feat(#875): LLM Router shadow mode + consolidate litellm master key

backend/python/app/api/chat_service.py

@@ -193,6 +193,23 @@ from ..middleware.log_service import LogService
 
 
 from ..capability_registry import resolve_auto_model
+try:
+    from ..core.llm_router import (
+        resolve_api_key as _llm_resolve_api_key,
+        route as _llm_route,
+        set_route_ctx as _set_route_ctx,
+        SHADOW as _LLM_ROUTER_SHADOW,
+    )
+except Exception as _llm_router_err:
+    import os as _os, logging as _logging
+    _logging.getLogger(__name__).warning(f"[chat] llm_router import failed, router disabled: {_llm_router_err}")
+    def _llm_resolve_api_key(api_url, model_api_key, provider_api_key):
+        if api_url and "qihu-aiflow-litellm" in api_url:
+            return _os.environ.get("LITELLM_MASTER_KEY") or "sk-litellm-master-key-aiflow-2026"
+        return model_api_key or provider_api_key
+    _llm_route = None
+    _set_route_ctx = None
+    _LLM_ROUTER_SHADOW = True
 
 
 
@@ -2820,6 +2837,16 @@ async def _send_message_core(
 
         _try_models.extend([m for m in _fallbacks if m not in _try_models])
 
+    # ===== LLM Router（#875 影子模式）=====
+    try:
+        if _llm_route is not None:
+            _set_route_ctx(tenant_id=getattr(current_user, "tenant_id", "") or "", claw_type="web")
+            _rm = _llm_route("text", db=db, legacy_primary=_try_models[0].name if _try_models else None)
+            if not _LLM_ROUTER_SHADOW and _rm.models:
+                _try_models = _rm.models
+    except Exception as _e:
+        logger.warning(f"[chat] LLM Router 观测失败(降级忽略): {_e}")
+
     _retry_log = []
 
     _all_failed = False
@@ -3082,8 +3109,8 @@ async def _send_message_core(
 
                 # 2026-07-29 fix: 调用 LiteLLM 端点时，必须用 master key (sk-litellm-...) 而非模型自身 key (tp-c...)
                 # LiteLLM 内置每个 model 的真实 key 在 litellm_config.yaml，chat 不应直接传模型的 key
+                api_key = _llm_resolve_api_key(api_url, api_key, provider.api_key if provider else None)
                 if api_url and "qihu-aiflow-litellm" in api_url:
-                    api_key = os.environ.get("LITELLM_MASTER_KEY") or "sk-litellm-master-key-aiflow-2026"
                     auth_header = None  # 用默认 Authorization: Bearer
                     logger.info("[chat] LiteLLM endpoint detected, using master key instead of model key")
 

backend/python/app/api/im_llm.py

@@ -31,6 +31,13 @@ from sqlalchemy.orm import Session
 from ..capability_registry import resolve_auto_model
 from ..models import AIModel, ModelProvider, TokenUsageLog, gen_uuid
 from app.core.observability import get_trace_id
+# #875 LLM Router：统一模型路由层（影子模式先行，收敛 litellm 主密钥硬编码）
+from ..core.llm_router import (
+    route as _llm_route,
+    set_route_ctx as _set_route_ctx,
+    SHADOW as _LLM_ROUTER_SHADOW,
+    resolve_api_key as _llm_resolve_api_key,
+)
 
 logger = logging.getLogger(__name__)
 
@@ -212,6 +219,19 @@ def _call_real_llm(db: Session, user=None, prompt: str = "", platform: str = "im
         ).order_by(AIModel.is_default.desc()).all()
         _fallbacks = [m for m in _fallbacks if _is_text_only(m)]
         _try_models.extend([m for m in _fallbacks[:4] if m not in _try_models])
+
+        # ===== LLM Router（#875 影子模式）=====
+        # 计算决策并打 claw_type 标签 metrics；影子模式下仅观测，不改流量。
+        # 切主路径：LLM_ROUTER_SHADOW=0 时改用 Router 候选链（已含文本过滤，等同此处逻辑）。
+        try:
+            _set_route_ctx(tenant_id=getattr(user, "tenant_id", "") or "", claw_type="im_wechat")
+            _rm = _llm_route("text", db=db, legacy_primary=_try_models[0].name if _try_models else None)
+            if not _LLM_ROUTER_SHADOW and _rm.models:
+                _try_models = _rm.models
+                _debug_log(f"ROUTER_LIVE models={[m.name for m in _try_models]}")
+        except Exception as _e:
+            logger.warning(f"[im_llm] LLM Router 观测失败(降级忽略): {_e}")
+
         _debug_log(f"TRY_MODELS user={user_id} count={len(_try_models)} names={[m.name for m in _try_models]}")
 
         if not _try_models:
@@ -268,10 +288,8 @@ def _call_real_llm(db: Session, user=None, prompt: str = "", platform: str = "im
             provider = db.query(ModelProvider).filter(ModelProvider.id == ai_model.provider_id).first()
             api_key = ai_model.api_key or (provider.api_key if provider else None)
             api_url = ai_model.endpoint
-            # 2026-07-30: LiteLLM 端点统一用 master key（与 chat_service.py 对齐）
-            if api_url and "qihu-aiflow-litellm" in api_url:
-                _litellm_key = os.environ.get("LITELLM_MASTER_KEY") or "sk-litellm-master-key-aiflow-2026"
-                api_key = _litellm_key
+            # 2026-07-30: LiteLLM 端点统一用 master key（收敛进 LLM Router，两端唯一来源）
+            api_key = _llm_resolve_api_key(api_url, api_key, provider.api_key if provider else None)
             if not api_url and provider and provider.api_base_url:
                 api_url = f"{provider.api_base_url.rstrip('/')}/v1/chat/completions"
             if not (api_key and api_url):