refactor(memory): consolidate memory config extraction and remove unused validator

- Add workspace default LLM fallback for emotion model in extraction orchestrator
- Consolidate memory config ID extraction logic into MemoryConfigService
- Remove duplicate extraction methods from AppService (_extract_memory_config_id_from_agent, _extract_memory_config_id_from_workflow)
- Remove unused validate_embedding_model function from validators
- Simplify AppService by delegating memory config extraction to MemoryConfigService
- Update validator exports to remove validate_embedding_model
- Improve code maintainability by centralizing memory configuration logic

api/app/services/memory_reflection_service.py

@@ -120,7 +120,14 @@ class WorkspaceAppService:
         return None
 
     def _get_memory_config(self, memory_content: str) -> Dict[str, Any]:
-        """Retrieve memory_config information based on memory_content"""
+        """Retrieve memory_config information based on memory_content
+        
+        Args:
+            memory_content: Memory config ID string
+            
+        Returns:
+            Dict containing memory config info including workspace_id for model fallback
+        """
         try:
             memory_content = resolve_config_id(memory_content, self.db)
             memory_config_result = MemoryConfigRepository.query_reflection_config_by_id(self.db, (memory_content))
@@ -128,6 +135,7 @@ class WorkspaceAppService:
             if memory_config_result:
                 return {
                     "config_id": memory_content,
+                    "workspace_id": memory_config_result.workspace_id,
                     "enable_self_reflexion": memory_config_result.enable_self_reflexion,
                     "iteration_period": memory_config_result.iteration_period,
                     "reflexion_range": memory_config_result.reflexion_range,
@@ -359,7 +367,17 @@ class MemoryReflectionService:
             }
 
     def _create_reflection_config_from_data(self, config_data: Dict[str, Any]) -> ReflectionConfig:
-        """Create reflective configuration objects from configuration data"""
+        """Create reflective configuration objects from configuration data
+        
+        If reflection_model_id is not set, falls back to workspace default LLM.
+        
+        Args:
+            config_data: Dict containing reflection config including workspace_id
+            
+        Returns:
+            ReflectionConfig object with model_id resolved
+        """
+        from app.repositories.workspace_repository import get_workspace_models_configs
 
         reflexion_range_value = config_data.get("reflexion_range")
         if reflexion_range_value is None or reflexion_range_value == "":
@@ -392,6 +410,17 @@ class MemoryReflectionService:
         if reflection_model_id:
             reflection_model_id = str(reflection_model_id)
         
+        # 如果 reflection_model_id 为空，回退到工作空间默认 LLM
+        if not reflection_model_id:
+            workspace_id = config_data.get("workspace_id")
+            if workspace_id:
+                workspace_models = get_workspace_models_configs(self.db, workspace_id)
+                if workspace_models and workspace_models.get("llm"):
+                    reflection_model_id = workspace_models["llm"]
+                    api_logger.info(
+                        f"reflection_model_id 为空，使用工作空间默认 LLM: {reflection_model_id}"
+                    )
+        
         return ReflectionConfig(
             enabled=config_data.get("enable_self_reflexion", False),
             iteration_period=str(iteration_period),  # ReflectionConfig期望字符串