Enhance workflow routing and agent configuration management

- Introduce WorkflowLLMRouter for pre-response LLM routing, allowing agents to determine the appropriate function to call based on user input.
- Implement UserTurnRoutingProcessor to manage user turns before reaching the LLM, ensuring proper routing and handling of user messages.
- Refactor WorkflowBrain to integrate new routing logic and enhance agent stage configuration, including entry modes and resource management.
- Update service factory to support dynamic LLM resource configuration based on workflow settings.
- Add tests for new routing functionality and ensure proper handling of user messages in various scenarios.

backend/routes/assistants.py

@@ -48,14 +48,23 @@ async def _validate_workflow_references(
     settings = graph.get("settings") or {}
     resource_expectations: dict[str, str] = {}
     for key, capability in (
+        ("defaultLlmResourceId", "LLM"),
         ("defaultAsrResourceId", "ASR"),
         ("defaultTtsResourceId", "TTS"),
     ):
         if settings.get(key):
             resource_expectations[str(settings[key])] = capability
-    knowledge_ids: set[str] = set()
+    knowledge_ids: set[str] = (
+        {str(settings["knowledgeBaseId"])}
+        if settings.get("knowledgeBaseId")
+        else set()
+    )
     for node in graph.get("nodes") or []:
         data = node.get("data") or {}
+        if node.get("type") == "agent" and data.get("inheritGlobalConfig", True):
+            continue
+        if data.get("llmResourceId"):
+            resource_expectations[str(data["llmResourceId"])] = "LLM"
         if data.get("asrResourceId"):
             resource_expectations[str(data["asrResourceId"])] = "ASR"
         if data.get("ttsResourceId"):