agent loop TDD

You type one sentence. A local agent ships it — tests first, then a pull request. agent-loop is an autonomous coding runner driven by a DAG. It plans the work, runs self-correcting TDD cycles (red → green), commits, pushes an `agent/*` branch, and opens the GitHub PR. Not a chatbot that pastes code into chat. Example: `yarn task "add JWT login on the API"` What it does: • Splits one intent into feature nodes • Writes failing tests, then the code that makes them pass • Commits, pushes, opens or reuses a PR • Works with TypeScript, Python, Go, and Rust when the toolchain is there • Stays local with `--no-push` if you want Get started: 🐙 GitHub → https://github.com/matteym/agent-loop... 🎵 TikTok →   / matt0uzz   If this helped, like and subscribe for more open-source AI coding tools. #OpenSource #TDD #AIAgents #Cursor #DevTools #DAG #Python #TypeScript