The Bots Always Find Something

Very similar situation here. Months of agentic coding and seem to find problems in almost every test. And every model seems to find different failures. It’s a major issue if you don’t know what’s worth fixing and what isn’t.