debug_forever

How does it decide what to test in a PR? And how does this code review tool connect with services that require configs? E.g. My checkout uses stripe, and for QA, we have a sandbox account connected.

The complexity in our team's code bases have only gotten worse with AI-integrated agents. Maybe it's the prompts we're using, but it's an ironic twist that these tools that promise so much productivity today ends up dumping more tech debt into our code.

It's funny reading the "key contributors to dependency-complexity" -- Duplication, Exceptions, Inheritance, Temporal Decomposition -- because those qualities seem like the standard for AI-generated code.