amux  by mixpeek

Control plane for running dozens of parallel AI coding agents

Created 4 months ago
282 stars

Top 92.3% on SourcePulse

GitHubView on GitHub
Project Summary

This project provides an open-source control plane for AI agents, specifically designed to multiplex and manage dozens of parallel Claude Code sessions using tmux. It targets developers and power users seeking to automate, orchestrate, and monitor AI coding agents efficiently, offering a significant productivity boost for tasks like parallel feature development, large-scale refactoring, and AI-driven testing.

How It Works

amux leverages tmux to create and manage independent agent sessions. A core self-healing watchdog monitors these sessions, automatically detecting and rectifying issues like context compaction failures, stuck prompts, and rate limits by restarting sessions or applying predefined actions. It facilitates agent-to-agent orchestration via a REST API and shared global memory, enabling complex workflows through features like a Kanban board for task management and real-time inter-session chat channels for coordination.

Quick Start & Requirements

  • Primary install / run command:
    git clone https://github.com/mixpeek/amux && cd amux && ./install.sh
    amux serve # → https://localhost:8822
    
  • Non-default prerequisites and dependencies: tmux, python3.
  • Links: amux.io, Getting started, FAQ, Blog

Highlighted Details

  • Self-healing: Automatically handles context compaction, restarts on corruption, and manages rate-limiting prompts.
  • Parallel Agents: Supports running dozens of independent agent sessions, each with a persistent UUID.
  • Orchestration: Features a REST API for inter-session communication, task delegation, and a Kanban board with atomic task claiming.
  • Web Dashboard & Mobile PWA: Provides a browser-based dashboard for live status, terminal peeking, file exploration, and a mobile-optimized PWA with offline support.
  • Single-file Architecture: The entire application is contained in a single Python file with inline HTML/CSS/JS, allowing for easy modification and automatic restarts on save.

Maintenance & Community

No specific details on maintainers, community channels (e.g., Discord/Slack), or roadmap were present in the provided README. The project is hosted on GitHub.

Licensing & Compatibility

  • License type: MIT + Commons Clause.
  • Compatibility notes: Free for use, modification, and self-hosting. Commercial resale requires a separate license. The project is local-first and designed for integration with Claude Code via tmux.

Limitations & Caveats

The project currently supports Claude Code via tmux; integration with other models or environments may require adaptation. It is local-first with no built-in authentication, necessitating careful network exposure management (e.g., using Tailscale or binding to localhost) to prevent unauthorized access to the dashboard. Mobile PWA functionality requires manual installation of a custom CA certificate for trust.

Health Check
Last Commit

1 day ago

Responsiveness

Inactive

Pull Requests (30d)
7
Issues (30d)
1
Star History
55 stars in the last 30 days

Explore Similar Projects

Starred by Andrew Ng Andrew Ng(Founder of DeepLearning.AI; Cofounder of Coursera; Professor at Stanford), Jack Lukic Jack Lukic(Author of Semantic UI), and
5 more.

ag2 by ag2ai

0.5%
5k
AgentOS for building AI agents and facilitating multi-agent cooperation
Created 1 year ago
Updated 1 day ago
Starred by Gagan Bansal Gagan Bansal(Coauthor of AutoGen; Research Scientist at Microsoft Research), Elvis Saravia Elvis Saravia(Founder of DAIR.AI), and
1 more.

agent-framework by microsoft

1.0%
12k
AI agent and multi-agent workflow framework
Created 1 year ago
Updated 1 day ago
Starred by Lilian Weng Lilian Weng(Cofounder of Thinking Machines Lab), Chip Huyen Chip Huyen(Author of "AI Engineering", "Designing Machine Learning Systems"), and
59 more.

AutoGPT by Significant-Gravitas

0.1%
185k
AI agent platform for building, deploying, and running autonomous workflows
Created 3 years ago
Updated 1 day ago
Feedback? Help us improve.