mux
Run and manage multiple AI coding agents in parallel workspaces
All Skills in This Repository (15)
Pure Emerald Level Indicatorsdev-server-sandbox
Launch isolated Mux dev-server instances with temporary MUX_ROOT directories.
tests
Enforce standardized unit and integration test organization and execution with bun x jest.
dev-desktop-sandbox
Run isolated Mux desktop Electron instances with unique temporary MUX_ROOT directories.
deep-review
Coordinate specialized sub-agents to review code changes and consolidate findings.
pull-requests
Enforce pull request governance for titles, reviews, and merge practices.
tbench
Automate Terminal-Bench 2.0 benchmarking and failure analysis for Mux agents.
react-effects
Guide React developers on when to use useEffect for side effects.
electron
Automate Electron desktop applications via Chrome DevTools Protocol using agent-browser.
dev-server-sandbox
Runs multiple isolated xum dev-server instances with temporary roots and free ports.
dev-desktop-sandbox
Launches isolated Electron desktop dev instances with temporary XUM_ROOT directories and free ports.
deep-review
Coordinates parallel sub-agents to review code changes across correctness, tests, and consistency.
pull-requests
Manages GitHub pull request creation, CI validation, and review resolution workflows.
Frequently Asked Questions
FAQPage SchemaHow to install Xum?▼
Run `npx skills add coder/mux --all -g -y` in your terminal to install all skills in this suite globally.
What does Xum do?▼
Xum is a desktop app that lets you run several AI coding agents at once, each in its own isolated workspace, with a central view of git changes, costs, and agent status.
Can Xum work with remote servers over SSH?▼
Yes. Workspaces can run locally, in git worktrees, or on a remote server over SSH, so agents can execute code wherever your project lives.
Which AI models does Xum support?▼
It supports major models like Claude, GPT, and Grok, plus local models through Ollama and many more through OpenRouter.
Can Xum test web apps automatically?▼
Yes. Its dogfood and browser automation skills let an agent explore a web app, reproduce bugs with screenshots and video, and write a structured issue report.
Related Repositories in Software Engineering
View All in Software Engineering→openclaw
Run a personal AI assistant across your devices and chat apps
superpowers
Gives coding agents a disciplined workflow from idea to merged code
react
AI agent skills for building, testing, and porting React core