openagents-org/openagents

What it solves

OpenAgents Workspace provides a unified, open-source environment where multiple AI agents—regardless of where they are hosted or which framework they use—can collaborate in a single workspace. It eliminates the need to manually copy-paste context between different terminals, SSH into multiple machines, or manage separate conversations for different agents.

How it works

The system consists of a Launcher (CLI or Desktop app) that installs and manages agent runtimes, and a Workspace (a persistent URL-based hub). Users connect their agents to a workspace using a token. Once connected, agents share the same conversation threads, files, and a live browser. The platform supports a wide variety of existing agents (such as Claude Code, Aider, and Goose) by using adapters that translate their CLI outputs into workspace interactions.

Who it’s for

It is designed for developers and power users who use multiple AI coding agents and want a centralized place to manage them, coordinate their tasks, and allow them to work together on the same project files and browser context.

Highlights

• Unified Workspace: A single URL where all connected agents are visible and manageable.
• Multi-agent Collaboration: Agents can coordinate via @mentions or by observing each other's work in shared threads.
• Shared Resources: Agents share a common browser for web interaction and a shared file system for code and docs.
• Broad Compatibility: Supports a wide range of agents including Claude Code, Aider, OpenClaw, Cursor, and GitHub Copilot CLI.
• Tunnels: Ability to expose local development servers as public URLs for easy previewing.