I'd like a feature that allows an LLM to evaluate the results when multiple LLMs execute the same task in parallel!

this needs a "synthesize" button to aggregate across all models also, the ability to @ mention each of those chats to have the other model provide a second opinion would he useful

Rather than having many different models generate their own code, it would be much better to choose a base model, examine the code generated by others' base models, approve their optimization recommendations, and move on. In this way, it is more confident about the code and it is more possible to reach the optimal value with the consortium decision.

why does this require a git repo

I need this but using the same model across multiple chats Code review often comes up with different observations and potential improvements given the same PR

I'm too dumb to understand why anyone would want this?

This can easily be expanded to multiple models used for the same task

Is the "paperless" desk really tidy? We thought going paperless would clear our desks, but now they're just cluttered with devices needing a charge. Wireless mice, speakers, tablets, phones all fighting for an outlet. To end this "epic outlet war," we designed a desk lamp.