logoalt Hacker News

AI Code Review Gets Better When I Ask Models to Debate: Claude, Gemini, Codex

6 pointsby Fendytoday at 6:02 PM1 commentview on HN

Comments

itmiticatoday at 8:36 PM

I did the exact same thing! Uncanny.

I agree with models being better at different tasks: gemini-cli is superficial, codex is stubborn as a mule and dependable, claude-cli just wants to get something working and done. qwen-cli, Qwen, in general, has a tendency to pendulate too much.

I also reduced the team to two, codex and claude, for me.

pro_underscoretoday at 6:29 PM

[dead]