Koredex - Automatically fixes failing pytest test and proves it worked

I kept running into the same thing again and again — a pytest failure, and then 20–30 minutes just trying to figure out what broke, fixing it, rerunning tests, and repeating the cycle. It’s even worse when one failure leads to a bunch of others. So I started building Koredex. The idea was simple: instead of just showing failing tests, what if something could actually fix them and then prove the fix works? Right now it: finds failing tests tries to fix them reruns everything to make sure it actually worked rolls back if it makes things worse It’s still early and works best on things like import errors, dependency issues, and simpler test failures. More complex cases are still a work in progress. Would love to know what you think — especially if you deal with pytest a lot.