AI AI Toolkit
Taking Off TodayPythonApache-2.0

Omnigent: Making Multi-Agent Orchestration Actually Usable

⭐ 4.6K Stars 🧭 542 Forks 🕓 Jun 24, 2026
🚀
Why it's hot: Unified meta-harness for multiple AI coding agents (Claude Code, Codex, Cursor) with real-time collaboration.

👤 Who This Is For

AI 应用开发者需要多 Agent 协同的团队对 Agent 治理有兴趣的研究者

If you"re using Claude Code or Cursor, you"ve probably hit this problem: these AI coding tools don"t talk to each other. You want one agent writing code while another does review, but you"re manually copying context between them.

Omnigent solves exactly this. It builds a "meta-harness" — a unified orchestration layer where Claude Code, Codex, Cursor, and your custom agents can all participate in a shared session. Start in your terminal, continue in the browser, pick it up on your phone. The session follows you.

The most practical bit is its sandbox and policy enforcement. You can require human approval for certain actions, or restrict agents to specific directories. For teams wanting to adopt AI coding tools but worried about security, this design is thoughtful.

Technically: Python 3.12+, clean architecture. They"ve got cross-device support including a mobile client. Why"s it trending? I"d guess Claude Code and Cursor"s user bases are growing fast, and people are starting to hit the "multi-agent coordination" pain point for real.

Heads up: it"s still Alpha, API might change. But Apache 2.0 means you can totally use it commercially or fork it. If your team is already using multiple AI coding tools, this project is worth watching.

🚀

Get Started

Open Source · Commercial Friendly

Quick Install
pip install omnigent
MIT License Python 🧭 542 forks
← Back to TrendingData source: GitHub API