danielealbano/android-remote-control-mcp

An MCP Server for Android running on the phone, optmized for token usage, supports also files downloads and cloudflare and ngrok automated tunnelling.

What it solves

This project provides a way for AI models to remotely and fully control an Android device. It eliminates the need for a host computer running ADB (Android Debug Bridge) by running the MCP server directly on the phone, allowing for lower latency, remote access over the internet, and more token-efficient communication between the AI and the device.

How it works

The application runs as a Model Context Protocol (MCP) server on Android using Ktor and Netty. It leverages Android's Accessibility Services to read UI elements, capture screenshots, and execute actions like tapping, typing, and swiping. It can be exposed to AI clients (like Claude or ChatGPT) via a local network, ADB port forwarding, or public HTTPS tunnels (Cloudflare/ngrok). To protect user data, it includes an on-device Privacy Mode that uses a local NER model to redact or pseudonymize sensitive information before it is sent to the AI provider.

Who it’s for

Developers and researchers interested in AI agents and embodied intelligence who want to give LLMs the ability to interact with mobile applications and the Android OS in a real-world, remote-capable environment.

Highlights

  • On-Device Server: Runs directly on Android, removing the dependency on a connected PC.
  • Extensive Toolset: Offers 57 MCP tools across 14 categories, including screen introspection, file operations, and app management.
  • Privacy-First: Features an on-device anonymizer to hide emails, credit cards, and other PII before data leaves the device.
  • Flexible Auth: Supports static bearer tokens and a self-contained OAuth 2.1 server for secure connections.
  • Remote Access: Integrated support for Cloudflare Quick Tunnels and ngrok for global accessibility.

Related

  • Project
    mobile-next/mobile-mcp

    Mobile MCP is an open‑source MCP server that lets AI agents control iOS/Android simulators, emulators, and real devices through a single API. It uses the native accessibility tree for fast, structured UI interaction, offers full device control (taps, swipes, app install, screen recording, etc.), and integrates with Claude, Gemini, Copilot, Codex and any MCP‑compatible client. Install via `npx @mobilenext/mobile-mcp@latest`, configure with a tiny JSON snippet, and run locally or as an SSE HTTP server.

  • Project
    joshuayoes/ios-simulator-mcp

    An MCP server that enables AI agents to control iOS simulators, perform UI interactions, and automate QA testing through programmatic tool calls.

  • Project
    witchan/ios-mcp

    An MCP server for jailbroken iPhones that enables AI agents to control the device via touch gestures, app management, and system-level access.

  • Project
    GoogleCloudPlatform/cloud-run-mcp

    An MCP server that enables AI agents to deploy and manage applications on Google Cloud Run through a standardized protocol.

  • Project
    AlexGladkov/claude-in-mobile

    An MCP server that enables AI assistants to control and debug mobile devices and desktop platforms through screenshots, UI interaction, and runtime debugging.