BeClaude

omnigent

New
4.4kGitHubGeneralby omnigent-ai

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 and sandboxing, and collaborate in real time from any device.

First seen 6/18/2026

Summary

Omnigent is an open-source AI agent framework that lets you orchestrate multiple AI agents like Claude Code, Codex, and Cursor from a single harness.

  • It enables you to swap agent backends without rewriting code, enforce policies and sandboxing, and collaborate in real time from any device, making it ideal for managing complex multi-agent workflows.

Install & Usage

1
Create the agents directory
mkdir -p .claude/agents
2
Save the agent file

Add the configuration to .claude/agents/omnigent.md

3
Invoke with @agent-name
@omnigent

Use Cases

Orchestrate Claude Code, Codex, and Cursor agents together to solve a complex software engineering task, switching between them seamlessly.
Enforce security policies and sandboxing rules across all agents in a multi-agent development pipeline to prevent data leaks.
Collaborate with a remote team in real time by sharing a single Omnigent session where each member controls different agents.
Build a custom agent that uses Claude Code for code generation and Pi for natural language reasoning, all within one framework.
Automate a CI/CD pipeline where Omnigent routes code review requests to different AI agents based on the programming language.
Migrate an existing single-agent project to a multi-agent setup without rewriting any agent-specific code, using Omnigent's harness abstraction.

Usage Examples

1

/omnigent orchestrate claude-code and codex to refactor this Python module and write unit tests

2

/omnigent enforce policy: no agent may access the /etc directory or execute shell commands

3

/omnigent collaborate --session my-project --invite [email protected] --agent claude-code,pi

View source on GitHub
agent

Security Audits

LicenseUnknownSourceWarnRepositoryPass

Frequently Asked Questions

What is omnigent?

Omnigent is an open-source AI agent framework that lets you orchestrate multiple AI agents like Claude Code, Codex, and Cursor from a single harness. It enables you to swap agent backends without rewriting code, enforce policies and sandboxing, and collaborate in real time from any device, making it ideal for managing complex multi-agent workflows.

How to install omnigent?

To install omnigent: create the agents directory (mkdir -p .claude/agents), then add the config to .claude/agents/omnigent.md. Finally, @omnigent in Claude Code.

What is omnigent best for?

omnigent is a agent categorized under General. It is designed for: agent. Created by omnigent-ai.

What can I use omnigent for?

omnigent is useful for: Orchestrate Claude Code, Codex, and Cursor agents together to solve a complex software engineering task, switching between them seamlessly.; Enforce security policies and sandboxing rules across all agents in a multi-agent development pipeline to prevent data leaks.; Collaborate with a remote team in real time by sharing a single Omnigent session where each member controls different agents.; Build a custom agent that uses Claude Code for code generation and Pi for natural language reasoning, all within one framework.; Automate a CI/CD pipeline where Omnigent routes code review requests to different AI agents based on the programming language.; Migrate an existing single-agent project to a multi-agent setup without rewriting any agent-specific code, using Omnigent's harness abstraction..