Alright, the moment of truth. Open your connected platform (Telegram, Discord, whatever you set up) and send your assistant a simple message like "Hello, what can you do?"
Now watch what happens in the Control UI at http://127.0.0.1:18789/. You'll see the incoming message, the model's reasoning, and the outgoing reply in real time. If the assistant uses any tools, each execution shows up as a separate event with its input and output.
No reply within 30 seconds? Check the daemon logs with openclaw gateway logs. Common issues: expired API key, wrong model ref, or a platform config typo. Fix the issue and try again.