An open-source, free AI agent app for iOS and Android that gives models like Claude and GPT a sandboxed Linux shell, browser automation and device tools.
mobile-next/mobile-mcp
An MCP server that lets AI agents and LLMs automate iOS and Android apps on simulators, emulators and real devices.
- Stars
- 9,755
- Forks
- 867
- Open issues
- 61
- Last commit
- 2 days ago
What is it?
Mobile MCP is a Model Context Protocol server that gives AI agents and LLMs a platform-agnostic way to control native iOS and Android apps. The same set of tools works across iOS simulators, Android emulators, and physical iOS and Android devices, so developers do not need to write XCUITest or Espresso code for each platform.
The server primarily drives apps through the native accessibility tree, which returns structured UI elements. When that is not enough, it falls back to screenshots and coordinate-based taps. The toolset covers device management, app install and launch, gestures, hardware buttons, deep links, screen recording, clipboard access, device logs and crash reports. It can run over stdio or as a Streamable HTTP server with optional Bearer token authorization.
Mobile MCP works with clients such as Claude Code, Cursor, Codex, Gemini CLI and Claude Desktop. Cloud-hosted real devices are available through the vendor's Mobile Next Cloud service.
Key features
- 01Drives apps via the native accessibility tree, with screenshot and coordinate fallback
- 02One tool set across iOS simulators, Android emulators and real devices
- 03Covers app install/launch, gestures, hardware buttons, deep links and orientation
- 04Provides screen recording, device logs and crash report retrieval
- 05Runs over stdio or Streamable HTTP with optional Bearer token auth
Use cases
- Letting an LLM agent run multi-step end-to-end test flows on a mobile app
- Automating repetitive form entry or data-entry tasks on native apps
- Exploring app behavior interactively from an AI coding assistant
- Running mobile automation in CI with simulators or cloud devices
Quick start
- 1
Install Node.js v20 or newer
- 2
Connect an iOS or Android device or start a simulator/emulator
- 3
Add the MCP server to your agent with the standard config (npx runs it)
npx -y @mobilenext/mobile-mcp@latest
Things to know
- Requires Xcode command line tools for iOS and Android Platform Tools for Android.
- Requires Node.js v20 or newer.
- Some features such as device mirroring only work in Claude Code under Ghostty or kitty.
Topics
FAQ
- What is mobile-mcp?
- An MCP server that lets AI agents and LLMs automate iOS and Android apps on simulators, emulators and real devices. Mobile MCP is a Model Context Protocol server that gives AI agents and LLMs a platform-agnostic way to control native iOS and Android apps. The same set of tools works across iOS simulators, Android emulators, and physical iOS and Android devices, so developers do not need to write XCUITest or Espresso code for each platform.
- Is mobile-mcp free to use? What license does it have?
- mobile-mcp is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
- Is mobile-mcp still maintained?
- Actively maintained. The last commit to mobile-mcp was on Oct 18, 2026.
- What can I use mobile-mcp for?
- Letting an LLM agent run multi-step end-to-end test flows on a mobile app Automating repetitive form entry or data-entry tasks on native apps Exploring app behavior interactively from an AI coding assistant Running mobile automation in CI with simulators or cloud devices
- How do I install mobile-mcp?
- Install Node.js v20 or newer Connect an iOS or Android device or start a simulator/emulator Add the MCP server to your agent with the standard config (npx runs it): npx -y @mobilenext/mobile-mcp@latest
- Where can I get mobile-mcp?
- The source code and installation instructions are on GitHub: https://github.com/mobile-next/mobile-mcp
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Natural-language driven Android automation for AI coding assistants and test suites, running on real phones or emulators.
A self-improving AI agent that creates and refines skills from experience, runs on many model providers, and works through CLI and messaging apps.
Compresses tool outputs, logs, files and RAG chunks before they reach an LLM, as a library, proxy or MCP server.
An open-source autonomous research agent that plans questions, gathers web and local sources, and writes cited reports with any LLM provider.
An open agent framework where the AI agent gets its own Dockerized Linux desktop, browser, and document tools to carry out real work.