revert: remove direct WebUI connection, focus on fixing LiteLLM streaming

- Reverted direct orchestrator connection to WebUI
- Added stream: true parameter to qwen-2.5-7b model config
- Keep LiteLLM as single proxy for all models

ai/litellm-config.yaml

@@ -40,6 +40,7 @@ model_list:
       timeout: 600  # 10 minutes for generation
       stream_timeout: 600
       supports_system_messages: false  # vLLM handles system messages differently
+      stream: true  # Enable streaming by default
 
   # Image Generation
   - model_name: flux-schnell