I like having different models review each other—for instance, having Claude write code for Codex to review, or having Codex write it and Claude review it.
Ask HN: Does a human still review your code?
31–32 of 32 posts
Re: Ask HN: Does a human still review your code?
#32Personal/side-business: Almost fully automated PR review flow. Local agents review before PR, Github Actions runs Codex/Claude to review the PR, if everything passes (Unit, e2e, lint, etc) then I merge. I'll review critical code paths if they are touched but that's not too often. Day Job: Similar but more human-in-the-loop and we are still feeling out "what needs human review" and "agent review is good enough". LLM r…
This matches my experience. For a lot of agent-written code, reviewing the final diff feels too low-level and too late. I'd rather review the plan&assumptions. If those are right, AI will probably get the implementation right as well.