I have a migrane and my search-fu seems lowered today, but I want to find out if someone tested how good o3 is at fixing real(-istic) code bugs and I don't care about GCC optimization flags, ozone killing insects, ARC scores or any leetcode style evals...

Could someone point me to a good write-up where people tried out o3 to do a real programming tasks a la debugging (pinpointing and fixing code that's causing an issue reported in Jira/GH issue and how it makes sure that the fix works), please? 🙏