AI Code Review Tools Are a Real Market Now

A thread around an open-source AI code review CLI tool drew practitioners sharing what they are already using in production. Coderabbit came up as a paid option at $30 per developer per month with positive results. One commenter described running a custom skill that routes code through Codex, Cursor, and Gemini simultaneously for a meta-review. The question underneath the discussion was not 'should we use AI for code review' but 'which tool wins and why.'

The pattern here is that AI code review has moved from experiment to infrastructure for many teams in under two years. The competitive question is whether purpose-built tools like Coderabbit justify their cost versus just using a general-purpose model in a side tab, which is a fair critique raised in the thread. The answer seems to be workflow integration, not raw capability.

A separate thread on Anthropic's open-source vulnerability discovery framework prompted someone to ask whether this is an existential threat to dedicated SAST vendors like Coverity. Nobody had a clean answer, but the fact that the question is being asked seriously is the signal.