fix(tg+km): 補齊 Telegram 操作紀錄揭露與 KM 分類修復 (ADR-076)

Telegram 訊息新增欄位： - alert_category 分類標籤（🏷️ 主機/K8s/資料庫/服務等） - playbook_name 顯示匹配到的 Playbook 名稱 - 頻率統計從 count_24h>1 降至 >=1（初次告警也顯示） - TelegramMessage 新增 alert_category/playbook_name 欄位 - decision_manager → send_approval_card 穿透 playbook_name KM 修復： - EntryType.PLAYBOOK → EntryType.AUTO_RUNBOOK（前者不存在，會 AttributeError） - category "auto_generated" → "ai_system"（前端 i18n 有對應翻譯） - runbook_generator 同步修正 category - KM 建立後推 Telegram 通知（best-effort） DB decision_chain 補欄位： - 新增 playbook_id / playbook_name / alert_category 2026-04-16 ogt + Claude Sonnet 4.6 Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-16 02:46:06 +08:00
parent 44ecf609e0
commit 62bcc50770
3 changed files with 77 additions and 5 deletions
--- a/apps/api/src/services/decision_manager.py
+++ b/apps/api/src/services/decision_manager.py
@@ -215,6 +215,8 @@ async def _push_decision_to_telegram(
        # 2026-04-09 Claude Sonnet 4.6: Tool Calling 模型/後端
        nemotron_tool_model = proposal_data.get("nemotron_tool_model", "")
        nemotron_tool_backend = proposal_data.get("nemotron_tool_backend", "")
+        # 2026-04-16 ogt + Claude Sonnet 4.6: Playbook 匹配名稱穿透到 TG 卡片 (ADR-076)
+        _playbook_name = proposal_data.get("playbook_name", "")

        # 建立 approval_id (使用 incident_id 作為追蹤)
        # 2026-03-27 ogt: 修復 INC-INC-INC- 重複前綴 bug
@@ -331,6 +333,7 @@ async def _push_decision_to_telegram(
                incident_id=incident.incident_id,
                alert_category=_alert_category,
                notification_type=_notification_type,
+                playbook_name=_playbook_name,
            )

        # 2026-04-09 Claude Sonnet 4.6: 存 message_id → 後續狀態更新在原訊息延續
@@ -511,8 +514,8 @@ async def _generate_playbook_draft_if_new(incident: "Incident") -> None:
            KnowledgeEntryCreate(
                title=f"[AI草稿] {alertname} Playbook",
                content=content,
-                entry_type=EntryType.PLAYBOOK,
-                category="auto_generated",
+                entry_type=EntryType.AUTO_RUNBOOK,
+                category="ai_system",
                tags=[alertname, severity, "ai_draft", "mcp_phase4c"],
                source=EntrySource.AI_EXTRACTED,
                status=EntryStatus.DRAFT,
@@ -539,6 +542,20 @@ async def _generate_playbook_draft_if_new(incident: "Incident") -> None:
            entry_id=entry.entry_id,
        )

+        # 2026-04-16 ogt + Claude Sonnet 4.6: KM 建立後推 Telegram 通知 (ADR-076)
+        # best-effort — 不阻塞主流程
+        try:
+            from src.services.telegram_gateway import get_telegram_gateway as _get_tg
+            _tg = _get_tg()
+            await _tg.send_as_nemotron(
+                f"📚 <b>KM 新條目</b> — <code>{entry.entry_id}</code>\n"
+                f"🏷️ 分類：<b>auto_generated</b>  狀態：DRAFT\n"
+                f"📋 告警：<code>{alertname}</code>  事件：<code>{incident.incident_id}</code>\n"
+                f"標題：{str(getattr(entry, 'title', alertname))[:60]}"
+            )
+        except Exception as _tg_err:
+            _sl.get_logger(__name__).debug("km_tg_notify_failed", error=str(_tg_err))
+
    except Exception as e:
        import structlog as _sl
        _sl.get_logger(__name__).debug("playbook_draft_failed", error=str(e))
@@ -2314,6 +2331,7 @@ class DecisionManager:
            import json as _json
            from src.utils.timezone import now_taipei

+            # 2026-04-16 ogt + Claude Sonnet 4.6: 補入 playbook_id / alert_category (ADR-076)
            entry = {
                "ts": now_taipei().isoformat(),
                "confidence": proposal_data.get("confidence", 0.0),
@@ -2321,6 +2339,9 @@ class DecisionManager:
                "provider": proposal_data.get("provider", proposal_data.get("source", "")),
                "source": proposal_data.get("source", ""),
                "diagnosis": str(proposal_data.get("diagnosis", proposal_data.get("description", "")))[:200],
+                "playbook_id": proposal_data.get("playbook_id", ""),
+                "playbook_name": proposal_data.get("playbook_name", ""),
+                "alert_category": proposal_data.get("alert_category", ""),
            }

            async with get_db_context() as db:
--- a/apps/api/src/services/runbook_generator.py
+++ b/apps/api/src/services/runbook_generator.py
@@ -90,7 +90,7 @@ class NemotronRunbookGenerator:
                title=f"[AUTO] {incident.incident_id} — {playbook.name}",
                content=content,
                entry_type=EntryType.AUTO_RUNBOOK,
-                category="auto_generated",
+                category="ai_system",
                tags=list(incident.affected_services or []) + ["auto_runbook", "nemotron"],
                source=EntrySource.AI_EXTRACTED,
                status=EntryStatus.DRAFT,
--- a/apps/api/src/services/telegram_gateway.py
+++ b/apps/api/src/services/telegram_gateway.py
@@ -164,6 +164,9 @@ class TelegramMessage:
    ai_provider: str = ""  # ollama/gemini/claude/expert_system/mock
    # 2026-04-04 ogt: 底層模型名稱 (e.g. qwen2.5:7b-instruct, nemotron-70b)
    ai_model: str = ""
+    # 2026-04-16 ogt + Claude Sonnet 4.6: 告警分類與修復鏈路顯示 (ADR-076)
+    alert_category: str = ""   # host/k8s/database/service/external_site/secops 等
+    playbook_name: str = ""    # 匹配到的 Playbook 名稱（空字串=規則匹配）

    # ==========================================================================
    # Phase 22: Nemotron 協作欄位 (ADR-044)
@@ -237,7 +240,7 @@ class TelegramMessage:

        # 2026-03-29 ogt: ADR-037 異常頻率顯示
        frequency_block = ""
-        if self.anomaly_frequency and self.anomaly_frequency.get("count_24h", 0) > 1:
+        if self.anomaly_frequency and self.anomaly_frequency.get("count_24h", 0) >= 1:
            freq = self.anomaly_frequency
            escalation_emoji = {
                None: "",
@@ -304,12 +307,37 @@ class TelegramMessage:
            "LOW": "低風險",
        }.get(self.risk_level.upper(), self.risk_level)

+        # ADR-076 分類標籤 (2026-04-16 ogt + Claude Sonnet 4.6)
+        _category_map = {
+            "host": "🖥️ 主機", "host_resource": "🖥️ 主機",
+            "k8s": "☸️ K8s", "kubernetes": "☸️ K8s",
+            "database": "🗄️ 資料庫",
+            "service": "⚙️ 服務",
+            "external_site": "🌐 外部網站",
+            "secops": "🔐 安全",
+            "auto_repair": "🔧 自動修復",
+            "alertchain_health": "📡 告警鏈路",
+            "flywheel_health": "🔄 飛輪健康",
+            "docker": "🐳 Docker",
+            "infrastructure": "🏗️ 基礎設施",
+        }
+        category_line = ""
+        if self.alert_category:
+            cat_display = html.escape(_category_map.get(self.alert_category, self.alert_category))
+            category_line = f"🏷️ 分類：<b>{cat_display}</b>\n"
+
+        # Playbook 顯示
+        playbook_line = ""
+        if self.playbook_name:
+            playbook_line = f"📖 Playbook：<code>{html.escape(self.playbook_name)}</code>\n"
+
        # ADR-075 TYPE-3 格式組裝
        message = (
            f"{self.status_emoji} ACTION REQUIRED | <b>{html.escape(risk_zh)}</b>\n"
            f"──────────────────────\n"
            f"📋 <code>{html.escape(incident_id)}</code>\n"
            f"🎯 資源：<code>{safe_resource}</code>\n"
+            f"{category_line}"
            f"\n"
            f"🧠 <b>AI 深度診斷</b>\n"
            f"├─ 分析：{safe_root_cause}\n"
@@ -317,6 +345,7 @@ class TelegramMessage:
            f"└─ {ai_source}\n"
            f"\n"
            f"⚡ <b>建議修復動作</b>\n"
+            f"{playbook_line}"
            f"<code>{safe_action}</code>\n"
        )
        if ai_cost_display:
@@ -448,10 +477,27 @@ class TelegramMessage:
            )

        # 2026-04-05 Claude Code: 重設計訊息格式，提升易讀性
+        # 2026-04-16 ogt + Claude Sonnet 4.6: 補 alert_category + playbook_name 顯示
+        _category_map = {
+            "host": "🖥️ 主機", "host_resource": "🖥️ 主機",
+            "k8s": "☸️ K8s", "kubernetes": "☸️ K8s",
+            "database": "🗄️ 資料庫", "service": "⚙️ 服務",
+            "external_site": "🌐 外部網站", "secops": "🔐 安全",
+            "auto_repair": "🔧 自動修復", "docker": "🐳 Docker",
+        }
+        category_line = ""
+        if self.alert_category:
+            cat_display = html.escape(_category_map.get(self.alert_category, self.alert_category))
+            category_line = f"🏷️ {cat_display}\n"
+        playbook_line = ""
+        if self.playbook_name:
+            playbook_line = f"📖 <code>{html.escape(self.playbook_name)}</code>\n"
+
        # 組裝訊息
        message = (
            f"{self.status_emoji} <b>{html.escape(self.risk_level)}</b>  <code>{html.escape(incident_id)}</code>\n"
            f"<b>{safe_resource}</b>\n"
+            f"{category_line}"
            f"\n"
            f"{conf_line}\n"
            f"👥 {resp_display}\n"
@@ -460,7 +506,7 @@ class TelegramMessage:
        if nemotron_block:
            message += f"\n{nemotron_block}"
        message += (
-            f"\n🔧 <b>建議:</b> {safe_action}\n"
+            f"\n🔧 <b>建議:</b> {playbook_line}{safe_action}\n"
            f"⏱️ 停機: {safe_downtime}"
        )

@@ -1488,6 +1534,8 @@ class TelegramGateway:
        # 2026-04-12 ogt: ADR-075 告警分類與通知類型（斷點 B 修復）
        alert_category: str = "",
        notification_type: str = "",
+        # 2026-04-16 ogt + Claude Sonnet 4.6: 修復鏈路顯示 (ADR-076)
+        playbook_name: str = "",
    ) -> dict:
        """
        推送待簽核卡片到 Telegram (v7.0 含 SignOz 整合)
@@ -1561,6 +1609,9 @@ class TelegramGateway:
            # 2026-04-09 Claude Sonnet 4.6: Tool Calling 模型/後端
            nemotron_tool_model=nemotron_tool_model,
            nemotron_tool_backend=nemotron_tool_backend,
+            # 2026-04-16 ogt + Claude Sonnet 4.6: 修復鏈路顯示 (ADR-076)
+            alert_category=alert_category,
+            playbook_name=playbook_name,
        )

        # 格式化訊息 — Phase 22: 如果 Nemotron 啟用，使用雙軌格式