diff --git a/apps/api/src/services/drift_interpreter.py b/apps/api/src/services/drift_interpreter.py
index 31baf523..ee88d30b 100644
--- a/apps/api/src/services/drift_interpreter.py
+++ b/apps/api/src/services/drift_interpreter.py
@@ -101,45 +101,24 @@ class NemotronDriftInterpreter:
 
     async def _call_nemotron(self, prompt: str) -> DriftInterpretation:
         """
-        呼叫本地 Ollama qwen2.5:7b-instruct 進行意圖分析
-        BUG-001 修復 2026-04-11: nvidia_provider 已重構為返回 NvidiaProviderResult 物件（非 4-tuple），
-        改用 Ollama httpx 直接呼叫，繞過 nvidia_provider，與 drift_narrator_service 一致
+        呼叫 OpenClaw AI Router 進行意圖分析。
+
+        2026-04-17 ogt + Claude Sonnet 4.6: 改用 OpenClaw 取代直接 Ollama httpx 呼叫
+        根因：直接呼叫 Ollama 繞過 AI Router，無 fallback → "All connection attempts failed"
+        修復：統一走 openclaw.call()，自動享有 Provider 降級與 fallback 機制
+        廢棄：BUG-001 的 httpx 直連繞過法（nvidia_provider 介面已於 7eb8375 穩定）
         """
-        import httpx
-        from src.core.config import get_settings
-        from src.services.model_registry import get_model as _get_model
-
-        # C1 修復 2026-04-11: 禁止寫死內網 IP（feedback_frontend_internal_ip_ban 鐵律）
-        # 改從 settings.OLLAMA_URL 讀取（已有此設定，default=http://192.168.0.111:11434）
-        _settings = get_settings()
-        OLLAMA_URL = getattr(_settings, "OLLAMA_URL", "http://192.168.0.111:11434")
-        # D1 集中化 2026-04-11: 從 models.json providers.ollama.models.drift_intent 讀取
-        MODEL = _get_model("ollama", "drift_intent")
-        TIMEOUT = 45.0
-
         try:
-            async with httpx.AsyncClient(timeout=TIMEOUT) as client:
-                resp = await client.post(
-                    f"{OLLAMA_URL}/api/generate",
-                    json={
-                        "model": MODEL,
-                        "prompt": prompt,
-                        "stream": False,
-                        "options": {"temperature": 0.2, "num_predict": 200},
-                    },
-                )
-                resp.raise_for_status()
-                data = resp.json()
-                response_text = data.get("response", "").strip()
+            from src.services.openclaw import get_openclaw
+            openclaw = get_openclaw()
+            response_text, _provider, success = await openclaw.call(prompt)
 
-            if not response_text:
-                return self._unknown_result("Ollama 回傳空值")
+            if not success or not response_text:
+                logger.warning("drift_interpreter_openclaw_failed", provider=_provider)
+                return self._unknown_result("AI Router 回傳失敗或空值")
 
             return self._parse_response(response_text)
 
-        except httpx.TimeoutException:
-            logger.warning("drift_interpreter_timeout", model=MODEL)
-            return self._unknown_result("Ollama 超時")
         except Exception as e:
             logger.warning("drift_interpreter_error", error=str(e))
             return self._unknown_result(str(e))