trackmcp
Back to directory

Control your Android devices with AI using Model Context Protocol Python-based implementation.

252 stars PythonAI & Machine Learning Updated Oct 18, 2025
adbaiandroidclaudecontextllmmcpmodelprotocol

Documentation

DroidMind is a powerful bridge between AI assistants and Android devices, enabling control, debugging, and system analysis through natural language. By implementing the Model Context Protocol (MCP), DroidMind allows AI models to directly interact with Android devices via ADB in a secure, structured way. When used as part of an agentic coding workflow, DroidMind can enable your assistant to build and debug with your device directly in the loop.

๐Ÿ’ซ Core Features

DroidMind empowers AI assistants to:

  • ๐Ÿ“ฑ Manage Devices: Connect via USB/TCP-IP, list devices, view properties, and reboot.
  • ๐Ÿ“Š Analyze Systems: Access logs (logcat, ANR, crash, battery), capture bug reports, and dump heap.
  • ๐Ÿ“‚ Handle Files: Browse, read, write, push, pull, delete, and manage device files/directories.
  • ๐Ÿ“ฆ Control Apps: Install, uninstall, start, stop, clear data, and inspect app details (manifest, permissions, activities).
  • ๐Ÿ–ผ๏ธ Automate UI: Perform taps, swipes, text input, and key presses.
  • ๐Ÿš Execute Shell Commands: Run ADB shell commands with a security-conscious framework.
  • ๐Ÿ”’ Operate Securely: Benefit from command validation, risk assessment, and sanitization.
  • ๐Ÿ’ฌ Integrate Seamlessly: Connect with any MCP-compatible client (Claude, Cursor, Cline, etc.).

For a detailed list of capabilities, see the **User Manual and MCP Reference**.

๐Ÿš€ Getting Started

Quickstart for IDEs (Zero Install with `uvx`)

For the fastest way to integrate DroidMind with an MCP-compatible IDE (like Cursor), you can configure it to run DroidMind directly from its GitHub repository using `uvx`. This method does not require you to manually clone or install DroidMind first.

Add the following to your IDE's MCP configuration (e.g., `.cursor/mcp.json` for Cursor):

json
{
  "mcpServers": {
    "droidmind": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/hyperb1iss/droidmind",
        "droidmind",
        "--transport",
        "stdio" // The default and preferred mode for most IDE integrations
      ]
    }
  }
}

Your IDE will be configured to launch DroidMind on demand. Full instructions for this setup are in the **Quick Start Guide**.

Prerequisites

  • Python 3.13 (3.14 not yet supported)
  • `uv` (Python package manager)
  • Android device with USB debugging enabled
  • ADB (Android Debug Bridge) installed and in your system's PATH

Installation

For detailed instructions on setting up DroidMind, including the quick IDE integration with `uvx` (covered in the Quick Start), manual installation from source, or using Docker, please see our comprehensive **Installation Guide**.

Running DroidMind

How you run DroidMind depends on your setup:

  • IDE Integration (via `uvx`): Your IDE automatically manages running DroidMind as configured in its MCP settings (e.g., `mcp.json`). See the Quick Start Guide.
  • Manual Installation: After installing from source, you can run DroidMind directly.
    • Stdio (for direct terminal interaction or some IDE setups):
bash
droidmind --transport stdio
    bash
    droidmind --transport sse

    This usually starts a server at `sse://localhost:4256/sse`.

    • Docker: Refer to the Docker Guide for commands to run DroidMind in a container.

    Refer to the **Installation Guide** for more details on running DroidMind in different environments.

    ๐Ÿณ Running with Docker

    DroidMind can also be run using Docker for a consistent, containerized environment. This is particularly useful for deployment and isolating dependencies.

    For comprehensive instructions on building the Docker image and running DroidMind in a container with `stdio` or `SSE` transport, including notes on ADB device access, please refer to our **Docker Guide**.

    ๐Ÿ”ฎ Example AI Assistant Queries

    With an AI assistant connected to DroidMind, you can make requests like:

    • "List all connected Android devices and show their properties."
    • "Take a screenshot of my Pixel."
    • "Install this APK on `emulator-5554`."
    • "Show me the recent crash logs from `your_device_serial`."
    • "Tap the 'Next' button on the current screen of `emulator-5554`."

    For more inspiration, check out our **Example Queries and Workflows** in the User Manual.

    ๐Ÿ”’ Security

    DroidMind incorporates a security framework to protect your devices:

    • Command Validation & Sanitization
    • Risk Assessment Categorization
    • Protected Path Operations
    • Comprehensive Logging

    High-risk operations are flagged, and critical ones are blocked by default. Learn more in our **Security Considerations** chapter.

    ๐Ÿ’ป Development

    DroidMind uses `uv` for dependency management and development workflows.

    bash
    # Install/update dependencies (creates/updates `.venv`)
    uv sync --all-groups
    
    # Run tests
    uv run pytest
    
    # Run linting
    uv run ruff check .
    
    # Run type checking
    uv run pyright

    ๐Ÿค Contributing

    Contributions are welcome! Please feel free to submit a Pull Request.

    1. Fork the repository.

    2. Create your feature branch (`git checkout -b feature/amazing-feature`).

    3. Set up your development environment with `uv`.

    4. Make your changes.

    5. Run tests, linting, and type checking.

    6. Commit your changes (`git commit -m 'Add some amazing feature'`).

    7. Push to the branch (`git push origin feature/amazing-feature`).

    8. Open a Pull Request.

    ๐Ÿ“ License

    This project is licensed under the Apache License - see the LICENSE file for details.


    Frequently asked questions

    What is droidmind?

    droidmind is Control your Android devices with AI using Model Context Protocol Python-based implementation.

    How do I install droidmind?

    Open the GitHub repository and follow its README. Most MCP servers are added to your client's MCP config, then called by your agent.

    Is droidmind open source?

    Yes โ€” it is hosted on GitHub at https://github.com/hyperb1iss/droidmind and has 252 stars.

    Related MCP tools

    Run your own MCP server? See who uses it and what to fix.

    Measure it with TrackMCP