omnigent-ai/omnigent

Project overview

An orchestration and management framework for heterogeneous AI coding agents. It provides a single layer to swap or combine agent harnesses, apply governance policies, and synchronize sessions across devices.

Project type

AI Agent · AI Coding

Use cases

Coding & Development

Deployment

Refer to project documentation

License

Apache-2.0

Best for

• Developers and AI engineers who need to orchestrate multiple heterogeneous AI coding agents across different devices with consistent policies and sandboxing.

Key capabilities

• Provides a common orchestration layer over Claude Code, Codex, Cursor, and custom agents to swap or combine harnesses without rewriting.
• Sessions follow the user across terminal, browser, and phone, keeping messages, sub-agents, terminals, and files in sync.
• Run agent sessions in disposable cloud sandboxes like Modal, Daytona, E2B, CoreWeave, Kubernetes, or Databricks.
• Create policies to pause for approval before risky actions, cap spend, or limit which tools an agent can access.
• Share live sessions so teammates can chat with agents, co-drive on a machine, or fork conversations.
• Define custom agents in YAML files with prompts, local Python functions, MCP servers, and sub-agents.

Limitations and risks

• Windows runs in a degraded mode without bwrap/seatbelt sandboxing or native tmux terminal wrappers.
• The project is currently in alpha status.

Getting started

• Setup difficulty is rated as easy. Users can run a curl install script, install via PyPI package (uv tool install omnigent), or use one command for docker compose. After installation, execute the omnigent command.

Evidence and sources

• GitHub project description: Omnigent is an open-source AI agent framework and meta-harness: orchestrate Claude Code, Codex, Cursor, Pi, and custom agents — swap harnesses without rewriting, enforce policies…
• README: 
• README: Mix Claude Code, Codex, Cursor, OpenCode, Hermes, Pi, and custom agents (defined in YAML) together in the same session. Ask one agent to review another's work, or split a task acr…
• README: 📱 Work with agents from any device, including your phone.** Sessions follow you: start in your terminal, continue in the browser, pick it up on your phone. Messages, sub-agents, t…
• README: ☁️ Run agents in cloud sandboxes.** No laptop required: run sessions in disposable [Modal](https://modal.com), [Daytona](https://www.daytona.io), [Islo](https://islo.dev), [E2B](h…

Reviewed at: 2026-07-18T06:28:38Z