Cette page est affichée en anglais. Une traduction française est en cours.
ConfigsMay 29, 2026·2 min de lecture

Oh My OpenAgent — Universal Agent Harness Platform

Oh My OpenAgent (omo) is an open-source agent harness that unifies Claude Code, Codex, Gemini, Cursor, and other coding agents into a single orchestration layer with shared context and tooling.

Prêt pour agents

Installation agent prête

Cet actif peut être installé après choix du runtime, vérification du plan et exécution de la commande adaptée.

Native · 98/100Policy : autoriser
Surface agent
Tout agent MCP/CLI
Type
Skill
Installation
Single
Confiance
Confiance : Established
Point d'entrée
Oh My OpenAgent Overview
Commande d'installation directe
npx -y tokrepo@latest install 0823a584-5b37-11f1-9bc6-00163e2b0d79 --target codex

À exécuter après confirmation du plan en dry-run.

Introduction

Oh My OpenAgent (omo) is an open-source harness that wraps multiple AI coding agents — Claude Code, Codex, Gemini, Cursor, and others — under one unified interface. It lets developers switch between agents, share context, and orchestrate multi-agent workflows from a single TUI.

What Oh My OpenAgent Does

  • Provides a unified terminal interface across multiple coding agent backends
  • Shares conversation context and memory between different agents
  • Supports hot-swapping between agents mid-session without losing state
  • Offers plugin architecture for extending agent capabilities
  • Includes built-in cost tracking and token usage monitoring

Architecture Overview

omo runs as a TypeScript process that manages agent sessions through adapter modules. Each supported agent (Claude Code, Codex, etc.) has a dedicated adapter that translates omo's internal protocol into the agent's native CLI interface. A shared context store backed by SQLite keeps conversation history accessible across agent switches.

Self-Hosting & Configuration

  • Install globally via npm or run from source with Node.js 20+
  • Run omo init to create a configuration file in your project
  • Add API keys for each agent backend in environment variables
  • Configure default agent, fallback agent, and routing rules
  • Set token budgets and cost limits per session or per day

Key Features

  • Multi-agent switching with preserved context
  • Unified command palette across all supported agents
  • Token and cost dashboard with per-agent breakdowns
  • Plugin system for custom tools and integrations
  • Session recording and replay for debugging workflows

Comparison with Similar Tools

  • Claude Code standalone — single agent; omo adds multi-agent orchestration
  • ECC — focuses on performance optimization; omo focuses on agent switching
  • Superset — desktop-first editor; omo is CLI-first and lightweight
  • Vibe Kanban — task management layer; omo is the agent runtime layer

FAQ

Q: Can I use omo with only one agent? A: Yes. It works as a lightweight wrapper even for a single agent, adding cost tracking and session management.

Q: Does omo modify agent behavior? A: No. It passes through commands to the native agent CLI and only adds orchestration around it.

Q: What agents are supported? A: Claude Code, Codex CLI, Gemini CLI, OpenCode, Cursor, and others via adapter plugins.

Q: Is there a VS Code extension? A: omo is CLI-focused. IDE integration is available through terminal panels.

Sources

Fil de discussion

Connectez-vous pour rejoindre la discussion.
Aucun commentaire pour l'instant. Soyez le premier à partager votre avis.

Actifs similaires