Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 

Repository files navigation

tmux-agent-teams

A high-level AI agent skill that abstracts away repetitive instructions for managing Claude Code agent teams in tmux. Instead of requiring users to describe every operation in detail, this skill converts a channel request into the necessary steps to start or manage an agent team, streamlining the process of creating and coordinating collaborative AI agents.

For a detailed setup, see the guide at https://tmuxcheatsheet.com/openclaw-tmux-setup/.

example usages

Creating an Agent Team

Prompt:

Create an agent team to review the repository in ~/example_repo. Use a security reviewer, a performance reviewer, and a QA reviewer. Send me the combined findings when they finish.

Response:

I can do that — I'll create a Claude Code agent team in a tmux session for ~/example_repo with security, performance, and QA reviewers. Quick confirmations before I start:
• Team name to use (I'll default to "review" if you don't care):
• Proceed now to create the tmux session and launch Claude Code in ~/example_repo? (Say "yes" to proceed or tell me changes.)

Listing Agent Teams

Prompt:

List agent teams.

Response:

I found 1 Claude Code agent team running:
Session name: claude-3a9ee5-review
working directory: ~/example_repo
status (inferred): actively working — "Doodling… (3m 54s · thinking)" and waiting at the prompt (❯).
Would you like me to capture more output, attach to the session, or do anything else with this team?

About

An AI agent skill that enables agents to interact with Claude Code through tmux.

Resources

Stars

3 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors