move the LLM instance directly to Assistant to make it cleaner to share with tests

[bcherry]

this gets rid of the awkward AGENT_MODEL constant by just making the LLM an inherent property of the Assistant, which seems more intuitive

I also switched the test judge model to base 5.2 instead of the chat version to make it clearer that you can (and should) use a different model for evals than for core chat.

I'd like to move STT and TTS as well, but I found a bug we need to fix first

also see livekit-examples/agent-starter-node#43

[u9g]

I also switched the test judge model to base 5.2 instead of the chat version to make it clearer that you can (and should) use a different model for evals than for core chat.

You did this by changing to 5.2 rather than 4.1-mini, I'm confused what you mean here, since the judge model was already discriminated from the agent model, and if anything you don't want a heavy judge model that will slow down every single test since IIRC we don't run tests in parallel by default.