Recently, we've started gettitng more and more patch submissions that
use undisclosed AI, in some cases nonsensical, in other cases not.
Should we add a policy like many other projects have, and if so, what
is our position? Personally I have noticed that people who use AI to
generate patches to send to mailing lists tend to not review them very
carefully themselves before firing them off.
I guess my W02jobs patch triggered this email?
I created it entirely on my phone using Claude Code. I verified that tests were passing, but couldn't see how long it was taking.
Sorry for not manually verifying. Sorry also that it wasn't clearly attributed as co-authored by Claude. I manually verified these things in my first patch, but I didn't verify it for this one.
My current understanding is that the assertion succeeded quickly, but the test was still waiting for the sleep. I'll reply to that thread about that.
Re: the main point of your thread: I think we should improve the test coverage, so it's easier for both humans and AI to send patches. That's what I was trying to do. Please let me know if you have suggestions.