Hi Product Hunt 👋 I built Ming because AI coding agents have made writing code dramatically faster, but reviewing that code still feels almost the same as before. After using tools like Codex and Claude Code, I kept running into the same problem: an agent could change 20–30 files in a few minutes, but I still had to understand the result file by file. AI-generated summaries helped a little, but they mostly described what the agent intended to do — not necessarily what actually changed. Ming is my attempt to improve the review side of the workflow. It reads your local Git diff directly in the browser and can organize related changes into focused review topics, while keeping the actual diff as the source of truth. I also wanted it to stay simple and developer-friendly: - open source and MIT licensed - local-first - no Ming account - no application backend - read-only repository access - works with local, uncommitted changes The project is still early, and I’d especially love feedback on the review workflow: When AI agents make large changes to your codebase, what part of reviewing those changes is the most painful for you today? Thanks for checking out Ming!

Ming — Review AI-Written Code

Open-source, local-first review for AI-written code

Open-source, local-first review for AI-written code