fix(ai): 修復假信心度 + 解除 Shadow Mode (Phase 22 P1)

1. openclaw.py: LLM 截斷時 confidence 0.82→0.0 (禁止偽造信心度)
2. prompts.py: NEMOTRON schema 範例值改用佔位符，防模型照抄 0.75
3. configmap: SHADOW_MODE_ENABLED=false，開放 low 風險自動執行
   條件門檻: confidence≥90% + trust_score≥5 + playbook_success≥95%

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

apps/api/src/core/prompts.py

@@ -140,7 +140,7 @@ You are an SRE AI. Analyze the alert and respond with ONLY valid JSON.
 
 ## Required JSON Schema:
 {
-  "confidence": 0.75,
+  "confidence": <YOUR_CALCULATED_VALUE>,
   "reasoning": "簡短理由 (繁體中文)",
   "primary_responsibility": "FE|BE|INFRA|DB|COLLAB",
   "risk_level": "low|medium|critical",
@@ -155,7 +155,7 @@ You are an SRE AI. Analyze the alert and respond with ONLY valid JSON.
 
 ## Rules:
 1. Response MUST be valid JSON.
-2. confidence MUST be a real number you calculated (0.0-1.0). NOT 0.75 or 0.82.
+2. confidence is a float 0.0-1.0 you CALCULATE from evidence. High evidence = 0.85-0.95. Low evidence = 0.40-0.65. NEVER copy example values.
 3. Language: Traditional Chinese (Taiwan).
 4. No explanation outside JSON.
 """