feat(api): LLM 整合 Expert System 診斷上下文

長期方案實作: Expert 診斷 + LLM 智能分析

變更:
1. decision_manager._dual_engine_analyze():
   - 測試資源跳過 LLM (省錢)
   - 傳遞 Expert 診斷上下文給 LLM
   - LLM 失敗時根據診斷調整回應

2. openclaw.generate_incident_proposal():
   - 新增 expert_context 參數
   - Prompt 包含 Expert 診斷結果
   - 引導 LLM 基於診斷做決策

流程:
Playbook → Expert診斷 → LLM(with context) → 智能建議

這是「先診斷根因，再決定行動」的正確實作

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

apps/api/src/services/openclaw.py

@@ -1068,17 +1068,25 @@ Trace URL: {signoz_trace_url}
         severity: str,
         signals: list[dict],
         affected_services: list[str],
+        expert_context: dict | None = None,
     ) -> tuple[dict | None, str, bool]:
         """
         為 Incident 生成 LLM-based 修復提案
 
         Phase 6.4: 賦予大腦「生成解決方案」的思考能力
+        2026-03-27: 整合 Expert System 診斷上下文
 
         Args:
             incident_id: Incident ID
             severity: 嚴重度 (P0/P1/P2/P3)
             signals: 關聯的告警訊號
             affected_services: 受影響服務
+            expert_context: Expert System 初步診斷 (可選)
+                - initial_diagnosis: 規則匹配結果
+                - diagnosis_description: 診斷描述
+                - suggested_diagnosis_commands: 建議診斷指令
+                - expert_confidence: 信心分數
+                - requires_human_review: 是否需人工介入
 
         Returns:
             (proposal_dict, provider, success)
@@ -1108,11 +1116,31 @@ Trace URL: {signoz_trace_url}
             signoz_context = f"""
 ## 📊 SignOz Real-time Metrics
 {signoz_metrics.to_summary()}
+"""
+
+        # 2026-03-27: 整合 Expert System 診斷上下文
+        expert_diagnosis_context = ""
+        if expert_context:
+            diagnosis_cmds = expert_context.get("suggested_diagnosis_commands", [])
+            diagnosis_cmds_str = "\n".join([f"  - `{cmd}`" for cmd in diagnosis_cmds]) if diagnosis_cmds else "  - (無)"
+            expert_diagnosis_context = f"""
+## 🔍 Expert System Initial Diagnosis
+- **Matched Rule**: {expert_context.get('initial_diagnosis', 'unknown')}
+- **Diagnosis**: {expert_context.get('diagnosis_description', 'N/A')}
+- **Confidence**: {expert_context.get('expert_confidence', 0.5):.0%}
+- **Requires Human Review**: {'Yes' if expert_context.get('requires_human_review') else 'No'}
+- **Suggested Diagnosis Commands**:
+{diagnosis_cmds_str}
+
+**IMPORTANT**: The Expert System has provided an initial diagnosis.
+Consider this context but apply your own analysis. If Expert says "human review required",
+provide diagnostic guidance rather than automated fixes.
 """
 
         proposal_prompt = f"""{OPENCLAW_SYSTEM_PROMPT}
 
 {signoz_context}
+{expert_diagnosis_context}
 
 ## 🚨 Incident Context
 - **Incident ID**: {incident_id}
@@ -1124,13 +1152,14 @@ Trace URL: {signoz_trace_url}
 {signal_summary}
 
 ## 🎯 Your Task
-Based on the above incident and signals, generate a remediation proposal.
+Based on the above incident, signals, and Expert System diagnosis, generate a remediation proposal.
 You MUST respond with ONLY valid JSON following the schema above.
 Focus on:
-1. Root cause analysis based on signals and SignOz data
-2. Specific kubectl command to remediate
+1. Root cause analysis based on signals, SignOz data, and Expert diagnosis
+2. Specific kubectl command to remediate (or diagnostic command if root cause unclear)
 3. Risk assessment for the proposed action
 4. Preventive recommendations
+5. If Expert System flagged "human review required", prioritize diagnostic commands over fixes
 """
 
         logger.info(