hummer98

hummer98 / using-cmux

Public

AI-oriented cmux terminal skill for Claude Code

18
0
69% credibility
Found Mar 27, 2026 at 18 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Shell
AI Summary

A plugin package that equips AI coding assistants with skills to manage and monitor sub-agents in a terminal multiplexer environment for better visibility and control.

How It Works

1
🔍 Discover a better way

You hear about a helpful add-on that lets your main AI assistant watch and guide smaller helper AIs working together on tasks in real time.

2
✅ Get ready

Make sure you have your AI coding helper and the shared workspace tool installed on your computer.

3
Pick your setup style
🚀
Quick plugin

Use a simple command in your AI to grab and set it up instantly.

📦
Skills only

Add just the core guidance without extra shortcuts.

🛠️
Hands-on copy

Download and run a setup script to place files where needed.

4
🚀 Launch in workspace

Open your shared workspace and start your main AI inside it – the add-on loads automatically.

5
👀 Watch magic happen

Tell your AI to handle a big task, and it splits the screen to launch helpers, sends them work, and shows every step live so you can jump in anytime.

6
📱 Check quick help

Type a simple shortcut command anytime to see handy tips on what your AI can do.

🎉 Perfect teamwork

Now your AI team works smoothly under your watchful eye, making debugging easy and results reliable.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 18 to 18 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is using-cmux?

using-cmux is an AI-oriented shell skill for Claude Code that equips AI agents to control cmux, a terminal multiplexer, for launching and managing sub-agents. It solves the black-box problem of Claude Code's built-in Agent tool by exposing sub-agent inputs, outputs, and states in real-time panes, letting you intervene, debug, or tweak on the fly. Run it in a cmux session, and skills auto-load via the CMUX_SOCKET_PATH env var—no extra setup.

Why is it gaining traction?

Unlike hashangit/cmux-skill, which buries sub-agent patterns under browser automation, using-cmux centers on core terminal ops: pane splitting, command sending with smart newline rules, full sub-agent lifecycles (launch to result harvest), screen reads with refresh fixes, notifications, and progress bars. Developers dig the streamlined patterns for trustworthy AI workflows and quick plugin install via `claude /plugin install hummer98/using-cmux`. The `/cmux` slash command gives instant refs too.

Who should use this?

Claude Code users building AI agent teams in terminal-heavy setups, especially those debugging opaque sub-agents during code gen or orchestration tasks. Ideal for AI-oriented dev GitHub tinkerers running cmux who need visible, intervenable multi-agent flows without browser cruft. Skip if you're not in the Claude + cmux ecosystem.

Verdict

Grab it if you're deep in Claude Code and cmux—solid docs and patterns make it a practical skill booster, despite 18 stars signaling early maturity. The 0.699999988079071% credibility score flags it as experimental; test in a side session first.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.