An open-source Mac app that gives Claude Code a real browser, a live iOS Simulator and its own git worktree per chat.
littleguys
Summary
Free, open-source Mac app that lets you chat with, approve and schedule Claude Code, Codex, OpenClaw and Hermes agents like contacts.
Screenshots
Description
littleguys is a free, open-source macOS app that gives every AI agent you run its own contact and chat thread, in the style of a messaging app. Instead of hunting through terminal tabs, you talk to Claude Code, Codex, Gemini CLI, Goose, OpenCode, OpenClaw or Hermes agents from one window and watch a small character in the menu bar act out who is working and who needs you.
Commands that need permission appear as approval cards (allow once, always allow or deny) rather than a buried [y/N] prompt. A status view shows every agent on every machine with a live ring, the last message and a quick reply, and finds gateways on your tailnet and LAN automatically, so agents on a Mac mini, a homelab and a laptop sit side by side. Routines let you schedule recurring jobs such as morning digests or nightly dependency bumps, with run-now and results delivered into a thread.
littleguys speaks the Agent Client Protocol, so any ACP agent can be added from the registry. It is a friendly front end for agents you already run, not a new place to run them. The app is MIT-licensed, built with a Rust core and React on Tauri, talks directly to your agents with no account, cloud or telemetry, and installs as a signed macOS 13+ build for Apple silicon and Intel or via Homebrew.
Reviews
No reviews yet — sign in to write the first one
Similar App Suggestions
Free, open-source macOS menu bar app that shows live status for every parallel Codex task, so you can see which agent needs you without reopening Codex.
SSH terminal for iPhone and iPad that alerts you when a coding agent is blocked or done, so you can answer it from anywhere.
A macOS script runner where you and your coding agent drive the same terminal processes — menu bar, CLI and a built-in MCP server over one local socket.
Menu-bar LLM inference server for Apple Silicon, with continuous batching and tiered KV caching that keeps local models fast enough for real coding work.