🚀 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.
- Ensure
uv
is installed. (See Prerequisites). - 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 useuvx
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.
-
Clone the Repository: Open your terminal and clone the DroidMind repository from GitHub:
-
Create and Activate Virtual Environment: Using
uv
, create a virtual environment:Activate the environment:
- On macOS and Linux:
- On Windows (PowerShell):
- On Windows (CMD):
-
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.):
- For Stdio Transport (Direct terminal interaction):
- For Development (includes all dependencies, plus dev tools):
(Note:
dev
typically includesstdio
andsse
specific dependencies if structured that way inpyproject.toml
)
🏃♀️ Running DroidMind¶
After installation (primarily for Method 2, as uvx
runs it directly):
- Stdio Mode (Direct Terminal Interaction):
Or, to be explicit:
- SSE Mode (for AI Assistants like Claude Desktop, Web UIs):
By default, this will start an SSE server at
http://localhost:4256
. The MCP connection URI for your AI assistant will typically besse://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! 💫