Hooking into chat input/output for local automation — where to look?

[LeoMonteiroF]

Hi! First of all, really nice project — super clean and exactly what I was looking for for local LLM usage.

I’m trying to experiment with a small automation layer on top of the app (basically routing prompts from a local server and reading responses back, possibly via Tailscale).

To avoid digging blindly through the codebase, could you point me to:

• Where the chat input state is managed
• Where the send/submit function is triggered
• Where assistant responses are stored/updated in state

I’m not planning to change the core logic — just hook into these points.

Also, from your experience with the codebase, would adding something like a periodic fetch (or simple external input injection) be relatively straightforward, or are there architectural constraints I should be aware of?

Thanks a lot!