Claude Squad 🤖 Agent Free

What it does

Claude Squad is a terminal app for running several AI coding agents at the same time without them tripping over each other. Each agent gets its own session and its own copy of your repository, so you can hand one a bug fix, another a refactor and a third a new feature, then review and push each result separately.

What is inside / how it works

Under the hood it combines two familiar tools. tmux provides an isolated terminal session for every agent, and git worktrees give each session its own branch and working directory, which avoids merge conflicts while agents work in parallel. A text-based interface lists your sessions and lets you switch between a live preview and a diff of each one's changes. Keyboard shortcuts create a session (optionally with a starting prompt), attach to it to give further instructions, commit and push its branch to GitHub, check out changes while pausing the session, resume it later or delete it. Tasks can keep running in the background, and an experimental auto-yes flag accepts agent prompts automatically for Claude Code and Aider. Configuration lives in ~/.claude-squad/config.json, where profiles let you pick which agent command a new session launches.

Works with

The default program is Claude Code. The README also documents Codex, Gemini CLI and Aider, launched with the -p flag or a profile, and other local terminal agents can be started the same way.

How to install or start

You need tmux and the GitHub CLI (gh). Install with Homebrew and add the short cs alias, or use the install script:

brew install claude-squad
ln -s "$(brew --prefix)/bin/claude-squad" "$(brew --prefix)/bin/cs"
cs

To start with another agent, run for example cs -p "codex".

Pricing

Free and open source under the AGPL-3.0 licence. Each agent session uses its own model quota or API credits, so running several in parallel multiplies usage.

Who should use it

Terminal-first developers who want to parallelise work across several agents while keeping each change reviewable on its own branch. It has no graphical interface and depends on tmux, so it is less suited to people who live in an IDE, and you still need to review and merge each branch yourself.