claudekit / tools / mobile-mcp
[ MCP · Development ]

Mobile MCP

MCP server for mobile automation and scraping on iOS, Android, emulators, simulators and real devices. Agents interact with native apps through structured accessibility snapshots or coordinate-based taps based on screenshots.

mobile-next/mobile-mcp ·updated
$ claude mcp add mobile-mcp -- npx -y @mobilenext/mobile-mcp@latest copy

Open source project by Mobile Next (mobile-next/mobile-mcp, Apache-2.0).

Why use it?

Checking how a mobile app actually behaves usually means tapping through a simulator by hand, or writing scripts in a different framework per platform (XCUITest on iOS, Espresso on Android). A coding agent can change the app’s code, but it can’t see or drive the result on a device.

Mobile MCP gives the agent that ability. It interacts with native iOS/Android applications and devices through structured accessibility snapshots, or coordinate-based taps based on screenshots, and the same tools work on every target, so no platform expertise is required.

What you can do with it

The README lists these main use cases:

  • Native app automation (iOS and Android) for testing or data-entry scenarios
  • Scripted flows and form interactions without manually controlling simulators/emulators or real devices (iPhone, Samsung, Google Pixel etc)
  • Automating multi-step user journeys driven by an LLM
  • General-purpose mobile application interaction for agent-based frameworks, including data extraction

In a development loop, after changing the app you can ask the agent to “open the login screen on the simulator, sign in with the test account, and navigate to Settings.” The agent lists the elements on screen, taps and types its way through each step, and checks the result. If the app crashes along the way, it can pull device logs and crash reports.

Key features

  • Accessibility-first: fast and cheap

    Drives apps from the native accessibility tree (no vision model, no image tokens), falling back to screenshots + coordinates only when needed.

  • One API, every target

    The same tools work across iOS and Android: iOS Simulator, iOS real device over USB, Android Emulator, and Android real device with USB debugging.

  • Full device control

    Taps, double taps, long presses, swipes, typing, hardware buttons (HOME, BACK, VOLUME_UP/DOWN, ENTER), orientation, GPS location override, and clipboard. App install (.apk, .ipa, .app, .zip), launch, terminate, and uninstall.

  • Screen recording, logs, and crash reports

    Record the device screen to a video file, collect live device logs (logcat on Android, unified log on iOS), list crash reports, and fetch the full content of a crash report.

  • Batch commands

    mobile_batch_commands runs multiple tools in sequence in a single call (e.g. click, type, click), optionally listing screen elements at the end.

How to use

Prerequisites: node.js v20+, plus Xcode command line tools for iOS and Android Platform Tools (adb) for Android.

Add it to Claude Code:

claude mcp add mobile-mcp -- npx -y @mobilenext/mobile-mcp@latest

For Claude Desktop and other MCP clients, use the standard config:

{
  "mcpServers": {
    "mobile-mcp": {
      "command": "npx",
      "args": ["-y", "@mobilenext/mobile-mcp@latest"]
    }
  }
}

Start a simulator or emulator first. For iOS, boot one with xcrun simctl boot "iPhone 16". Then verify it works by asking your agent:

list available devices

You should get back your running simulators, emulators, and connected devices. If the list is empty, make sure a simulator or emulator is running.

Notes

  • Telemetry is on by default — Mobile MCP collects anonymous usage telemetry via PostHog and Scarf (tool name and duration on each call). Add "MOBILEMCP_DISABLE_TELEMETRY": "1" to the env block of your config to disable it.
  • Platform requirements — iOS Simulators on macOS/Linux, Android Emulators on Linux/Windows/macOS. Real iOS devices must be connected over USB and trusted; Android devices need USB debugging enabled and authorized.
  • Streamable HTTP mode — It runs over stdio by default. --listen 3000 starts a Streamable HTTP server instead. Without MOBILEMCP_AUTH, the server accepts unauthenticated connections, so set a token before exposing it.
  • Repeatable tests — For turning agent-driven exploration into repeatable, deterministic tests, the authors point to mobilewright.
  • Cloud devices are optional — The four remote-device tools are for Mobile Next Cloud, which rents real iOS and Android devices. They aren’t needed for local devices.
  • Source — mobile-next/mobile-mcp (Apache-2.0).
§ 6

See also

same category · curated
[01]
[Plugin] Ponytail · Makes your AI agent think like the laziest senior dev in the room. Before writing code the agent stops at the first rung that holds — does this need to exist, is it already in the codebase, does the stdlib do it — and never cuts validation, error handling, security, or accessibility.
tool · claudekit.io / tools / ponytail
[02]
[Tool] Paseo · Run any coding agent from your phone, desktop, or terminal. Self-hosted, multi-provider, open source. A local daemon runs the agents on your machine and every client — desktop, mobile, web, CLI — connects to it, with the native mobile app at full feature parity with desktop.
tool · claudekit.io / tools / paseo
[03]
[Tool] Orca · Orca is the ADE for working with a fleet of parallel agents. Run any coding agent with your own subscription. It runs each agent in an isolated git worktree, lets you compare outputs side by side, and merge the ones you want. Works with Claude Code, Codex, Grok, and 30+ CLI agents. Available on desktop and mobile.
tool · claudekit.io / tools / orca
§ 7

Frequently Asked Questions

frequently asked
§ 7.1
What is Mobile MCP?
An MCP server that enables scalable mobile automation through a platform-agnostic interface. Agents and LLMs interact with native iOS and Android apps on emulators, simulators, and real devices.
§ 7.2
How do I install Mobile MCP in Claude Code?
Run `claude mcp add mobile-mcp -- npx -y @mobilenext/mobile-mcp@latest`. You also need node.js v20+, plus Xcode command line tools for iOS or Android Platform Tools for Android.
§ 7.3
Do I need to know XCUITest or Espresso?
No. There is no XCUITest, no Espresso, and no per-platform glue: the same tools work across iOS and Android, and you describe the goal while the agent does it.
§ 7.4
Does it rely on screenshots?
It is accessibility-first: it drives apps from the native accessibility tree with no vision model and no image tokens. It falls back to screenshots and coordinates only when needed.
§ 7.5
Is Mobile MCP free?
It is open source under Apache-2.0 and runs against simulators, emulators, and devices on your own machine. Mobile Next Cloud, which provides real iOS and Android devices in the cloud, is a separate service.
§ 7.6
Does Mobile MCP collect telemetry?
Yes, it collects anonymous usage telemetry via PostHog and Scarf. Set `MOBILEMCP_DISABLE_TELEMETRY=1` to disable it.