feat: new chat working stateless. Added citation logic.

surfsense_backend/app/agents/new_chat/llm_config.py

@@ -87,10 +87,11 @@ def create_chat_litellm_from_config(llm_config: dict) -> ChatLiteLLM | None:
         provider_prefix = provider_map.get(provider, provider.lower())
         model_string = f"{provider_prefix}/{llm_config['model_name']}"
 
-    # Create ChatLiteLLM instance
+    # Create ChatLiteLLM instance with streaming enabled
     litellm_kwargs = {
         "model": model_string,
         "api_key": llm_config.get("api_key"),
+        "streaming": True,  # Enable streaming for real-time token streaming
     }
 
     # Add optional parameters

surfsense_backend/app/tasks/chat/stream_new_chat.py

@@ -12,10 +12,9 @@ from langchain_core.messages import HumanMessage
 from sqlalchemy.ext.asyncio import AsyncSession
 
 from app.agents.new_chat.chat_deepagent import (
-    create_chat_litellm_from_config,
     create_surfsense_deep_agent,
-    load_llm_config_from_yaml,
 )
+from app.agents.new_chat.llm_config import create_chat_litellm_from_config, load_llm_config_from_yaml
 from app.services.connector_service import ConnectorService
 from app.services.new_streaming_service import VercelStreamingService