Documentation

teamrc Guide

teamrc lets you define one agent team and apply it across Claude Code, Cursor, Codex, Gemini, and OpenClaw. Use this page to choose the shortest path to what you need.

Start here

Get Started in 2 Minutes

The shortest path from an empty repo to a working, synced team.

Core Concepts

Read this if you want the mental model behind members, skills, and knowledge.

CLI Reference

Use this when you want exact commands, flags, and common workflows.

Web UI Tour

Take the visual tour if you plan to manage teams from the browser.

The mental model

There are three moving parts:

1.

Team definition

The source of truth. One definition of your agents, rules, and instructions, stored on the relay and optionally mirrored locally as .teamrc.yaml .

2.

Platform config files

The files each platform actually reads. teamrc generates them from the team definition, so you do not need to hand-edit each platform separately.

3.

The relay

The shared coordination point. Changes go up through the CLI or web UI, and machines pull them back down with sync , pull , or background sync.

Why teamrc exists

AI agent platforms all expect different file layouts, config formats, and skill models. teamrc gives you one shared team definition, then generates the platform-native files and keeps them synced across machines without forcing those files into Git.

Explore the guide

Platforms

How teamrc writes config for Claude Code, Cursor, Codex, Gemini, and OpenClaw.

Syncing

The relay, invites, multi-project teams, and auth.

Configuration

The .teamrc.yaml schema, scope rules, machine config, and validation limits.

FAQ

Short answers about Git, accounts, privacy, syncing, and recovery.