agent-desktop
agent-desktop is a command-line tool that lets AI agents read and operate desktop apps on macOS, Windows and Linux.
What it is good for
- Letting an AI agent click and type in desktop apps
- Reading what is on screen without relying on pixel matching
- Automating repetitive steps in apps that have no API
- Checking that a desktop action worked before continuing
At a glance
- Who it is for
- Setup
- Guided command-line setup
Level 4 of 5. You copy and paste commands into a terminal by following the instructions. Comfortable territory for vibe coders.
- Cost
- Free and open source
- Type
- Agent app
- Tasks
- Works with
- Claude Code, Codex, skills.sh, Homebrew, Scoop, npm
Momentum
- Rank this week
- Not ranked yet
- Best rank
- None yet
- GitHub stars
- 7,622
GitHub numbers as of Nov 2, 2026. Updated weekly.
Official links
Setup steps are on the official pages. We do not repeat them here, since they change.
Questions
- Is agent-desktop free?
- Yes. It is free and open source (Apache-2.0 license).
- Do I need to know how to code to use agent-desktop?
- You do not need to write code, but setup means copying and pasting commands into a terminal by following its instructions.
- What does agent-desktop work with?
- According to its own documentation, it works with Claude Code, Codex, skills.sh, Homebrew, Scoop, and npm.
- What is agent-desktop good for?
- Letting an AI agent click and type in desktop apps, reading what is on screen without relying on pixel matching, automating repetitive steps in apps that have no API, and checking that a desktop action worked before continuing.
Similar agents
- cua
Cua is a set of tools that gives AI agents full desktops they can operate for people who need work done on their computer.
- OpenBot
OpenBot is a desktop app that runs persistent AI teammates on your own computer, each with its own workspace and task queue, for people who want agents working in the background.
- esp32-iphone-agent
ESP32 iPhone Agent is a hardware and app setup that lets an AI model operate a real iPhone for hands-free phone tasks, using a small board instead of a computer.
- AutoCua
AutoCua is a Python agent that controls a computer, browser, or phone by clicking and typing for users who need repetitive desktop or web tasks done.
- Octopus-SI-OS
Octopus-SI-OS is a research preview agent app that runs agents on your Mac and lets them reshape their own environment, for people exploring self-directed AI work.
- iphone-use
iphone-use is a self-hosted tool that lets AI agents read and control a real iPhone for people who need phone tasks automated.