAgentInteractionLearningProof(), apiClient.getAiAgentLiveReadModelGate(), @@ -626,6 +629,7 @@ export function AutomationInventoryTab() { providerRouteMatrixResult, deploymentLayoutResult, warRoomResult, + professionalTaskExpansionResult, proactiveOperationsResult, interactionLearningProofResult, liveReadModelGateResult, @@ -704,6 +708,7 @@ export function AutomationInventoryTab() { setProviderRouteMatrix(providerRouteMatrixResult.status === 'fulfilled' ? providerRouteMatrixResult.value : null) setDeploymentLayout(deploymentLayoutResult.status === 'fulfilled' ? deploymentLayoutResult.value : null) setWarRoom(warRoomResult.status === 'fulfilled' ? warRoomResult.value : null) + setProfessionalTaskExpansion(professionalTaskExpansionResult.status === 'fulfilled' ? professionalTaskExpansionResult.value : null) setProactiveOperations(proactiveOperationsResult.status === 'fulfilled' ? proactiveOperationsResult.value : null) setInteractionLearningProof(interactionLearningProofResult.status === 'fulfilled' ? interactionLearningProofResult.value : null) setLiveReadModelGate(liveReadModelGateResult.status === 'fulfilled' ? liveReadModelGateResult.value : null) @@ -780,6 +785,7 @@ export function AutomationInventoryTab() { providerRouteMatrixResult, deploymentLayoutResult, warRoomResult, + professionalTaskExpansionResult, proactiveOperationsResult, interactionLearningProofResult, liveReadModelGateResult, @@ -2094,7 +2100,7 @@ export function AutomationInventoryTab() { ) } - if (error || !snapshot || !backlog || !backupTargets || !backupReadiness || !backupPolicy || !offsiteEscrow || !giteaHealth || !observabilityMatrix || !providerRouteMatrix || !deploymentLayout || !warRoom || !proactiveOperations || !interactionLearningProof || !liveReadModelGate || !redisDryRunGate || !learningWritebackPackage || !telegramReceiptPackage || !ownerApprovedLearningDryRun || !runtimeWriteGateReview || !postWriteVerifierPackage || !runtimeVerifierEvidenceReview || !reportAutomationReview || !reportStatusBoard || !reportRuntimeReadiness || !reportRuntimeDryRun || !reportRuntimeFixtureReadback || !runtimeWorkerShadowGate || !operationPermissionModel || !candidateOperationDryRunEvidence || !taskResultAuditTrail || !matchedPlaybookLearningGap || !criticReviewerResultCapture || !ownerApprovedResultCaptureDryRun || !ownerApprovedResultCaptureReadback || !runtimeReadbackApprovalPackage || !runtimeReadbackImplementationReview || !reportLiveDeliveryApprovalPackage || !runtimeReadbackFixtureApproval || !runtimeReadbackPromotionGate || !ownerApprovedFixturePromotionGate || !canonicalRuntimeReadbackOwnerAcceptance || !failureReceiptNoSendReplay || !reviewerQueueNoWriteReadback || !resultCaptureNoWriteReadback || !resultCapturePromotionApprovalGate || !ownerApprovedResultCapturePromotionDryRun || !resultCaptureWriteGateReview || !resultCaptureWriterImplementationReview || !resultCaptureWriterDryRunFixture || !resultCaptureWriterDryRunReadback || !resultCaptureOwnerPromotionReview || !resultCaptureOwnerApprovedExecutionRehearsal || !resultCaptureOwnerAcceptanceMaintenanceGate || !resultCaptureOwnerAcceptanceReadbackPreflightHold || !resultCaptureOwnerApprovedPreflightReleasePackage || !resultCaptureOwnerApprovedReleaseReadinessReadback || !resultCaptureOwnerReleaseApprovalGate || !resultCapturePostReleaseVerifierRollbackGate || !resultCaptureFinalReleaseCandidateReadback || !resultCaptureReleaseAuthorizationHold || !resultCaptureReleaseAuthorizationReadbackGate || !resultCaptureReleaseVerifierPreflightGate || !resultCaptureReleaseVerifierOwnerReviewPacket || !resultCaptureReleaseDecisionHold || !resultCaptureReleaseDecisionReadback || !resultCaptureReleaseDecisionNextHandoff || !resultCaptureReleaseDecisionInputPrep || !resultCaptureReleaseDecisionOwnerResponsePreflight || !resultCaptureReleaseDecisionOwnerResponseReadback || !resultCaptureReleaseDecisionOwnerResponseAcceptanceGate || !reportTruthActionabilityReview || !ownerDryRunPackage || !hostStatefulInventory || !serviceHealthGapMatrix || !serviceHealthNotificationPolicy) { + if (error || !snapshot || !backlog || !backupTargets || !backupReadiness || !backupPolicy || !offsiteEscrow || !giteaHealth || !observabilityMatrix || !providerRouteMatrix || !deploymentLayout || !warRoom || !professionalTaskExpansion || !proactiveOperations || !interactionLearningProof || !liveReadModelGate || !redisDryRunGate || !learningWritebackPackage || !telegramReceiptPackage || !ownerApprovedLearningDryRun || !runtimeWriteGateReview || !postWriteVerifierPackage || !runtimeVerifierEvidenceReview || !reportAutomationReview || !reportStatusBoard || !reportRuntimeReadiness || !reportRuntimeDryRun || !reportRuntimeFixtureReadback || !runtimeWorkerShadowGate || !operationPermissionModel || !candidateOperationDryRunEvidence || !taskResultAuditTrail || !matchedPlaybookLearningGap || !criticReviewerResultCapture || !ownerApprovedResultCaptureDryRun || !ownerApprovedResultCaptureReadback || !runtimeReadbackApprovalPackage || !runtimeReadbackImplementationReview || !reportLiveDeliveryApprovalPackage || !runtimeReadbackFixtureApproval || !runtimeReadbackPromotionGate || !ownerApprovedFixturePromotionGate || !canonicalRuntimeReadbackOwnerAcceptance || !failureReceiptNoSendReplay || !reviewerQueueNoWriteReadback || !resultCaptureNoWriteReadback || !resultCapturePromotionApprovalGate || !ownerApprovedResultCapturePromotionDryRun || !resultCaptureWriteGateReview || !resultCaptureWriterImplementationReview || !resultCaptureWriterDryRunFixture || !resultCaptureWriterDryRunReadback || !resultCaptureOwnerPromotionReview || !resultCaptureOwnerApprovedExecutionRehearsal || !resultCaptureOwnerAcceptanceMaintenanceGate || !resultCaptureOwnerAcceptanceReadbackPreflightHold || !resultCaptureOwnerApprovedPreflightReleasePackage || !resultCaptureOwnerApprovedReleaseReadinessReadback || !resultCaptureOwnerReleaseApprovalGate || !resultCapturePostReleaseVerifierRollbackGate || !resultCaptureFinalReleaseCandidateReadback || !resultCaptureReleaseAuthorizationHold || !resultCaptureReleaseAuthorizationReadbackGate || !resultCaptureReleaseVerifierPreflightGate || !resultCaptureReleaseVerifierOwnerReviewPacket || !resultCaptureReleaseDecisionHold || !resultCaptureReleaseDecisionReadback || !resultCaptureReleaseDecisionNextHandoff || !resultCaptureReleaseDecisionInputPrep || !resultCaptureReleaseDecisionOwnerResponsePreflight || !resultCaptureReleaseDecisionOwnerResponseReadback || !resultCaptureReleaseDecisionOwnerResponseAcceptanceGate || !reportTruthActionabilityReview || !ownerDryRunPackage || !hostStatefulInventory || !serviceHealthGapMatrix || !serviceHealthNotificationPolicy) { return (
@@ -3666,6 +3672,34 @@ export function AutomationInventoryTab() { warRoom.display_redaction_contract.conversation_transcript_display_allowed === false && warRoom.display_redaction_contract.redaction_required === true ) + const professionalTaskOverall = professionalTaskExpansion.program_status.overall_completion_percent + const professionalTaskTotal = professionalTaskExpansion.rollups.professional_task_count + const professionalTaskDomains = professionalTaskExpansion.rollups.domain_count + const professionalTaskTelegramStages = professionalTaskExpansion.rollups.telegram_stage_count + const professionalTaskMessageTypes = professionalTaskExpansion.rollups.telegram_message_type_count + const professionalTaskApprovals = professionalTaskExpansion.rollups.approval_required_count + const professionalTaskHighCritical = ( + professionalTaskExpansion.rollups.high_risk_task_count + + professionalTaskExpansion.rollups.critical_risk_task_count + ) + const professionalTaskLiveWrites = ( + professionalTaskExpansion.rollups.current_live_count + + professionalTaskExpansion.rollups.gateway_queue_write_count + + professionalTaskExpansion.rollups.telegram_send_count + + professionalTaskExpansion.rollups.bot_api_call_count + + professionalTaskExpansion.rollups.delivery_receipt_write_count + + professionalTaskExpansion.rollups.production_write_count + + professionalTaskExpansion.rollups.secret_read_count + + professionalTaskExpansion.rollups.paid_api_call_count + + professionalTaskExpansion.rollups.host_write_count + + professionalTaskExpansion.rollups.kubectl_action_count + ) + const professionalTaskRedactionLocked = ( + professionalTaskExpansion.redaction_contract.conversation_transcript_display_allowed === false + && professionalTaskExpansion.redaction_contract.telegram_message_must_be_sanitized === true + && professionalTaskExpansion.redaction_contract.redaction_required === true + ) + const visibleProfessionalTasks = professionalTaskExpansion.professional_tasks.slice(0, 8) const backlogProgressPercent = backlog.progress_summary.overall_percent const explicitApprovalItemCount = backlog.item_approval_boundary_rollup.items_requiring_explicit_approval.length const taskBoundaryCount = snapshot.task_approval_boundary_rollup.total_tasks @@ -4052,6 +4086,98 @@ export function AutomationInventoryTab() {
+ +
+
+
+ + + {t('professionalTaskExpansion.title')} + +
+
+ + +
+
+ +

+ {professionalTaskExpansion.program_status.status_note} +

+ +
+ } /> + } /> + } /> + } /> + } /> + } /> + } /> + } /> +
+ +
+
+ {t('professionalTaskExpansion.telegramTitle')} +

+ {professionalTaskExpansion.telegram_runtime_bridge.canonical_room} +

+
+ + + + + +
+
+ +
+ {t('professionalTaskExpansion.redactionTitle')} +

+ {redactPublicText(professionalTaskExpansion.redaction_contract.frontend_display_policy)} +

+
+ + + + +
+
+
+ + {t('professionalTaskExpansion.tasksTitle')} +
+ {visibleProfessionalTasks.map(task => { + const taskTone = task.risk_tier === 'critical' ? 'danger' : task.risk_tier === 'high' ? 'warn' : 'neutral' + const taskColor = toneColor(taskTone) + return ( +
+
+ + {task.title} + + +
+
+ + + + +
+
+ ) + })} +
+
+ +
diff --git a/apps/web/src/lib/api-client.ts b/apps/web/src/lib/api-client.ts index 4c678022..e3bd6681 100644 --- a/apps/web/src/lib/api-client.ts +++ b/apps/web/src/lib/api-client.ts @@ -334,6 +334,11 @@ export const apiClient = { return handleResponse(res) }, + async getAiAgentProfessionalTaskExpansion() { + const res = await fetch(`${API_BASE_URL}/agents/agent-professional-task-expansion`) + return handleResponse(res) + }, + async getAiAgentProactiveOperationsContract() { const res = await fetch(`${API_BASE_URL}/agents/agent-proactive-operations-contract`) return handleResponse(res) @@ -1418,6 +1423,110 @@ export interface AiAgent12AgentWarRoomSnapshot { } } +export interface AiAgentProfessionalTaskExpansionSnapshot { + schema_version: 'ai_agent_professional_task_expansion_v1' + generated_at: string + program_status: { + overall_completion_percent: number + current_priority: 'P0' | 'P1' | 'P2' | 'P3' + current_task_id: 'P2-405A' + next_task_id: 'P2-405B' + read_only_mode: true + runtime_authority: 'professional_task_expansion_and_telegram_bridge_read_only_no_send' + status_note: string + } + source_refs: string[] + telegram_runtime_bridge: { + canonical_room: string + canonical_room_env: string + gateway_required: boolean + gateway_owner: string + arbiter: string + receipt_owner: string + replay_owner: string + no_send_preview_ready: boolean + queue_preview_readback_ready: boolean + approved_canary_required: boolean + direct_bot_api_allowed: boolean + bot_api_call_enabled: boolean + gateway_queue_write_enabled: boolean + telegram_send_enabled: boolean + delivery_receipt_write_enabled: boolean + stages: Array<{ + stage_id: string + title: string + status: string + allowed_output: string + live_send_enabled: boolean + exit_condition: string + }> + message_types: Array<{ + message_type_id: string + title: string + cadence: string + risk_scope: string + delivery_gate: string + }> + } + professional_task_domains: Array<{ + domain_id: string + title: string + summary: string + primary_owner: string + task_count: number + }> + professional_tasks: Array<{ + task_id: string + title: string + domain_id: string + owner_agent: string + risk_tier: 'low' | 'medium' | 'high' | 'critical' + automation_mode: string + telegram_policy: string + required_mcp: string[] + required_rag: string[] + approval_required: boolean + current_live_count_24h: number + blocked_actions: string[] + }> + reporting_contract: Record<'daily' | 'weekly' | 'monthly' | 'action_required', { + required: boolean + delivery_mode: string + sections: string[] + }> + redaction_contract: { + redaction_required: boolean + conversation_transcript_display_allowed: boolean + raw_prompt_display_allowed: boolean + private_reasoning_display_allowed: boolean + secret_value_display_allowed: boolean + raw_runtime_payload_display_allowed: boolean + telegram_message_must_be_sanitized: boolean + frontend_display_policy: string + } + rollups: { + professional_task_count: number + domain_count: number + telegram_stage_count: number + telegram_message_type_count: number + approval_required_count: number + low_risk_task_count: number + medium_risk_task_count: number + high_risk_task_count: number + critical_risk_task_count: number + current_live_count: number + gateway_queue_write_count: number + telegram_send_count: number + bot_api_call_count: number + delivery_receipt_write_count: number + production_write_count: number + secret_read_count: number + paid_api_call_count: number + host_write_count: number + kubectl_action_count: number + } +} + export interface AiAgentProactiveOperationsContractSnapshot { schema_version: 'ai_agent_proactive_operations_contract_v1' generated_at: string diff --git a/docs/LOGBOOK.md b/docs/LOGBOOK.md index f4c80d2b..b3661655 100644 --- a/docs/LOGBOOK.md +++ b/docs/LOGBOOK.md @@ -83,6 +83,39 @@ - S4.9 owner response、source-control owner acceptance、Nginx reload、host restart、firewall change、active scan、secret collection、runtime execution、action button 全部維持 `0 / false`。 - 本輪未 SSH、未讀 live conf、未改主機、未重啟 Docker / Nginx、未修改 firewall / iptables、未收 secrets 明文、未執行 active scan、未 force push。 +## 2026-06-15｜AI Agent 專業任務擴展與 Telegram Runtime Bridge 只讀契約 + +**背景**：統帥要求繼續盤點 AI Agent 還可以承接哪些專業工作，並質疑 Telegram 群組 / TG Bot 尚未整合到 AI Agent 報表、告警、審核與自動化作業鏈。既有 12-Agent War Room 已定義 OpenClaw / Hermes / NemoTron / SRE / Security / DevOps / Data / Supply Chain / Product / QA / Market / Telegram Ops 分工，但仍需要一份可由 API 讀回、可測試、可被 guard 阻擋誤啟用的專業任務擴展契約。 + +**完成項目**： +- 新增 `ai_agent_professional_task_expansion_v1` schema、committed snapshot、API loader 與 `GET /api/v1/agents/agent-professional-task-expansion`。 +- `/zh-TW/governance?tab=automation-inventory` 已接入 P2-405A 卡片，顯示 24 類專業任務、8 個領域、5 段 Telegram bridge、6 種訊息、需批准 19、高風險 / critical 11 與 live / send / write = 0。 +- 新增 `docs/ai/AI_AGENT_PROFESSIONAL_TASK_EXPANSION_2026-06-15.md`，把 AI Agent 可承接的專業工作固定成 24 類任務、8 個領域、MCP/RAG 需求、風險層級、Telegram policy 與 blocked actions。 +- Telegram Runtime Bridge 已拆成 5 段：no-send preview、queue preview readback、approved canary、日報 / 週報 / 月報 digest、action-required digest。 +- 任務 rollup 固定：professional task `24`、domain `8`、Telegram stage `5`、message type `6`、需批准 `19`、low / medium / high / critical = `3 / 10 / 6 / 5`。 +- 邊界固定：Gateway queue write、Telegram send、Bot API call、delivery receipt write、production write、secret read、paid API、host write、kubectl action 全部 `0 / false`。 + +**完成度與狀態**： +- P2-405A 專業任務擴展契約：`82%`。 +- 專業任務定義 / MCP-RAG / 風險分層：`100%`。 +- Telegram no-send bridge contract：`100%`。 +- Telegram 實發、Gateway queue 寫入、Bot API call、delivery receipt E2E：`0%`，全部仍需後續批准包與 canary gate。 +- 下一步：P2-405B 只允許把完整 no-send message preview、dedup key、receipt expectation 與 canary approval package 顯示到治理頁；本輪卡片只顯示 bridge / rollup，不代表 Telegram 實發。 + +**本地驗證**： +- `python3 -m json.tool docs/evaluations/ai_agent_professional_task_expansion_2026-06-15.json` 通過。 +- `python3 -m json.tool docs/schemas/ai_agent_professional_task_expansion_v1.schema.json` 通過。 +- `python3 -m json.tool apps/web/messages/zh-TW.json` 與 `python3 -m json.tool apps/web/messages/en.json` 通過。 +- `python3 -m py_compile apps/api/src/services/ai_agent_professional_task_expansion.py apps/api/src/api/v1/agents.py` 通過。 +- `DATABASE_URL=postgresql+asyncpg://test:test@localhost/test pytest -q apps/api/tests/test_ai_agent_professional_task_expansion.py apps/api/tests/test_ai_agent_professional_task_expansion_api.py` → `7 passed`。 +- `pnpm --filter @awoooi/web typecheck` 通過。 +- `NEXT_PUBLIC_API_URL=https://awoooi.wooo.work pnpm --filter @awoooi/web build` 通過；`/zh-TW/governance` First Load JS `451 kB`。 +- `python3 scripts/security/security-mirror-progress-guard.py --root .` → `SECURITY_MIRROR_PROGRESS_GUARD_OK`。 +- `python3 scripts/ops/doc-secrets-sanity-check.py docs .gitea apps/web/messages/zh-TW.json apps/web/messages/en.json 'apps/web/src/app/[locale]/governance/tabs/automation-inventory-tab.tsx' apps/web/src/lib/api-client.ts` → `DOC_SECRET_SANITY_OK scanned_files=872`。 +- `git diff --check` 通過。 + +**邊界**：本輪未 SSH、未讀 live conf、未改主機、未修改 Telegram Bot、未發 Telegram、未寫 Gateway queue、未呼叫 Bot API、未讀 secret、未啟用 paid API、未跑 kubectl、未做 production write，也沒有把工作視窗對話內容放到前端。 + ## 2026-06-15｜Monitoring / Alerting / Observability owner response acceptance 只讀帳本 **背景**：Monitoring / Alerting / Observability 已有 repo-only 清冊與 owner request draft，但仍缺少「owner 回覆收件後如何驗收、哪些欄位必填、哪些證據可接受、哪些動作必須阻擋」的固定帳本。這會讓 Prometheus / Alertmanager / Grafana / SigNoz / Sentry / Langfuse / OTEL / Telegram 告警鏈路在後續收件時，可能被誤判成 reload、receiver route change、silence change、Telegram send 或 alert chain smoke 授權。 diff --git a/docs/ai/AI_AGENT_12_AGENT_WAR_ROOM_2026-06-14.md b/docs/ai/AI_AGENT_12_AGENT_WAR_ROOM_2026-06-14.md index c14b9b58..8f835d17 100644 --- a/docs/ai/AI_AGENT_12_AGENT_WAR_ROOM_2026-06-14.md +++ b/docs/ai/AI_AGENT_12_AGENT_WAR_ROOM_2026-06-14.md @@ -184,5 +184,5 @@ Telegram Ops 工位負責 Telegram Gateway、Channel Hub、日報/週報/月報 狀態變更：12 位 Agent 只讀審查全部回收並彙整；已建立 `ai_agent_12_agent_war_room_v1` schema / committed snapshot / API / tests / governance UI 區塊。 證據：`docs/schemas/ai_agent_12_agent_war_room_v1.schema.json`、`docs/evaluations/ai_agent_12_agent_war_room_2026-06-14.json`、`GET /api/v1/agents/agent-12-agent-war-room`、治理頁 12-Agent War Room 區塊、12 位 Agent 只讀回饋。 阻擋：runtime writer、Telegram send、Bot API、production write、SDK 安裝、付費 API、shadow/canary、host update、DB migration、restore 仍未批准。 -下一步：本地驗證、推送 Gitea CD、production API readback 與 desktop / mobile Browser smoke；P2-143 承接 report receipt / 月報 / Agent 工作量 runtime data model。 +下一步：P2-405B 承接 Telegram no-send preview UI；讓統帥先看見 TG 訊息、dedup、receipt 與風險標籤長相。P2-146 owner response receipt preview 仍屬另一條 release gate 主線；兩者都不得直接打開 Gateway queue、Telegram send、Bot API 或 production write。 ``` diff --git a/docs/ai/AI_AGENT_AUTOMATION_WORKLIST_2026-06-04.md b/docs/ai/AI_AGENT_AUTOMATION_WORKLIST_2026-06-04.md index ade3cdb8..d86ca41e 100644 --- a/docs/ai/AI_AGENT_AUTOMATION_WORKLIST_2026-06-04.md +++ b/docs/ai/AI_AGENT_AUTOMATION_WORKLIST_2026-06-04.md @@ -15,6 +15,7 @@ | OpenClaw / Hermes / NemoTron 主動溝通、學習與成長證據 | 100% | P2-401A 到 P2-144 已完成只讀證據面、runtime / report / result-capture gates、no-write readback、promotion review、writer implementation review、writer dry-run fixture、writer dry-run readback、owner promotion execution gate、owner-approved execution rehearsal、owner acceptance / maintenance window gate、owner acceptance readback / preflight hold、owner-approved preflight release package、owner-approved release readiness readback、owner release approval gate、post-release verifier / rollback gate、final release candidate readback、release authorization hold / readback gate、release verifier preflight / owner review packet、release decision hold / readback、release decision next handoff、release decision input prep、12-Agent War Room、owner response 預檢與 owner response 回讀；P2-141 基線與 S4.9 owner release packet 補強皆已正式驗證，P2-142 12-Agent War Room 已完成 production readback 與 desktop / mobile smoke，P2-143 owner response 預檢已完成 production readback 與 in-app browser smoke，P2-144 owner response 回讀已完成 production API readback 與 desktop / mobile smoke。runtime worker、DB migration、production Redis consumer group、canonical runtime readback、live query、runtime score、result capture write、Telegram 實發、delivery receipt E2E、live report delivery、reviewer queue write、Gateway queue write、AI analysis runtime、中低風險 auto worker、KM / LOGBOOK / audit DB / timeline / PlayBook trust 寫入、SDK / 付費服務仍未開 gate | `ai_agent_result_capture_release_decision_owner_response_readback_v1`、`GET /api/v1/agents/agent-result-capture-release-decision-owner-response-readback`、`docs/evaluations/ai_agent_result_capture_release_decision_owner_response_readback_2026-06-14.json`、feature commit `8795f100`、deploy marker `ac938037`、Gitea code-review `2965` / CD `2964` success、5 個回覆讀回 lane、18 個 owner 必填欄位、6 個 readback validation check、6 個 rejection guard、5 個 operator action、等待外部回覆 `5`、未收件 lane `5`、正式寫入 / 發送 `0`；P2-142 feature commit `5de4b3f3`、deploy marker `1a2c9e36`、Gitea CD run `4232` success、production API readback、desktop / mobile in-app browser smoke；P2-143 feature commit `755b0a8d`、deploy marker `667d6329`、Gitea code-review `2961` / CD `2960` success、production API readback、desktop / mobile in-app browser smoke；MASTER §3.2.1b / §3.2.1d / §3.4.3 | | AI Agent 主動營運委派與版本生命週期 | 100% | P2-402A / P2-402B / P2-402C / P2-402D / P2-402E / P2-402F / P2-402G 已完成；已建立 repo-only 版本新鮮度快照、工具採用批准包、Telegram action-required digest policy、Gitea PR 草案 lane、host / K3s / stateful 版本只讀盤點、API 與 governance UI。定期排程、外部版本查詢、工具安裝、CI 變更、套件升級、主機更新、container pull、實際 PR creation、auto merge、Telegram 實發、SSH、kubectl、重啟仍未開 gate | `ai_agent_proactive_operations_contract_v1`、`ai_agent_version_freshness_snapshot_v1`、`ai_agent_tool_adoption_approval_package_v1`、`ai_agent_telegram_action_required_digest_policy_v1`、`ai_agent_gitea_pr_draft_lane_v1`、`ai_agent_host_stateful_version_inventory_v1`、`GET /api/v1/agents/agent-proactive-operations-contract`、`GET /api/v1/agents/agent-version-freshness-snapshot`、`GET /api/v1/agents/agent-tool-adoption-approval-package`、`GET /api/v1/agents/agent-telegram-action-required-digest-policy`、`GET /api/v1/agents/agent-gitea-pr-draft-lane`、`GET /api/v1/agents/agent-host-stateful-version-inventory`、`/zh-TW/governance?tab=automation-inventory`、MASTER §3.2.1c | | 12-Agent War Room 編組 | 72% | 12 個邏輯工位與分批派工規則已正式部署；OpenClaw / Hermes / NemoTron / SRE / Security / DevOps / Data/DR / Supply Chain / Product/UI / QA / Market / Telegram 共 12 份只讀審查已回收；schema / committed snapshot / API / tests / governance UI 區塊 / production API readback / desktop + mobile in-app browser smoke 已完成；runtime writer、Telegram send、Bot API、production write 仍未批准 | `ai_agent_12_agent_war_room_v1`、`docs/evaluations/ai_agent_12_agent_war_room_2026-06-14.json`、`GET /api/v1/agents/agent-12-agent-war-room`、feature commit `5de4b3f3`、deploy marker `1a2c9e36`、Gitea CD run `4232` success、`/zh-TW/governance?tab=automation-inventory`、12 份 Codex sub-agent 只讀回饋 | +| AI Agent 專業任務擴展與 Telegram Runtime Bridge | 84% | P2-405A 已完成本地只讀契約與治理頁 P2-405A 卡片；24 類專業任務、8 個領域、5 段 Telegram bridge、6 種訊息類型、MCP/RAG stack、日報 / 週報 / 月報 / action-required 報告契約已固定；Telegram 實發、Gateway queue、Bot API、delivery receipt、production write、secret read、paid API、host write、kubectl action 仍全部關閉 | `ai_agent_professional_task_expansion_v1`、`docs/evaluations/ai_agent_professional_task_expansion_2026-06-15.json`、`GET /api/v1/agents/agent-professional-task-expansion`、`/zh-TW/governance?tab=automation-inventory`、`docs/ai/AI_AGENT_PROFESSIONAL_TASK_EXPANSION_2026-06-15.md`、需批准任務 `19`、low / medium / high / critical = `3 / 10 / 6 / 5`；下一步 P2-405B 完整 no-send message preview / canary approval package | | Owner response 預檢與拒收邊界 | 100% | P2-143 已完成正式部署與 production readback；承接 P2-141 input prep 與 P2-142 War Room，只建立 owner / verifier / rollback / maintenance / live-apply 五類外部回覆的 intake 預檢、必填欄位與拒收規則；正式 owner response 尚未收到、未接受、未寫入 | `ai_agent_result_capture_release_decision_owner_response_preflight_v1`、`GET /api/v1/agents/agent-result-capture-release-decision-owner-response-preflight`、feature commit `755b0a8d`、deploy marker `667d6329`、Gitea code-review `2961` / CD `2960` success、5 個 response intake lane、18 個 required owner field、6 個 validation check、6 個 rejection guard、5 個 operator action；owner response received / accepted / redacted payload / reviewer queue / Gateway / Telegram / Bot API / production write / secret read / destructive operation 全為 `0` | | Owner response 回讀狀態 | 100% | P2-144 已完成正式部署與 production readback；承接 P2-143 preflight，只讀回五類外部回覆仍未收到、未接受、未拒絕、未保存 | `ai_agent_result_capture_release_decision_owner_response_readback_v1`、`GET /api/v1/agents/agent-result-capture-release-decision-owner-response-readback`、feature commit `8795f100`、deploy marker `ac938037`、Gitea code-review `2965` / CD `2964` success、5 個 response readback lane、18 個 required owner field、6 個 readback validation check、6 個 readback rejection guard、5 個 operator action、waiting external response `5`、no external response received `5`；owner response received / accepted / redacted payload / reviewer queue / Gateway / Telegram / Bot API / production write / secret read / destructive operation 全為 `0` | | 本工作清單與分析報告 | 100% | 已完成 | 本 MD 文件 | @@ -1416,7 +1417,8 @@ UI： ## 13. 立即執行順序 1. P2-004：依賴 / 供應鏈漂移監控，保持只讀觀察與批准包邊界。 -2. P3-001：外部 Agent / SDK / API 相關能力仍需證據、費用批准與 shadow / canary 關卡。 +2. P2-405B：AI Agent Telegram 完整 no-send message preview / canary approval package，先讓統帥看見未來 TG 訊息、dedup、receipt 與風險標籤長相；仍不得實發。 +3. P3-001：外部 Agent / SDK / API 相關能力仍需證據、費用批准與 shadow / canary 關卡。 ## 14. 目前風險 diff --git a/docs/ai/AI_AGENT_PROFESSIONAL_TASK_EXPANSION_2026-06-15.md b/docs/ai/AI_AGENT_PROFESSIONAL_TASK_EXPANSION_2026-06-15.md new file mode 100644 index 00000000..c8a428e4 --- /dev/null +++ b/docs/ai/AI_AGENT_PROFESSIONAL_TASK_EXPANSION_2026-06-15.md @@ -0,0 +1,78 @@ +# AI Agent 專業任務擴展與 Telegram Runtime Bridge 工作報告 + +> 日期：2026-06-15（台北時間） +> 狀態：P2-405A 已完成只讀契約、snapshot、API、測試與治理頁卡片；Telegram 實發仍未啟用。 +> 事實來源：`ai_agent_professional_task_expansion_v1` + +## 1. 結論 + +本輪把「AI Agent 還能處理哪些專業工作」正式產品化成 24 類專業任務，並把 Telegram 群組 / TG Bot 整合拆成 5 段啟動前閘門。 + +這不是直接讓 AI Agent 發 Telegram 或改 production；目前只允許 no-send preview、queue preview readback、owner review 與批准包。真正送到 **AwoooI SRE 戰情室** 必須先通過 approved canary、dedup、receipt、redaction、OpenClaw 仲裁、Security gate 與 QA verifier。 + +## 2. 完成度 + +| 範圍 | 完成度 | 狀態 | +|---|---:|---| +| 專業任務擴展契約 | 100% | 24 類任務已定義 owner、風險、MCP、RAG、Telegram policy 與 blocked actions | +| Telegram Runtime Bridge 契約 | 100% | no-send preview、queue preview、approved canary、日週月報、action-required digest 已分段 | +| API / loader | 100% | `GET /api/v1/agents/agent-professional-task-expansion` 只讀輸出 | +| 治理頁可視化 | 100% | `/zh-TW/governance?tab=automation-inventory` 顯示任務、風險、TG bridge 與 live/send/write=0 | +| Telegram 實發 | 0% | `telegram_send_count=0`、`bot_api_call_count=0`、`gateway_queue_write_count=0` | +| Runtime 自動優化 | 0% | production write、host write、kubectl、paid API、secret read 全部維持 0 | + +## 3. Telegram Bridge + +| 階段 | 狀態 | 可做 | 不可做 | +|---|---|---|---| +| no-send preview | ready for owner review | 產生脫敏 TG 訊息預覽、dedup key、風險層級 | 不送 Telegram、不寫 queue | +| queue preview readback | ready for owner review | 讀回 queue envelope hash、receipt expectation | 不寫 production queue | +| approved canary | blocked until approval | 單一 canary 訊息與 receipt 驗證 | 未批准不得 Bot API send | +| 日週月報 digest | blocked until canary receipt | Agent 工作量、完成度、阻擋項、圖表摘要 | 不洗版、不發成功噪音 | +| action-required digest | blocked until dual gate | 高風險審核卡、failure-only escalation | 不取代 runtime authorization | + +## 4. 專業任務總覽 + +| 領域 | 任務數 | 代表任務 | 主責 | +|---|---:|---|---| +| SRE / Observability | 4 | incident correlation、告警降噪、synthetic check、DB slow query triage | SRE Sentinel / QA | +| Security / IwoooS | 3 | public bundle redaction、secret metadata parity、owner response completeness | Security Sentinel | +| DevOps / Release | 4 | image digest pin、GitOps diff、gateway route impact、candidate PR draft | DevOps Commander | +| Data / Backup / DR | 3 | backup readiness、escrow evidence、data freshness | Data DR / Hermes | +| Supply Chain | 2 | CVE triage、SBOM/license drift | Supply Chain Scout | +| Product / UI / QA | 1 | mobile overflow、console、a11y regression | Product UI / QA | +| AI Governance / Replay / Market | 4 | market watch、NemoTron replay、cost forecast、runbook/postmortem | OpenClaw / NemoTron / Hermes | +| Telegram / Reports / Receipts | 3 | digest preview、report truth gate、post-action verifier | Telegram Ops / Hermes / OpenClaw | + +## 5. 專業能力層級 + +| 層級 | AI Agent 可自動做 | Gate | +|---|---|---| +| 低風險 | 文件、報表、UI smoke、只讀摘要、truth gate | guard 後自動摘要 | +| 中風險 | proposal、no-write dry-run、PR 草案、Telegram preview | QA + OpenClaw 或 owner review | +| 高風險 | 只產批准包、rollback plan、failure-only digest 草案 | 統帥批准 | +| Critical | production write、kubectl、ArgoCD sync、Telegram 實發、secret、restore、host write | 預設 blocked | + +## 6. MCP / RAG + +首批 MCP：Gitea、Browser、Observability、Telegram Gateway、Package Registry、Database Readonly、Backup Status、ArgoCD Readonly、HTTP Probe、Fixture Store。 + +首批 RAG：LOGBOOK chunks、HARD_RULES / ADR chunks、Runbook chunks、Incident / alert traces、Agent market evidence、Owner response templates。 + +成長指標：KM entries、PlayBook updates、recommendations、replay score delta、blocked action prevented count、receipt missing count。 + +## 7. 邊界 + +- 不直接發 Telegram。 +- 不寫 Telegram Gateway queue。 +- 不呼叫 Bot API。 +- 不讀 `TELEGRAM_BOT_TOKEN` 或任何機密明文。 +- 不把工作視窗對話、未遮罩提示、私人推理或未遮罩 runtime payload 放進前端或 Telegram。 +- 不做 production write、host write、kubectl、ArgoCD sync、restore、rollback、paid API、SDK install。 + +## 8. 下一步 + +1. P2-405B：把完整 no-send message preview、dedup key、receipt expectation 與 canary approval package 顯示到治理頁，讓統帥先看 TG 訊息會長什麼樣。 +2. P2-405C：產生 approved canary approval package，明確列出發送時間窗、目標群組、dedup、receipt 與 rollback / mute plan。 +3. P2-405D：canary 通過後才開日報 / 週報 / 月報 digest delivery。 +4. P2-405E：Action-required digest 只對 failure / high-risk / approval-required 事件開啟。 diff --git a/docs/evaluations/ai_agent_professional_task_expansion_2026-06-15.json b/docs/evaluations/ai_agent_professional_task_expansion_2026-06-15.json new file mode 100644 index 00000000..3447ad62 --- /dev/null +++ b/docs/evaluations/ai_agent_professional_task_expansion_2026-06-15.json @@ -0,0 +1,244 @@ +{ + "schema_version": "ai_agent_professional_task_expansion_v1", + "generated_at": "2026-06-15T10:40:00+08:00", + "program_status": { + "current_priority": "P2", + "current_task_id": "P2-405A", + "next_task_id": "P2-405B", + "overall_completion_percent": 82, + "read_only_mode": true, + "runtime_authority": "professional_task_expansion_and_telegram_bridge_read_only_no_send", + "status_note": "AI Agent 專業任務擴展與 Telegram Runtime Bridge 已產品化為只讀契約；目前只允許 no-send preview、queue preview readback 與 owner review，不寫 Gateway queue、不送 Telegram、不呼叫 Bot API、不讀 secret、不做 production write。" + }, + "source_refs": [ + "docs/ai/AI_AGENT_12_AGENT_WAR_ROOM_2026-06-14.md", + "docs/ai/AI_AGENT_PROACTIVE_OPERATIONS_2026-06-11.md", + "docs/ai/AI_AGENT_INTERACTION_LEARNING_PROOF_2026-06-11.md", + "docs/ai/AI_AGENT_TELEGRAM_ACTION_REQUIRED_DIGEST_POLICY_2026-06-11.md", + "docs/ai/AI_AGENT_DEPLOYMENT_LAYOUT_2026-06-11.md", + "docs/security/MONITORING-OWNER-RESPONSE-ACCEPTANCE.md", + "apps/api/src/services/telegram_gateway.py", + ".gitea/workflows/cd.yaml", + ".gitea/workflows/code-review.yaml" + ], + "telegram_runtime_bridge": { + "canonical_room": "AwoooI SRE 戰情室", + "canonical_room_env": "SRE_GROUP_CHAT_ID", + "gateway_required": true, + "gateway_owner": "telegram_ops_liaison", + "arbiter": "openclaw", + "receipt_owner": "hermes", + "replay_owner": "nemotron", + "no_send_preview_ready": true, + "queue_preview_readback_ready": true, + "approved_canary_required": true, + "direct_bot_api_allowed": false, + "bot_api_call_enabled": false, + "gateway_queue_write_enabled": false, + "telegram_send_enabled": false, + "delivery_receipt_write_enabled": false, + "stages": [ + { + "stage_id": "tg_bridge_01_no_send_preview", + "title": "Telegram no-send preview", + "status": "ready_for_owner_review", + "allowed_output": "脫敏訊息預覽、dedup key、目標群組 metadata、風險層級、需審核選項", + "live_send_enabled": false, + "exit_condition": "OpenClaw + Security + QA 確認不含 secret、raw payload、工作視窗內容與誤導性執行按鈕。" + }, + { + "stage_id": "tg_bridge_02_queue_preview_readback", + "title": "Gateway queue preview readback", + "status": "ready_for_owner_review", + "allowed_output": "queue envelope hash、message type、receipt expectation、rate-limit / dedup plan", + "live_send_enabled": false, + "exit_condition": "Hermes 讀回 preview hash，QA 確認不會寫入 production queue。" + }, + { + "stage_id": "tg_bridge_03_approved_canary", + "title": "Approved canary delivery", + "status": "blocked_until_explicit_approval", + "allowed_output": "單一 approved canary 訊息與 receipt 驗證", + "live_send_enabled": false, + "exit_condition": "統帥明確批准 canary、指定時間窗與 rollback / mute plan。" + }, + { + "stage_id": "tg_bridge_04_daily_weekly_monthly_digest", + "title": "日報 / 週報 / 月報 digest delivery", + "status": "blocked_until_canary_receipt", + "allowed_output": "Agent 工作量、完成度、阻擋項、圖表摘要與 low-noise digest", + "live_send_enabled": false, + "exit_condition": "canary receipt 正常、dedup 有效、成功訊息壓制策略通過。" + }, + { + "stage_id": "tg_bridge_05_action_required_digest", + "title": "Action-required / high-risk approval digest", + "status": "blocked_until_dual_gate", + "allowed_output": "高風險審核卡、owner response 缺口、failure-only escalation", + "live_send_enabled": false, + "exit_condition": "OpenClaw 仲裁、Security gate、QA verifier 與 Telegram receipt 全通過。" + } + ], + "message_types": [ + { + "message_type": "daily_agent_workload_digest", + "cadence": "daily", + "risk_tier": "low", + "owner_agent": "hermes", + "send_policy": "send_after_canary_and_dedup_gate" + }, + { + "message_type": "weekly_governance_summary", + "cadence": "weekly", + "risk_tier": "medium", + "owner_agent": "hermes", + "send_policy": "send_after_canary_and_truth_gate" + }, + { + "message_type": "monthly_growth_report", + "cadence": "monthly", + "risk_tier": "medium", + "owner_agent": "hermes", + "send_policy": "send_after_owner_review" + }, + { + "message_type": "high_risk_approval_card", + "cadence": "event", + "risk_tier": "high", + "owner_agent": "openclaw", + "send_policy": "approval_required_before_send" + }, + { + "message_type": "failure_only_escalation", + "cadence": "event", + "risk_tier": "critical", + "owner_agent": "sre_sentinel", + "send_policy": "failure_only_after_route_lock" + }, + { + "message_type": "report_receipt_gap_alert", + "cadence": "event", + "risk_tier": "medium", + "owner_agent": "telegram_ops_liaison", + "send_policy": "no_send_preview_until_receipt_contract" + } + ] + }, + "professional_task_domains": [ + {"domain_id": "sre_observability", "label": "SRE / Observability", "owner_agent": "sre_sentinel"}, + {"domain_id": "security_iwooos", "label": "Security / IwoooS", "owner_agent": "security_sentinel"}, + {"domain_id": "devops_release", "label": "DevOps / Release", "owner_agent": "devops_commander"}, + {"domain_id": "data_dr", "label": "Data / Backup / DR", "owner_agent": "data_dr_guardian"}, + {"domain_id": "supply_chain", "label": "Supply Chain / Versions", "owner_agent": "supply_chain_scout"}, + {"domain_id": "product_quality", "label": "Product / UI / QA", "owner_agent": "product_ui_curator"}, + {"domain_id": "ai_governance", "label": "AI Governance / Replay / Market", "owner_agent": "openclaw"}, + {"domain_id": "telegram_reporting", "label": "Telegram / Reports / Receipts", "owner_agent": "telegram_ops_liaison"} + ], + "professional_tasks": [ + {"task_id": "pro_task_001_incident_correlation", "title": "跨來源 incident correlation 與根因候選排序", "domain_id": "sre_observability", "owner_agent": "sre_sentinel", "supporting_agents": ["openclaw", "hermes"], "risk_tier": "medium", "automation_mode": "no_write_dry_run", "telegram_policy": "action_required_preview", "required_mcp": ["observability", "gitea", "browser"], "required_rag": ["incident_traces", "runbooks"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["alert_rule_change", "silence_change", "restart_service"]}, + {"task_id": "pro_task_002_alert_noise_budget", "title": "告警噪音預算與降噪提案", "domain_id": "sre_observability", "owner_agent": "sre_sentinel", "supporting_agents": ["hermes", "openclaw"], "risk_tier": "medium", "automation_mode": "proposal_only", "telegram_policy": "weekly_digest", "required_mcp": ["prometheus", "alertmanager"], "required_rag": ["alert_traces", "runbooks"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["receiver_route_change", "silence_policy_change", "prometheus_reload"]}, + {"task_id": "pro_task_003_synthetic_check_plan", "title": "公開路由 synthetic check 與失敗分類計畫", "domain_id": "sre_observability", "owner_agent": "qa_verifier", "supporting_agents": ["product_ui_curator", "sre_sentinel"], "risk_tier": "low", "automation_mode": "auto_summary_after_guard", "telegram_policy": "daily_digest", "required_mcp": ["browser", "http_probe"], "required_rag": ["route_runbooks"], "approval_required": false, "current_live_count_24h": 0, "blocked_actions": ["active_scan", "production_write"]}, + {"task_id": "pro_task_004_backup_restore_readiness", "title": "備份 freshness、restore drill 與 DR readiness 報告", "domain_id": "data_dr", "owner_agent": "data_dr_guardian", "supporting_agents": ["hermes", "openclaw"], "risk_tier": "high", "automation_mode": "approval_required_before_execution", "telegram_policy": "action_required_preview", "required_mcp": ["backup_status", "gitea"], "required_rag": ["dr_runbooks"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["restore_run", "offsite_sync", "credential_escrow_marker_write"]}, + {"task_id": "pro_task_005_non_secret_escrow_evidence", "title": "credential escrow 非敏感證據補件路由", "domain_id": "data_dr", "owner_agent": "data_dr_guardian", "supporting_agents": ["security_sentinel", "hermes"], "risk_tier": "critical", "automation_mode": "blocked_until_owner_response", "telegram_policy": "action_required_preview", "required_mcp": ["gitea"], "required_rag": ["dr_runbooks", "hard_rules"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["secret_value_collection", "credential_read", "marker_write_without_owner"]}, + {"task_id": "pro_task_006_dependency_cve_triage", "title": "依賴 CVE / EOL / license 風險分流", "domain_id": "supply_chain", "owner_agent": "supply_chain_scout", "supporting_agents": ["openclaw", "security_sentinel"], "risk_tier": "medium", "automation_mode": "proposal_only", "telegram_policy": "weekly_digest", "required_mcp": ["package_registry", "osv"], "required_rag": ["dependency_policy"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["package_upgrade", "lockfile_write", "external_scan_without_approval"]}, + {"task_id": "pro_task_007_sbom_license_drift", "title": "SBOM / license / Docker surface drift 批准包", "domain_id": "supply_chain", "owner_agent": "supply_chain_scout", "supporting_agents": ["hermes", "security_sentinel"], "risk_tier": "medium", "automation_mode": "proposal_only", "telegram_policy": "weekly_digest", "required_mcp": ["syft", "trivy"], "required_rag": ["supply_chain_runbooks"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["tool_install", "image_pull", "sbom_upload"]}, + {"task_id": "pro_task_008_image_digest_pin_plan", "title": "Docker image digest pinning 與 rollback 計畫", "domain_id": "supply_chain", "owner_agent": "devops_commander", "supporting_agents": ["supply_chain_scout", "qa_verifier"], "risk_tier": "high", "automation_mode": "approval_required_before_execution", "telegram_policy": "action_required_preview", "required_mcp": ["registry", "gitea"], "required_rag": ["release_runbooks"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["image_pull", "image_push", "kustomization_write_without_review"]}, + {"task_id": "pro_task_009_gitops_diff_review", "title": "K8s / ArgoCD manifest diff 與 blast-radius review", "domain_id": "devops_release", "owner_agent": "devops_commander", "supporting_agents": ["openclaw", "qa_verifier"], "risk_tier": "critical", "automation_mode": "approval_required_before_execution", "telegram_policy": "action_required_preview", "required_mcp": ["gitea", "argocd"], "required_rag": ["gitops_runbooks"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["argocd_sync", "kubectl_apply", "helm_upgrade"]}, + {"task_id": "pro_task_010_gateway_route_impact", "title": "Nginx / public gateway route impact analysis", "domain_id": "devops_release", "owner_agent": "devops_commander", "supporting_agents": ["security_sentinel", "qa_verifier"], "risk_tier": "critical", "automation_mode": "approval_required_before_execution", "telegram_policy": "action_required_preview", "required_mcp": ["gitea", "browser"], "required_rag": ["gateway_runbooks"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["nginx_reload", "route_change", "certbot_renew"]}, + {"task_id": "pro_task_011_public_bundle_redaction", "title": "前台 bundle / DOM / i18n 敏感資訊防洩漏掃描", "domain_id": "security_iwooos", "owner_agent": "security_sentinel", "supporting_agents": ["product_ui_curator", "qa_verifier"], "risk_tier": "medium", "automation_mode": "no_write_dry_run", "telegram_policy": "daily_digest", "required_mcp": ["browser", "gitea"], "required_rag": ["hard_rules"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["frontend_deploy_without_smoke", "secret_storage"]}, + {"task_id": "pro_task_012_secret_metadata_parity", "title": "Secret name parity、injection route 與 redaction gate", "domain_id": "security_iwooos", "owner_agent": "security_sentinel", "supporting_agents": ["devops_commander", "openclaw"], "risk_tier": "critical", "automation_mode": "blocked_until_owner_response", "telegram_policy": "action_required_preview", "required_mcp": ["gitea"], "required_rag": ["secret_runbooks"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["secret_value_read", "secret_rotation", "workflow_modification"]}, + {"task_id": "pro_task_013_telegram_digest_preview", "title": "AI Agent 日週月報 Telegram no-send preview", "domain_id": "telegram_reporting", "owner_agent": "telegram_ops_liaison", "supporting_agents": ["hermes", "qa_verifier"], "risk_tier": "medium", "automation_mode": "no_write_dry_run", "telegram_policy": "no_direct_telegram", "required_mcp": ["telegram_gateway"], "required_rag": ["logbook_chunks"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["telegram_send", "bot_api_call", "gateway_queue_write"]}, + {"task_id": "pro_task_014_report_truth_gate", "title": "日報 / 週報 / 月報 truth gate 與可處置性審查", "domain_id": "telegram_reporting", "owner_agent": "hermes", "supporting_agents": ["openclaw", "sre_sentinel"], "risk_tier": "medium", "automation_mode": "auto_summary_after_guard", "telegram_policy": "weekly_digest", "required_mcp": ["gitea", "observability"], "required_rag": ["logbook_chunks", "incident_traces"], "approval_required": false, "current_live_count_24h": 0, "blocked_actions": ["false_green_report", "report_send_without_truth_gate"]}, + {"task_id": "pro_task_015_agent_market_watch", "title": "主流 AI Agent / SDK / 模型版本市場雷達", "domain_id": "ai_governance", "owner_agent": "market_scout", "supporting_agents": ["openclaw", "nemotron"], "risk_tier": "medium", "automation_mode": "proposal_only", "telegram_policy": "weekly_digest", "required_mcp": ["web_primary_sources", "gitea"], "required_rag": ["agent_market_evidence"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["sdk_install", "paid_api_call", "production_route_change"]}, + {"task_id": "pro_task_016_model_replay_scorecard", "title": "NemoTron / 候選模型 replay scorecard 與 failure mode 分析", "domain_id": "ai_governance", "owner_agent": "nemotron", "supporting_agents": ["openclaw", "qa_verifier"], "risk_tier": "high", "automation_mode": "no_write_dry_run", "telegram_policy": "action_required_preview", "required_mcp": ["fixture_store", "gitea"], "required_rag": ["replay_fixtures"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["shadow_route", "canary_route", "full_replay_without_approval"]}, + {"task_id": "pro_task_017_cost_capacity_forecast", "title": "AI provider、主機、K3s 與外部服務成本 / 容量預測", "domain_id": "ai_governance", "owner_agent": "openclaw", "supporting_agents": ["supply_chain_scout", "sre_sentinel"], "risk_tier": "high", "automation_mode": "proposal_only", "telegram_policy": "monthly_digest", "required_mcp": ["observability", "billing_metadata"], "required_rag": ["cost_policy"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["paid_provider_switch", "limit_increase", "paid_api_call"]}, + {"task_id": "pro_task_018_db_query_triage", "title": "PostgreSQL 慢查詢、索引與資料新鮮度異常 triage", "domain_id": "sre_observability", "owner_agent": "sre_sentinel", "supporting_agents": ["data_dr_guardian", "openclaw"], "risk_tier": "high", "automation_mode": "proposal_only", "telegram_policy": "action_required_preview", "required_mcp": ["database_readonly", "observability"], "required_rag": ["db_runbooks"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["migration", "reindex", "write_query"]}, + {"task_id": "pro_task_019_data_freshness_source_reliability", "title": "資料來源新鮮度、fallback 與來源可信度監控", "domain_id": "data_dr", "owner_agent": "hermes", "supporting_agents": ["sre_sentinel", "qa_verifier"], "risk_tier": "medium", "automation_mode": "auto_summary_after_guard", "telegram_policy": "daily_digest", "required_mcp": ["http_probe", "database_readonly"], "required_rag": ["data_quality_runbooks"], "approval_required": false, "current_live_count_24h": 0, "blocked_actions": ["source_switch_without_owner", "data_write"]}, + {"task_id": "pro_task_020_ui_accessibility_regression", "title": "前後台 mobile overflow、console error、可讀性與 a11y regression", "domain_id": "product_quality", "owner_agent": "product_ui_curator", "supporting_agents": ["qa_verifier", "security_sentinel"], "risk_tier": "low", "automation_mode": "auto_summary_after_guard", "telegram_policy": "daily_digest", "required_mcp": ["browser"], "required_rag": ["frontend_guidelines"], "approval_required": false, "current_live_count_24h": 0, "blocked_actions": ["unrelated_redesign", "public_secret_display"]}, + {"task_id": "pro_task_021_runbook_postmortem_draft", "title": "Runbook、postmortem 與 owner handoff 草稿", "domain_id": "ai_governance", "owner_agent": "hermes", "supporting_agents": ["openclaw", "sre_sentinel"], "risk_tier": "low", "automation_mode": "auto_summary_after_guard", "telegram_policy": "weekly_digest", "required_mcp": ["gitea"], "required_rag": ["logbook_chunks", "runbooks"], "approval_required": false, "current_live_count_24h": 0, "blocked_actions": ["canonical_km_write_without_review", "secret_in_doc"]}, + {"task_id": "pro_task_022_owner_response_completeness", "title": "Owner response 欄位完整性、補件與拒收分流", "domain_id": "security_iwooos", "owner_agent": "security_sentinel", "supporting_agents": ["hermes", "openclaw"], "risk_tier": "high", "automation_mode": "proposal_only", "telegram_policy": "action_required_preview", "required_mcp": ["gitea"], "required_rag": ["owner_response_templates"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["mark_accepted_without_reviewer", "runtime_gate_open"]}, + {"task_id": "pro_task_023_candidate_pr_draft_packaging", "title": "低 / 中風險候選 PR 草案、測試與 rollback 封包", "domain_id": "devops_release", "owner_agent": "devops_commander", "supporting_agents": ["qa_verifier", "openclaw"], "risk_tier": "medium", "automation_mode": "proposal_only", "telegram_policy": "action_required_preview", "required_mcp": ["gitea"], "required_rag": ["release_runbooks"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["branch_push_without_approval", "auto_merge", "workflow_trigger"]}, + {"task_id": "pro_task_024_post_action_verifier", "title": "Post-action verifier、rollback trigger 與 receipt reconciliation", "domain_id": "telegram_reporting", "owner_agent": "openclaw", "supporting_agents": ["telegram_ops_liaison", "qa_verifier"], "risk_tier": "critical", "automation_mode": "approval_required_before_execution", "telegram_policy": "failure_only_preview", "required_mcp": ["telegram_gateway", "observability", "gitea"], "required_rag": ["verification_runbooks"], "approval_required": true, "current_live_count_24h": 0, "blocked_actions": ["verifier_live_readback_without_gate", "rollback_without_owner", "telegram_send_without_receipt"] } + ], + "mcp_rag_stack": { + "mcp_connectors": [ + "gitea", + "browser", + "observability", + "telegram_gateway", + "package_registry", + "database_readonly", + "backup_status", + "argocd_readonly", + "http_probe", + "fixture_store" + ], + "rag_sources": [ + "LOGBOOK chunks", + "HARD_RULES / ADR chunks", + "runbook chunks", + "incident / alert traces", + "agent market evidence", + "owner response templates" + ], + "growth_metrics": [ + "km_entries_created", + "playbook_updates", + "recommendations_created", + "replay_score_delta", + "blocked_action_prevented_count", + "receipt_missing_count" + ] + }, + "risk_policy": { + "low": "文件、報表、UI smoke 與只讀摘要可在 guard 後自動處理並進每日 digest。", + "medium": "中風險只能先產 proposal、no-write dry-run 或 PR 草案；涉及設定、外部服務、成本、路由或未知狀態時升級審核。", + "high": "高風險必須 OpenClaw 仲裁、QA / Security gate 與統帥批准；不得自動執行。", + "critical": "production write、kubectl、ArgoCD sync、Telegram 實發、secret、付費 API、restore、rollback、host write 均維持 blocked until explicit approval。" + }, + "reporting_contract": { + "daily": {"required": true, "owner_agent": "hermes", "telegram_stage": "tg_bridge_04_daily_weekly_monthly_digest"}, + "weekly": {"required": true, "owner_agent": "hermes", "telegram_stage": "tg_bridge_04_daily_weekly_monthly_digest"}, + "monthly": {"required": true, "owner_agent": "hermes", "telegram_stage": "tg_bridge_04_daily_weekly_monthly_digest"}, + "action_required": {"required": true, "owner_agent": "openclaw", "telegram_stage": "tg_bridge_05_action_required_digest"} + }, + "redaction_contract": { + "redaction_required": true, + "conversation_transcript_display_allowed": false, + "raw_prompt_display_allowed": false, + "private_reasoning_display_allowed": false, + "secret_value_display_allowed": false, + "raw_runtime_payload_display_allowed": false, + "telegram_message_must_be_sanitized": true, + "forbidden_terms": [ + "work_window_transcript", + "raw prompt", + "private reasoning", + "chain-of-thought", + "telegram token", + "authorization header", + "secret value" + ] + }, + "rollups": { + "professional_task_count": 24, + "domain_count": 8, + "telegram_stage_count": 5, + "telegram_message_type_count": 6, + "approval_required_count": 19, + "low_risk_task_count": 3, + "medium_risk_task_count": 10, + "high_risk_task_count": 6, + "critical_risk_task_count": 5, + "current_live_count": 0, + "gateway_queue_write_count": 0, + "telegram_send_count": 0, + "bot_api_call_count": 0, + "delivery_receipt_write_count": 0, + "production_write_count": 0, + "secret_read_count": 0, + "paid_api_call_count": 0, + "host_write_count": 0, + "kubectl_action_count": 0 + } +} diff --git a/docs/schemas/ai_agent_professional_task_expansion_v1.schema.json b/docs/schemas/ai_agent_professional_task_expansion_v1.schema.json new file mode 100644 index 00000000..c7dedfb3 --- /dev/null +++ b/docs/schemas/ai_agent_professional_task_expansion_v1.schema.json @@ -0,0 +1,329 @@ +{ + "$schema": "https://json-schema.org/draft/2020-12/schema", + "$id": "https://awoooi.wooo.work/schemas/ai_agent_professional_task_expansion_v1.schema.json", + "title": "AI Agent Professional Task Expansion And Telegram Runtime Bridge", + "type": "object", + "required": [ + "schema_version", + "generated_at", + "program_status", + "source_refs", + "telegram_runtime_bridge", + "professional_task_domains", + "professional_tasks", + "mcp_rag_stack", + "risk_policy", + "reporting_contract", + "redaction_contract", + "rollups" + ], + "properties": { + "schema_version": { + "const": "ai_agent_professional_task_expansion_v1" + }, + "generated_at": { + "type": "string", + "format": "date-time" + }, + "program_status": { + "type": "object", + "required": [ + "current_priority", + "current_task_id", + "next_task_id", + "overall_completion_percent", + "read_only_mode", + "runtime_authority", + "status_note" + ], + "properties": { + "current_priority": { + "enum": ["P0", "P1", "P2", "P3"] + }, + "current_task_id": { + "const": "P2-405A" + }, + "next_task_id": { + "const": "P2-405B" + }, + "overall_completion_percent": { + "type": "integer", + "minimum": 0, + "maximum": 100 + }, + "read_only_mode": { + "const": true + }, + "runtime_authority": { + "const": "professional_task_expansion_and_telegram_bridge_read_only_no_send" + }, + "status_note": { + "type": "string", + "minLength": 1 + } + }, + "additionalProperties": true + }, + "source_refs": { + "type": "array", + "minItems": 1, + "items": { + "type": "string" + } + }, + "telegram_runtime_bridge": { + "type": "object", + "required": [ + "canonical_room", + "canonical_room_env", + "gateway_required", + "no_send_preview_ready", + "queue_preview_readback_ready", + "approved_canary_required", + "direct_bot_api_allowed", + "bot_api_call_enabled", + "gateway_queue_write_enabled", + "telegram_send_enabled", + "delivery_receipt_write_enabled", + "stages", + "message_types" + ], + "properties": { + "canonical_room": { + "const": "AwoooI SRE 戰情室" + }, + "canonical_room_env": { + "const": "SRE_GROUP_CHAT_ID" + }, + "gateway_required": { + "const": true + }, + "no_send_preview_ready": { + "const": true + }, + "queue_preview_readback_ready": { + "const": true + }, + "approved_canary_required": { + "const": true + }, + "direct_bot_api_allowed": { + "const": false + }, + "bot_api_call_enabled": { + "const": false + }, + "gateway_queue_write_enabled": { + "const": false + }, + "telegram_send_enabled": { + "const": false + }, + "delivery_receipt_write_enabled": { + "const": false + }, + "stages": { + "type": "array", + "minItems": 5, + "items": { + "type": "object" + } + }, + "message_types": { + "type": "array", + "minItems": 6, + "items": { + "type": "object" + } + } + }, + "additionalProperties": true + }, + "professional_task_domains": { + "type": "array", + "minItems": 8, + "items": { + "type": "object" + } + }, + "professional_tasks": { + "type": "array", + "minItems": 24, + "items": { + "type": "object", + "required": [ + "task_id", + "title", + "domain_id", + "owner_agent", + "risk_tier", + "automation_mode", + "telegram_policy", + "required_mcp", + "required_rag", + "approval_required", + "current_live_count_24h", + "blocked_actions" + ], + "properties": { + "task_id": { + "type": "string" + }, + "title": { + "type": "string" + }, + "domain_id": { + "type": "string" + }, + "owner_agent": { + "type": "string" + }, + "risk_tier": { + "enum": ["low", "medium", "high", "critical"] + }, + "automation_mode": { + "enum": [ + "auto_summary_after_guard", + "no_write_dry_run", + "proposal_only", + "approval_required_before_execution", + "blocked_until_owner_response" + ] + }, + "telegram_policy": { + "enum": [ + "daily_digest", + "weekly_digest", + "monthly_digest", + "action_required_preview", + "failure_only_preview", + "no_direct_telegram" + ] + }, + "required_mcp": { + "type": "array", + "items": { + "type": "string" + } + }, + "required_rag": { + "type": "array", + "items": { + "type": "string" + } + }, + "approval_required": { + "type": "boolean" + }, + "current_live_count_24h": { + "const": 0 + }, + "blocked_actions": { + "type": "array", + "items": { + "type": "string" + } + } + }, + "additionalProperties": true + } + }, + "mcp_rag_stack": { + "type": "object" + }, + "risk_policy": { + "type": "object" + }, + "reporting_contract": { + "type": "object" + }, + "redaction_contract": { + "type": "object" + }, + "rollups": { + "type": "object", + "required": [ + "professional_task_count", + "domain_count", + "telegram_stage_count", + "telegram_message_type_count", + "approval_required_count", + "low_risk_task_count", + "medium_risk_task_count", + "high_risk_task_count", + "critical_risk_task_count", + "current_live_count", + "gateway_queue_write_count", + "telegram_send_count", + "bot_api_call_count", + "delivery_receipt_write_count", + "production_write_count", + "secret_read_count", + "paid_api_call_count", + "host_write_count", + "kubectl_action_count" + ], + "properties": { + "professional_task_count": { + "const": 24 + }, + "domain_count": { + "const": 8 + }, + "telegram_stage_count": { + "const": 5 + }, + "telegram_message_type_count": { + "const": 6 + }, + "approval_required_count": { + "const": 19 + }, + "low_risk_task_count": { + "const": 3 + }, + "medium_risk_task_count": { + "const": 10 + }, + "high_risk_task_count": { + "const": 6 + }, + "critical_risk_task_count": { + "const": 5 + }, + "current_live_count": { + "const": 0 + }, + "gateway_queue_write_count": { + "const": 0 + }, + "telegram_send_count": { + "const": 0 + }, + "bot_api_call_count": { + "const": 0 + }, + "delivery_receipt_write_count": { + "const": 0 + }, + "production_write_count": { + "const": 0 + }, + "secret_read_count": { + "const": 0 + }, + "paid_api_call_count": { + "const": 0 + }, + "host_write_count": { + "const": 0 + }, + "kubectl_action_count": { + "const": 0 + } + }, + "additionalProperties": true + } + }, + "additionalProperties": true +} 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 a719dc34..cb65970f 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 @@ -638,6 +638,7 @@ Alert / Sentry / SigNoz / Gitea / Market Watch / Operator | `docs/evaluations/ai_agent_report_runtime_dry_run_2026-06-12.json` + `GET /api/v1/agents/agent-report-runtime-dry-run` | P2-403M 報表 runtime no-write dry-run 證據包；建立 5 個 dry-run artifact、3 個 SRE 戰情室 queue digest 草案、4 個 readback verifier case、3 個 Agent dry-run role 與 6 個 operator checkpoint；不寫 Gateway queue、不送 Telegram、不呼叫 Bot API、不寫 delivery receipt、不啟動 worker、不跑 verifier live readback、不讀 secret，已由 P2-403N fixture readback 承接 | | `docs/evaluations/ai_agent_report_runtime_fixture_readback_2026-06-12.json` + `GET /api/v1/agents/agent-report-runtime-fixture-readback` | P2-403N fixture smoke / queue preview readback / verifier dry-run 證據包；建立 5 個 fixture smoke、3 個 SRE 戰情室 queue preview readback、4 個 verifier dry-run case、3 個 Agent fixture role 與 5 個 operator checkpoint；不寫 Gateway queue、不送 Telegram、不呼叫 Bot API、不寫 delivery receipt、不啟動 worker、不跑 verifier live readback、不讀 secret，下一步 P2-404 | | `docs/evaluations/ai_agent_runtime_worker_shadow_gate_2026-06-12.json` + `GET /api/v1/agents/agent-runtime-worker-shadow-gate` | P2-404 runtime worker shadow / no-write execution evidence gate；建立 5 個 shadow candidate、4 個 no-write replay、4 個 verifier shadow case、3 個 Agent shadow role 與 6 個 operator checkpoint；shadow live worker、Gateway queue write、Telegram send、Bot API、delivery receipt、auto worker、verifier live readback、production write 與 secret read 全部 `0 / false`，下一步 P2-101 | +| `docs/evaluations/ai_agent_professional_task_expansion_2026-06-15.json` + `GET /api/v1/agents/agent-professional-task-expansion` | P2-405A AI Agent 專業任務擴展與 Telegram Runtime Bridge；承接 12-Agent War Room、P2-403 report/runtime 鏈與 monitoring owner response acceptance，固定 24 類專業任務、8 個任務領域、5 段 Telegram bridge、6 種訊息類型、MCP/RAG stack、日報 / 週報 / 月報 / action-required 報告契約，並接入 governance automation inventory 卡片；需批准任務 `19`、low / medium / high / critical = `3 / 10 / 6 / 5`；Gateway queue write、Telegram send、Bot API、delivery receipt、production write、secret read、paid API、host write 與 kubectl action 全部 `0 / false`，下一步 P2-405B 完整 no-send message preview / canary approval package | | `docs/evaluations/ai_agent_operation_permission_model_2026-06-12.json` + `GET /api/v1/agents/agent-operation-permission-model` | P2-101 操作類別權限模型；建立 5 條 permission lane、13 類 operation category、3 個 Agent permission role、8 個 gate transition 與 5 個 operator decision template；runtime execution、Gateway queue write、Telegram send、Bot API、delivery receipt、auto worker、verifier live readback、production write、secret / paid provider、host command 與 destructive action 全部 `0 / false`，已由 P2-102 承接 | | `docs/evaluations/ai_agent_candidate_operation_dry_run_evidence_2026-06-12.json` + `GET /api/v1/agents/agent-candidate-operation-dry-run-evidence` | P2-102 候選操作 dry-run 證據；13 類候選操作全部具備 input / output evidence hash、side-effect count、verifier plan、rollback/no-op plan 與人工 handoff；6 個 verifier plan、7 個 gate evidence requirement、5 個 operator handoff；runtime、Gateway queue、Telegram、production write、secret / paid provider 與 destructive action 全部 `0 / false`，已由 P2-103 承接 | | `docs/evaluations/ai_agent_task_result_audit_trail_2026-06-13.json` + `GET /api/v1/agents/agent-task-result-audit-trail` | P2-103 任務結果稽核軌跡；8 條 result route、6 個 writeback contract、7 個 audit checkpoint、5 個 operator handoff；把 diagnostic-only、repair candidate、execution failed、provider unmatched、report zero-signal 等結果固定到 KM 草稿、LOGBOOK 證據、audit trail、timeline 與人工下一步；KM / LOGBOOK / audit DB / timeline / PlayBook trust / Gateway queue / Telegram 寫入全為 `0 / false`，已由 P2-104 承接 | @@ -840,6 +841,7 @@ Repo / registry / release notes / K8s / host / observability / backup evidence 64. 建立 owner response preflight 與拒收邊界。✅ P2-143 已完成正式驗證；承接 P2-141 decision input prep 與 P2-142 War Room 基線，固定 response intake lane `5`、required owner field `18`、intake validation check `6`、rejection guard `6`、operator action `5`、waiting external response `5`；owner response received / accepted / rejected、redacted payload ingested、reviewer queue write、Gateway queue write、Telegram send、Bot API、result capture、learning、PlayBook trust、production write、secret read、destructive operation 仍為 `0 / false`；feature commit `755b0a8d`、deploy marker `667d6329`、Gitea code-review `2961` / CD `2960` success、本地 P2-142 War Room + P2-139 至 P2-143 regression `37 passed`、JSON parse、Python compile、Web typecheck、guard、doc secret sanity、禁用外露值掃描、production API readback 與 desktop / mobile smoke 通過。下一步 P2-144 owner response readback。 65. 建立 owner response readback。✅ P2-144 已完成正式驗證；承接 P2-143 preflight，固定 response readback lane `5`、required owner field `18`、readback validation check `6`、readback rejection guard `6`、operator action `5`、waiting external response `5`、no external response received lane `5`；owner response received / accepted / rejected、redacted payload ingested、reviewer queue write、Gateway queue write、Telegram send、Bot API、result capture、learning、PlayBook trust、production write、secret read、destructive operation 仍為 `0 / false`；feature commit `8795f100`、deploy marker `ac938037`、Gitea code-review `2965` / CD `2964` success；本地 P2-139 至 P2-144 regression `45 passed`，rebase 後含 tenants regression 的推送前回歸 `47 passed`、JSON parse、Python compile、Web typecheck、guard、doc secret sanity、value-only 禁用外露值掃描、i18n key parity 與 diff check 通過；production API readback、desktop / mobile smoke、水平溢位 `0`、危險控制 `0`、工作溝通片語命中 `0` 已完成。下一步 P2-145 owner response acceptance gate。 66. 建立 owner response acceptance gate。✅ P2-145 已完成並正式驗證；承接 P2-144 readback，固定 acceptance gate lane `5`、required owner field `18`、acceptance validation check `6`、acceptance rejection guard `6`、operator action `5`、blocked no external response `5`、no acceptable external response `5`；owner response received / accepted / rejected、redacted payload ingested、reviewer queue write、Gateway queue write、Telegram send、Bot API、result capture、learning、PlayBook trust、production write、secret read、destructive operation 仍為 `0 / false`；feature commit `386dbd07`、deploy marker `36fbfc6b`、Gitea code-review `2969` / CD `2968` success；P2-144 + P2-145 regression `16 passed`、JSON parse、Python compile、Web typecheck、guard、doc secret sanity、顯示值工作視窗污染掃描、i18n key parity、diff check、production API readback、in-app browser smoke、desktop / mobile smoke、水平溢位 `0`、P2-145 卡片操作控制 `0`、工作溝通片語命中 `0` 已完成。下一步 P2-146 acceptance receipt preview，且必須等合格、遮罩、欄位完整、可驗證來源的外部正式回覆後才能建立 receipt preview。 +66a. 建立 AI Agent 專業任務擴展與 Telegram Runtime Bridge。✅ P2-405A 本地完成；承接 12-Agent War Room、P2-403 report/runtime 鏈與 monitoring owner response acceptance，固定 24 類專業任務、8 個領域、5 段 Telegram bridge、6 種訊息類型與 MCP/RAG stack，並接入 governance automation inventory 卡片；需批准任務 `19`、low / medium / high / critical = `3 / 10 / 6 / 5`；Gateway queue write、Telegram send、Bot API、delivery receipt、production write、secret read、paid API、host write、kubectl action 仍為 `0 / false`。下一步 P2-405B 完整 no-send message preview / canary approval package；這不占用 P2-146 owner response receipt preview。 67. 新增 P0 配置控管優先序前台可視化。✅ 正式驗證完成；`/zh-TW/iwooos` 已集中顯示 Nginx public gateway、DNS / TLS / certbot、K8s / ArgoCD / production manifests、Workflow / runner / secret metadata、Public / admin / API runtime config、agent-bounty runtime / treasury 六類即時風險配置；owner response `0 / 0`、live evidence `0`、執行期 `0`、操作按鈕 `0`；feature commit `e992af89`、deploy marker `ed651a98`、Gitea code-review `2971` / CD `2970` success；本地與正式 in-app browser、desktop `1440x1100`、mobile `390x844` smoke 通過。這不是 Nginx live conf 讀取、`nginx -t`、reload、DNS / TLS probe、certbot renew、ArgoCD sync、kubectl、workflow / secret 修改、public route change、agent-bounty runtime、payout / withdrawal、production write 或 runtime gate。 68. 補強 P0 高價值配置 Gate path pattern、工作樹 preflight、owner packet 與 coverage snapshot。✅ 本地完成；`k8s/nginx/**`、`scripts/ops/**/*cert*`、`scripts/ops/**/*tls*` 已納入 high-value config classification，Nginx public gateway 與 DNS / TLS / certbot sample 從 `matched=0 / C0=0` 收斂到 `matched=3 / C0=2`；gate snapshot 顯示 `changed_files=6`、`matched=6`、`categories=3`、`c0=2`、`c1=0`；預設模式已可讀取 staged / unstaged / untracked，臨時 `k8s/nginx/*` smoke 命中 C0；owner packet snapshot `packets=3 / c0=2 / runtime_gate=0`；coverage snapshot `categories=14 / c0=8 / avg=67 / runtime_gate=0`；owner evidence 仍 `provided=false / complete=false`，runtime execution 仍 `false`。這不是 live config read、`nginx -t`、reload、certbot renew、DNS / TLS probe、host write、active scan、workflow 修改、secret 收集、production write 或 runtime gate。 69. 同步高價值配置 Owner Packet 前台 projection。✅ 已完成並正式驗證；`/zh-TW/iwooos` 與 `/zh-TW/awooop` 已顯示 owner packet snapshot `packet=3 / c0=2`、最高命中 `C0 / P0`、Nginx public gateway、DNS / TLS / certbot 與 security tooling 影響範圍；feature commit `e999c16b`、deploy marker `16c6b983`、Gitea code-review `2973` / CD `2972` success；request sent、received、accepted、runtime gate 與 action buttons 仍為 `0`；本地與正式 desktop / mobile / in-app browser smoke 已通過，水平溢位 `0`、卡片內操作控制 `0`、危險連結 `0`、工作溝通片語命中 `0`。不得因此調高 IwoooS headline。 @@ -892,6 +894,7 @@ Repo / registry / release notes / K8s / host / observability / backup evidence | `docs/evaluations/ai_agent_report_runtime_dry_run_2026-06-12.json` + `GET /api/v1/agents/agent-report-runtime-dry-run` | P2-403M 報表 runtime no-write dry-run 證據包；5 個 dry-run artifact、3 個 queue digest 草案、4 個 readback verifier case、3 個 Agent dry-run role、6 個 operator checkpoint；不寫 Gateway queue、不送 Telegram、不呼叫 Bot API、不寫 delivery receipt、不啟動 worker、不跑 verifier live readback | | `docs/evaluations/ai_agent_report_runtime_fixture_readback_2026-06-12.json` + `GET /api/v1/agents/agent-report-runtime-fixture-readback` | P2-403N fixture smoke / queue preview readback / verifier dry-run 證據包；5 個 fixture smoke、3 個 queue preview readback、4 個 verifier dry-run case、3 個 Agent fixture role、5 個 operator checkpoint；不寫 Gateway queue、不送 Telegram、不呼叫 Bot API、不寫 delivery receipt、不啟動 worker、不跑 verifier live readback | | `docs/evaluations/ai_agent_runtime_worker_shadow_gate_2026-06-12.json` + `GET /api/v1/agents/agent-runtime-worker-shadow-gate` | P2-404 runtime worker shadow / no-write execution evidence gate；5 個 shadow candidate、4 個 no-write replay、4 個 verifier shadow case、3 個 Agent shadow role、6 個 operator checkpoint；不啟動 live worker、不寫 Gateway queue、不送 Telegram、不呼叫 Bot API、不寫 production target | +| `docs/evaluations/ai_agent_professional_task_expansion_2026-06-15.json` + `GET /api/v1/agents/agent-professional-task-expansion` | P2-405A AI Agent 專業任務擴展與 Telegram Runtime Bridge；24 類專業任務、8 個領域、5 段 Telegram bridge、6 種訊息類型、MCP/RAG stack、治理頁 P2-405A 卡片；不寫 Gateway queue、不送 Telegram、不呼叫 Bot API、不寫 production target、不讀 secret、不用 paid API | | `docs/evaluations/ai_agent_operation_permission_model_2026-06-12.json` + `GET /api/v1/agents/agent-operation-permission-model` | P2-101 操作類別權限模型；5 條 permission lane、13 類 operation category、3 個 Agent permission role、8 個 gate transition、5 個 operator decision template；不啟動 runtime worker、不寫 Gateway queue、不送 Telegram、不呼叫 Bot API、不寫 production target、不讀 secret | | `docs/evaluations/ai_agent_candidate_operation_dry_run_evidence_2026-06-12.json` + `GET /api/v1/agents/agent-candidate-operation-dry-run-evidence` | P2-102 候選操作 dry-run 證據；13 類候選操作、13 組 dry-run evidence、6 個 verifier plan、7 個 gate evidence requirement、5 個 operator handoff；不啟動 runtime worker、不寫 Gateway queue、不送 Telegram、不呼叫 Bot API、不寫 production target、不讀 secret、不執行 destructive action | | `docs/evaluations/ai_agent_task_result_audit_trail_2026-06-13.json` + `GET /api/v1/agents/agent-task-result-audit-trail` | P2-103 任務結果稽核軌跡；8 條 result route、6 個 writeback contract、7 個 audit checkpoint、5 個 operator handoff；不寫 KM、不 runtime append LOGBOOK、不寫 audit DB、不寫 timeline、不更新 PlayBook trust、不寫 Gateway queue、不送 Telegram | @@ -4977,3 +4980,16 @@ Trigger commit `f5cd37b7` 與 deploy marker `0ba92357` 已把 governance UI 的 - `iwooos-config-control-guard.py`、`security-mirror-progress-guard.py` 與高價值配置 coverage 已鎖住新 artifact；`monitoring_alerting_observability` 只讀治理成熟度 `62% -> 66%`，高價值配置平均只讀治理成熟度 `68% -> 69%`。 **裁決：** 這是 owner response acceptance 只讀帳本，不是 owner response received / accepted / rejected、live evidence、Prometheus reload、Alertmanager reload、Grafana import、SigNoz apply、Sentry deploy、Langfuse config change、OTEL reload、receiver route change、silence change、Telegram send、live alert fire、alert chain smoke、secret collection、host write、production write 或 runtime gate；IwoooS headline 仍維持 `64%`，active runtime gate 仍 `0`。 + +### 2026-06-15 10:40 (台北) — §3.2 / §5 — 新增 AI Agent 專業任務擴展與 Telegram Runtime Bridge — 把專業委派與 TG 群組整合前置門檻固定成 no-send guard artifact + +**觸發**：統帥要求繼續專業評估 AI Agent 還可以承接哪些工作，並指出 Telegram 群組 / TG Bot 需要真正納入整體監控、報表、告警與審核路徑。既有 12-Agent War Room 已定義分工，但還缺一份可由 API 讀回、可測試、可禁止誤啟用的專業任務擴展契約。 + +**已推進：** +- 新增 `docs/schemas/ai_agent_professional_task_expansion_v1.schema.json`、`docs/evaluations/ai_agent_professional_task_expansion_2026-06-15.json`、`apps/api/src/services/ai_agent_professional_task_expansion.py` 與 `GET /api/v1/agents/agent-professional-task-expansion`。 +- 新增 `docs/ai/AI_AGENT_PROFESSIONAL_TASK_EXPANSION_2026-06-15.md`，把 24 類專業任務拆成 SRE / Observability、Security / IwoooS、DevOps / Release、Data / Backup / DR、Supply Chain、Product / UI / QA、AI Governance / Replay / Market、Telegram / Reports / Receipts 8 個領域。 +- `/zh-TW/governance?tab=automation-inventory` 新增 P2-405A 卡片，顯示任務數、風險分層、Telegram bridge、脫敏狀態與 live / send / write = 0。 +- Telegram Runtime Bridge 固定 5 段：no-send preview、queue preview readback、approved canary、日週月報 digest、action-required digest；所有 live send / queue write / Bot API call / receipt write 仍為 `0 / false`。 +- 任務風險 rollup 固定：總任務 `24`、需批准 `19`、low / medium / high / critical = `3 / 10 / 6 / 5`；高風險與 critical 任務必須保留 approval gate。 + +**裁決：** 這是專業任務與 Telegram 整合的只讀擴展契約，不是 Telegram send、Gateway queue write、Bot API call、delivery receipt write、production write、host write、kubectl action、secret read、paid API call 或 runtime authorization；下一步 P2-405B 只能做完整 no-send message preview 與 canary approval package，可讓統帥看到將來 TG 訊息長相，但仍不得實發。