Interesting finger pointing game getting played in corporate. They create issue using slop AI, then fix is implemented with slop AI, then everyone sees the issue wasn't actually fixed, but all the tests are passing so?