cmux

Orchestrate multi-AI terminals and browser automation via a single cmux interface.

17|3|Updated Jan 23, 2026
One-click install
npx skills add https://github.com/insightflo/claude-impl-tools --skill cmux-insightflo
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: cmux
Source: https://github.com/insightflo/claude-impl-tools/tree/main/plugin/skills/cmux
Command: npx skills add https://github.com/insightflo/claude-impl-tools --skill cmux-insightflo

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

cmux provides unified control over multiple AI coding agents and their browser tasks from a single terminal interface, simplifying orchestration and automation.

Core Features & Use Cases

  • Centralized workspace management: create and organize workspaces, panes, surfaces, and panels for parallel tasks.
  • Browser automation and command routing: automate browser sessions and terminal commands across agents.
  • Notifications and sidebar metadata: real-time status updates and progress badges integrated into a lightweight UI.

Quick Start

Open Claude Code and command cmux to create a new workspace and split panels for parallel agents.

Frequently Asked Questions about cmux

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I manage multiple AI coding agents from a single terminal interface?▼

You can manage multiple AI agents from a single terminal by using a multiplexer interface to orchestrate workspaces, panes, and browser tasks across your macOS environment. This centralizes control for parallel processing.

Can I automate browser tasks alongside terminal commands for parallel AI agents?▼

Yes, you can automate browser tasks alongside terminal commands by routing them through a single interface. This allows you to coordinate browser sessions and command execution across multiple agents simultaneously.

What is the best way to organize workspaces and panels for parallel AI tasks?▼

The best way to organize parallel AI tasks is by using a multiplexer to create and manage distinct workspaces, panels, and surfaces. This enables centralized workspace management for concurrent agent operations.

Does this terminal multiplexer work on macOS environments?▼

Yes, this terminal multiplexer works across macOS environments, providing a JSON-RPC interface over a local socket. It requires YAML frontmatter setup to define the workspace configuration.

How do I get real-time status updates for ongoing browser automation tasks?▼

You get real-time status updates for browser automation tasks through integrated notifications and sidebar metadata. This provides progress badges and lightweight UI elements directly within your terminal interface.

What are the limitations of using a single interface for multi-AI terminal orchestration?▼

A limitation of single-interface multi-AI orchestration is that all terminal and browser automation routing depends on a single JSON-RPC socket connection. If this socket fails, centralized control over parallel agents is interrupted.