Hey Product Hunt 👋 I built rook because I kept losing the thread. Once you run more than one AI coding agent, it falls apart fast: one is stuck on a permission prompt, another finished ten minutes ago, a third is quietly burning tokens on the wrong thing — and you're flipping through terminal tabs to figure out which is which. rook is one screen over all of them. Open it in your browser and you see every agent, what it's actually doing (in plain English), and which ones are waiting on you — with the controls right there: approve a prompt, reply, jump into the live terminal, review the diff, or read the PR it's working on. A few things I care about: - 100% local & single-user. It runs on localhost. Nothing leaves your machine, and GitHub stays read-only until you turn on write actions. - Works with what you already run — Claude Code, Codex, Aider, Gemini. Auto-discovers agents; no config. - One line to try: curl -fsSL overclockhq.github.io/rook/install.sh | sh - Open source, MIT. It's day one and I'll be here all day — tell me what's missing, what's broken, and what you'd want it to watch next. I'll ship fixes live. 🛠️

rook

Watch and drive all your AI coding agents, locally

Watch and drive all your AI coding agents, locally