Update system to use remote Ollama host with graceful fallback handling

README.md

@@ -24,6 +24,6 @@ A personal development assistant powered by LLMs.
 
 ## Deployment
 
-The application is designed to work in Hugging Face Spaces environment. For local LLM inference, it connects to a remote Ollama instance via ngrok tunnel at . This allows the application to access powerful local models without requiring them to be installed directly in the Space.
+The application is designed to work in Hugging Face Spaces environment. For local LLM inference, it connects to a remote Ollama instance via ngrok tunnel at `https://a877ef1aa487.ngrok-free.app`. This allows the application to access powerful local models without requiring them to be installed directly in the Space.
 
 In case the remote Ollama instance is unavailable, the system gracefully falls back to checking a local instance, and handles unavailability by showing appropriate status messages in the UI.