Auto Pause when rate limit exceeded

[lukerogers]

Today I was running /fleet with /autopilot to build a new app and after a while it started showing the following:

✗ Sorry, you've hit a rate limit that restricts the number of Copilot model requests you can make within a specific time period. Please try again in 46 minutes. Please review our Terms of Service
(https://docs.github.com/site-policy/github-terms/github-terms-of-service).

I have no doubt I exceeded the rate limit and I have no problem with that, but it kept throwing the error with the same time value. I don't know how long it was showing the error, but it wasn't increasing or decreasing, so I'm guessing it was resetting the time with each request.

It seems like an opportunity to have the CLI pause for the time specified to allow the rate limit condition to pass. Is this something that is doable?

[AndreaGriffiths11]

Thanks for the detailed report. This makes sense. What you’re describing (the “try again in 46 minutes” value not changing) can happen if the CLI keeps sending requests while you’re already rate limited. Depending on how the backend enforces limits, each new request can effectively restart the cooldown window, so the retry time looks stuck. I'll submit this for the team to consider/review.
In the meantime, the only workaround is to stop sending requests and wait out the window manually, or split the run into smaller batches..

Thanks for taking the time!