mi is an autonomous coding agent that fits in a single JavaScript file. no framework, no dependencies beyond Node builtins, works with any OpenAI-compatible API — OpenAI, Ollama, local models, whatever you have. the core is a loop: call the llm, check if it wants to use tools, execute them, feed results back, repeat. two built-in tools — bash (full system access) and skills (markdown playbooks loaded on demand) — are enough for it to read repos, write code, run tests, and debug failures.
You open your terminal, type mi, and you have a coding agent. It reads your codebase, runs commands, fixes bugs, writes tests, refactors files — whatever you need. One command to install, zero config to start.
Bring any model. Qwen, Gemma, GPT, Claude, if you can run it locally or hit it through an API, mi works with it. Switch models with an env var. No vendor lock-in, no account required for local setups.
Here's the thing that makes mi different: the entire agent is 30 lines of JavaScript. Not a wrapper around a framework. Not a scaffold with plugins. Thirty actual lines you can read in a minute. You'll understand exactly what your agent is doing and why - no black box, no magic.
And because it's that simple, it's yours to extend. Drop a .mjs file into the tools folder and mi picks it up on the next call. Write a skill as a markdown file and it's available immediately. The agent can even write its own tools while it's running.
Need to throw more firepower at a problem? mi spawns subagents that work in the same repo and share context through progress files. Set a goal with a pass/fail check and let it iterate until the job is done.
mi is open source, MIT licensed, and runs anywhere Node does.
No comment highlights available yet. Please check back later!
About Mi on Product Hunt
“30-line zero-config CLI agent for bug fixes + refactoring”
Mi launched on Product Hunt on May 13th, 2026 and earned 0 upvotes and 1 comments, placing #43 on the daily leaderboard. mi is an autonomous coding agent that fits in a single JavaScript file. no framework, no dependencies beyond Node builtins, works with any OpenAI-compatible API — OpenAI, Ollama, local models, whatever you have. the core is a loop: call the llm, check if it wants to use tools, execute them, feed results back, repeat. two built-in tools — bash (full system access) and skills (markdown playbooks loaded on demand) — are enough for it to read repos, write code, run tests, and debug failures.
Mi was featured in Open Source (68.4k followers), Artificial Intelligence (468.1k followers), GitHub (41.2k followers) and Vibe coding (443 followers) on Product Hunt. Together, these topics include over 125.4k products, making this a competitive space to launch in.
Who hunted Mi?
Mi was hunted by Ivan Charapanau. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.
Want to see how Mi stacked up against nearby launches in real time? Check out the live launch dashboard for upvote speed charts, proximity comparisons, and more analytics.
Ivan Charapanau