LLM fallback #76
Description
Hi, I noticed that the current LLM fallback when a step fails is quite basic and could be significantly improved. I’d like to start working on this.
Could you please confirm if this is something I can take up?
Also, I couldn’t find a contributor guide is there any process I should follow before starting