The gist

Hermes Workspace is an open-source, self-hosted command center for AI agents. It provides a complete workspace to orchestrate agents, browse their memory, manage skills, and control operations from a single interface. Unlike simple chat wrappers, it offers integrated tools for managing files and a full PTY terminal, unifying all aspects of agent interaction. It is designed to work with the vanilla NousResearch/hermes-agent, connecting directly to its gateway.

What it does

Connect to any OpenAI-compatible backend or a Hermes agent gateway with real-time streaming.
Browse, search, and edit an agent's memory files directly from the UI.
Manage and explore over 2,000 available agent skills from an integrated registry.
Navigate, preview, and edit workspace files using an integrated Monaco editor.
Access a full, persistent PTY terminal within the command center interface.
Install as a Progressive Web App (PWA) for a native experience on desktop and mobile.

How it works

Hermes Workspace is a self-hosted web application installed by cloning its GitHub repository and running it locally or via Docker. The user configures it to point to an OpenAI-compatible backend, such as a running Hermes agent gateway. All interaction happens through the web UI, which provides a chat interface, file browser, memory editor, and terminal. The tool is open-source and free to use under an MIT license.

Best for

This tool is best for developers and AI researchers who need a comprehensive command center to manage and interact with self-hosted AI agents, especially for tasks requiring file system access, memory inspection, and terminal control.

Watch out for

The setup is technical and requires running and configuring a separate AI agent backend, like NousResearch/hermes-agent, for full functionality. It is not a standalone, out-of-the-box chat application.