codejury

什麼是 codejury？

AI coding agents often produce code that looks good but still needs manual review. I’ve been exploring whether multiple independent LLM reviewers can do better than having the same model review its own code. I built CodeJury, a small harness with AI judges focused on correctness, security, architecture, testing, performance, and overall quality. Would you use this alongside your coding agent? I’d love honest feedback, criticism, and suggestions. ⭐s appreciated!