An open-source Computer Use service exposed as an MCP server, letting AI agents control macOS, Linux, and Windows desktops.
leeguooooo/iphone-use
Lets AI agents like Claude Code or Codex read the screen, tap, type and swipe on a real iPhone over USB or Wi-Fi via MCP.
- Stars
- 310
- Forks
- 34
- Open issues
- 3
- Last commit
- 2 hours ago
What is it?
iphone-use is a self-hosted tool that lets AI agents operate a real iPhone. A daemon on macOS reads the screen as an accessibility tree (text rather than pixels), sends taps, typing and swipes through a small XCTest runner installed on the phone, and waits for the screen to settle before returning a result. It works in any app, including apps that expose no API, and it can be driven over USB or Wi-Fi.
Agents connect through an MCP server for Claude Code, Codex, Cursor, Claude Desktop and other MCP clients, or through an HTTP API. Each action reports whether it was applied, and taps on covered elements are refused. Repeated tasks can be saved as flows that replay without a model. The same daemon also offers a browser page and a native iOS app so a person can view the live screen and take control.
The project requires a Mac running macOS 15 or later, Xcode, and an iPhone on iOS 15 or later with Developer Mode enabled. It is licensed under PolyForm Shield 1.0.0, which is source-available rather than standard open source.
Key features
- 01Reads the screen as text from the accessibility tree, without screenshots.
- 02Taps, types and swipes in any app, including apps without an API.
- 03Connects over USB, or over Wi-Fi through an encrypted tunnel after setup.
- 04Exposes MCP tools for Claude Code, Codex, Cursor and other MCP clients, plus an HTTP API.
- 05Saves repeated tasks as flows that replay without a model or tokens.
- 06Provides a browser page and native iOS app for live viewing and manual control.
Use cases
- Letting Claude Code or Codex run multi-step tasks on a phone, such as reading app data or sending messages.
- Running login or flow tests on a real iPhone and reporting which step fails.
- Exporting data from apps that offer no API, such as health or notes apps.
- Sharing one iPhone between an agent and a person who needs to step in.
Topics
FAQ
- What is iphone-use?
- Lets AI agents like Claude Code or Codex read the screen, tap, type and swipe on a real iPhone over USB or Wi-Fi via MCP. iphone-use is a self-hosted tool that lets AI agents operate a real iPhone. A daemon on macOS reads the screen as an accessibility tree (text rather than pixels), sends taps, typing and swipes through a small XCTest runner installed on the phone, and waits for the screen to settle before returning a result. It works in any app, including apps that expose no API, and it can be driven over USB or Wi-Fi.
- Is iphone-use free to use? What license does it have?
- iphone-use is open source under the Other license. Check the license terms before using it in commercial products.
- Is iphone-use still maintained?
- Actively maintained. The last commit to iphone-use was on Oct 19, 2026.
- What can I use iphone-use for?
- Letting Claude Code or Codex run multi-step tasks on a phone, such as reading app data or sending messages. Running login or flow tests on a real iPhone and reporting which step fails. Exporting data from apps that offer no API, such as health or notes apps. Sharing one iPhone between an agent and a person who needs to step in.
- Where can I get iphone-use?
- The source code and installation instructions are on GitHub: https://github.com/leeguooooo/iphone-use
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A Codex, WorkBuddy and Doubao plugin that lets an AI assistant operate a real iPhone over USB via WebDriverAgent, with a live screen preview.
A CLI and skill layer that installs, configures and health-checks channels so AI agents can read and search Twitter, Reddit, YouTube, Bilibili, XiaoHongShu and more.
A browser built for AI agents and you to share: agents run tasks in isolated Spaces while your own tabs and logins stay untouched.
Lets an AI vision model operate a real iPhone using a $7 ESP32-S3 board that acts as an AirPlay screen receiver and USB HID mouse, with no computer needed at run time.
GitHub's official MCP server that lets AI assistants read repositories, manage issues and pull requests, and work with GitHub Actions.