Skip to content

🚀 Installation Guide

Get DroidMind up and running on your system. This guide covers the primary ways to install DroidMind, whether you want a quick setup for IDE integration or a full manual installation.

📋 Prerequisites

Before you begin, ensure you meet the following requirements:

  • Python: DroidMind requires Python 3.13 or higher. You can download Python from the official website.
  • UV: We strongly recommend using uv for project and package management. It's a fast, modern Python package installer and resolver. Follow the official uv installation guide.
  • Android Device: An Android device (physical or emulator) with USB debugging enabled.
  • ADB (Android Debug Bridge): ADB must be installed and accessible in your system's PATH. ADB is part of the Android SDK Platform Tools.
  • (Optional for Network Control): If you plan to connect to devices over Wi-Fi, ensure your Android device has ADB over TCP/IP enabled and is on the same network as the machine running DroidMind.

⚙️ Installation Methods

Choose the installation method that best suits your needs:

Method 1: Quick IDE Integration (via uvx)

This method is ideal for quickly running DroidMind as an MCP server, for example, within an IDE that supports the Model Context Protocol (like Cursor). It uses uvx to run DroidMind directly from its latest version on GitHub, meaning no manual cloning or installation of DroidMind is required first.

This setup configures your IDE to launch DroidMind, typically using stdio transport for direct interaction or sse if needed by the client.

  1. Ensure uv is installed. (See Prerequisites).
  2. Configure your IDE: You'll instruct your IDE to run DroidMind by adding a configuration to its MCP server settings file (e.g., .cursor/mcp.json for Cursor). See the Quick Start Guide for the exact JSON configuration and details on how your IDE will use uvx to fetch and run DroidMind.

This uvx-based approach is excellent for a seamless experience, development, and testing, as your IDE handles DroidMind's lifecycle.

Method 2: Manual Installation (from Source)

This method gives you a local copy of the DroidMind codebase, allowing for development or more permanent setups.

  1. Clone the Repository: Open your terminal and clone the DroidMind repository from GitHub:

    git clone https://github.com/hyperb1iss/droidmind.git
    cd droidmind
    
  2. Create and Activate Virtual Environment: Using uv, create a virtual environment:

    uv venv
    

    Activate the environment:

    • On macOS and Linux:
      source .venv/bin/activate
      
    • On Windows (PowerShell):
      .venv\Scripts\Activate.ps1
      
    • On Windows (CMD):
      .venv\Scripts\activate.bat
      
  3. Install Dependencies: With the virtual environment activated, install DroidMind and its dependencies. Choose the extras based on your needs:

    • For SSE Transport (Recommended for AI Assistant web UIs, Claude Desktop, etc.):
      uv pip install -e .[sse]
      
    • For Stdio Transport (Direct terminal interaction):
      uv pip install -e .[stdio]
      
    • For Development (includes all dependencies, plus dev tools):
      uv pip install -e .[dev,sse]
      
      (Note: dev typically includes stdio and sse specific dependencies if structured that way in pyproject.toml)

🏃‍♀️ Running DroidMind

After installation (primarily for Method 2, as uvx runs it directly):

  • Stdio Mode (Direct Terminal Interaction):
droidmind

Or, to be explicit:

droidmind --transport stdio
  • SSE Mode (for AI Assistants like Claude Desktop, Web UIs):
    droidmind --transport sse
    
    By default, this will start an SSE server at http://localhost:4256. The MCP connection URI for your AI assistant will typically be sse://localhost:4256/sse.

🐳 Docker Installation

For a containerized setup, DroidMind can also be run using Docker. This is useful for creating a consistent environment and simplifying deployment. Refer to our Docker Guide for detailed instructions.

✅ Next Steps

With DroidMind installed and running:

  • Quick Start Guide: Learn how to connect DroidMind to your AI assistant and start issuing commands.
  • Configure your AI Assistant: Refer to your AI assistant's documentation (e.g., Claude Desktop, Cursor) on how to connect to an MCP server using the appropriate URI (e.g., sse://localhost:4256/sse for SSE mode).

Happy DroidMinding! 💫