← Today's AI open source

Today's AI open source · 2026-11-17

milind-soni/OpenMausBot

OpenMausBot is a local-first desktop app for running multiple AI agents like contacts in a chat app.

AgentsCoding toolsApps · interfaces

Repositorygithub.com/milind-soni/OpenMausBot ↗
Main languageTypeScript
LicenseApache-2.0
GitHub stars · forks★ 24,182 · 3,356 forks (as of 2026-11-17)
Created · last update2026-08-11 · 2026-11-16
GitHub topicsai-agents, claude-code, codex, computer-use, docker, electron, grok, grok-bot

Overview

OpenMausBot is an open-source desktop app that places multiple AI bots in a sidebar like messenger contacts, letting you chat with them and assign them work. According to the README, each bot is a real agent with its own personality, model, computer and connected apps.

Bots run on top of the claude, codex and grok CLIs installed on the user's computer, so they use your existing logins and subscriptions with no proxy in between. A small harness server running on 127.0.0.1 manages all agent processes, and chat history, keys and events are stored in ~/.openmausbot rather than in the cloud. Each bot can use a cloud Linux desktop or an isolated Local VM to do actual work.

It suits developers and practitioners who already use Claude Code or Codex and want to split roles across several agents, watch their progress and approve only the important steps. Installers are provided for macOS, Windows and Ubuntu.

Why it is worth a look

Its distinguishing design treats agents not as a single assistant window but as multiple bots in a messenger. Because it uses the official agent CLIs the user has already installed as its engine, with no new account or intermediary server, and keeps data local, it is an implementation you can use to experiment directly with running several agents together.

Key features

What it is used for

Getting started

  1. Download the installer for your operating system (.dmg, .exe, .deb) from the official site or GitHub Releases and install it.
  2. In the engine check step of onboarding, install and log in to a supported agent such as Claude Code or Codex (this can also be done under App settings → Agent engines).
  3. Click New or share → New bot, set a name, and choose a connected engine and an available model (a working folder is needed only for file tasks).
  4. Send one small, clear task, such as organizing the day's schedule or summarizing notes, and check the result.

What to weigh

Related terms and companies

Sources

Written with AI assistance from the repository README and GitHub data. Star counts and other numbers are GitHub values as of the date shown; features and licensing can change, so check the repository.