---
title: "About · yokoten"
description: "yokoten is a free, open-source command-line tool that guards AI coding agents and builds a good software team’s habits into your project. It is made in the open by yokoten-labs, with Claude."
canonical: https://yokoten.app/about
last-updated: 2026-10-22
lang: en
---

> For AI agents: the documentation index is at https://yokoten.app/llms.txt. To set up yokoten in a project, follow https://yokoten.app/agents.md.

# About · yokoten

> yokoten is a free, open-source command-line tool that guards AI coding agents and builds a good software team’s habits into your project. It is made in the open by yokoten-labs, with Claude.

## What yokoten is

yokoten is a free command-line tool. It puts a good software team’s process into your project: rules your AI agents read, a hook that stops a risky command before it runs, a commit check that catches personal data and secrets, and CI that runs the same check last. On the server, branch protection and a bot with limited rights make sure nothing merges or ships without you.

`doctor` reads your project and says where it is on a road of five stages: safety net, issues and pull requests, release, a team of agents, going public. Then it names the one next step.

## Who it’s for

Anyone who builds software with AI coding agents, and above all people new to how good teams work: issues, tests, reviews and releases. You learn the habits by using them in your own project, one stage at a time. The guard runs in Claude Code, Codex CLI, GitHub Copilot CLI, Antigravity CLI, Muse Code, DeepSeek Harness and Grok CLI. With any other agent, or none, the commit check and CI still stop a bad change.

## Open source

yokoten is open source, under the [MIT license](https://github.com/yokoten-labs/yokoten/blob/main/LICENSE). It costs nothing and needs only Node 22.18 or later. The code, the issues, the pull requests, the reviews and the releases are all public, in the [yokoten-labs](https://github.com/yokoten-labs) organization on GitHub. The project works the way it teaches: every change starts as an issue, goes through a pull request and a review, and passes CI before it merges.

## Built with Claude

yokoten-labs builds yokoten together with Claude, Anthropic’s AI model, working in Claude Code. The agents that write the code run under the same guard and the same rules yokoten gives you. A person approves every merge and decides when to release. The [decision records](https://github.com/yokoten-labs/yokoten/tree/main/docs/records) say why it is built this way.

## The name

Yokoten (横展) comes from Toyota’s factories: when one production line finds a fix for a defect, the fix goes to every line that could have the same defect. yokoten does that for your projects: a lesson one project learns is offered to the rest.

---

- [yokoten: guardrails and good team habits for AI coding agents](https://yokoten.app/index.md)
- [How it works · yokoten](https://yokoten.app/how-it-works.md)
- [Roadmap · yokoten](https://yokoten.app/roadmap.md)
- [Quick start · yokoten](https://yokoten.app/start.md)
- [Docs · yokoten](https://yokoten.app/docs.md)
- [Contact · yokoten](https://yokoten.app/contact.md)
- [Privacy · yokoten](https://yokoten.app/privacy.md)
- [agents.md](https://yokoten.app/agents.md)
- [llms.txt](https://yokoten.app/llms.txt)

Language: [简体中文](https://yokoten.app/zh/about.md) · [繁體中文](https://yokoten.app/zh-tw/about.md) · [日本語](https://yokoten.app/ja/about.md)

HTML: https://yokoten.app/about

© 2026 yokoten-labs