Skip to main content
Glama

JetsonMCP

by ajeetraina
QUICKSTART.mdโ€ข5.18 kB
# JetsonMCP Quick Start Guide Get up and running with JetsonMCP in 15 minutes! ## ๐Ÿ Prerequisites - **Jetson Nano** with JetPack 4.6+ installed - **Development Machine** with Python 3.8+ - **Network Connection** between your machine and Jetson - **Claude Desktop** installed on your development machine ## ๐Ÿš€ Step 1: Prepare Your Jetson Nano ### Option A: Automatic Setup (Recommended) ```bash # On your Jetson Nano, run: wget -O - https://raw.githubusercontent.com/ajeetraina/jetsonMCP/main/scripts/setup-jetson.sh | bash ``` ### Option B: Manual Setup ```bash # Enable SSH sudo systemctl enable ssh sudo systemctl start ssh # Configure passwordless sudo (for system management) sudo visudo # Add: yourusername ALL=(ALL) NOPASSWD:ALL # Get your Jetson's IP address hostname -I ``` ## ๐Ÿ“‹ Step 2: Install JetsonMCP ### On Your Development Machine: ```bash # Clone the repository git clone https://github.com/ajeetraina/jetsonMCP.git cd jetsonMCP # Create virtual environment python -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate # Install JetsonMCP pip install -e . ``` ## โš™๏ธ Step 3: Configure Connection ```bash # Copy configuration template cp .env.example .env # Edit configuration nano .env # or your preferred editor ``` **Minimum required settings in `.env`:** ```bash JETSON_HOST=192.168.1.100 # Your Jetson's IP address JETSON_USERNAME=your_username # SSH username JETSON_PASSWORD=your_password # SSH password ``` ## ๐Ÿงช Step 4: Test Connection ```bash # Test SSH connection to your Jetson jetsonmcp test-connection # Should show: โœ… Connection successful! ``` ## ๐Ÿ”— Step 5: Configure Claude Desktop ### Option A: Automatic Setup ```bash # Run the integration script ./scripts/install-claude-desktop.sh ``` ### Option B: Manual Setup Add to your Claude Desktop configuration: **macOS**: `~/Library/Application Support/Claude/claude_desktop_config.json` **Windows**: `%APPDATA%\Claude\claude_desktop_config.json` **Linux**: `~/.config/Claude/claude_desktop_config.json` ```json { "mcpServers": { "jetsonmcp": { "command": "python", "args": ["-m", "jetsonmcp.server"], "cwd": "/absolute/path/to/jetsonMCP" } } } ``` ## ๐ŸŽ‰ Step 6: Start Using JetsonMCP 1. **Restart Claude Desktop** to load the new configuration 2. **Test the integration** by asking Claude: > "What's the current temperature on my Jetson Nano?" > "Switch my Jetson to 5W power mode" > "Check if CUDA is working on my Jetson" > "Install TensorFlow on my Jetson Nano" ## ๐Ÿ“‹ Common First Commands ### System Information - "Get system information from my Jetson" - "Check disk space on my Jetson Nano" - "What services are running on my Jetson?" ### Hardware Management - "What's the current power mode?" - "Monitor GPU temperature for 30 seconds" - "Check GPU memory usage" ### AI/ML Setup - "Check if CUDA is installed and working" - "Install PyTorch on my Jetson" - "Set up Jupyter Notebook for remote access" ### Container Operations - "Check if Docker is running" - "Install NVIDIA Container Runtime" - "List all Docker containers" ## ๐Ÿ”ง Troubleshooting ### Connection Issues **Problem**: โŒ Connection failed **Solutions**: 1. Verify Jetson IP address: `hostname -I` (on Jetson) 2. Test SSH manually: `ssh username@jetson-ip` 3. Check firewall: `sudo ufw status` (on Jetson) 4. Verify SSH service: `sudo systemctl status ssh` (on Jetson) ### Permission Issues **Problem**: "sudo: no password" errors **Solution**: Configure passwordless sudo on Jetson: ```bash sudo visudo # Add: yourusername ALL=(ALL) NOPASSWD:ALL ``` ### Claude Desktop Issues **Problem**: Claude doesn't recognize JetsonMCP **Solutions**: 1. Verify configuration file location 2. Check that Python can import jetsonmcp: `python -c "import jetsonmcp"` 3. Restart Claude Desktop completely 4. Check Claude Desktop logs ### Python Import Errors **Problem**: "ModuleNotFoundError: No module named 'jetsonmcp'" **Solution**: Install in development mode: ```bash cd jetsonMCP pip install -e . ``` ## ๐Ÿ“„ Next Steps ### Explore Advanced Features - Set up model deployment workflows - Configure container-based AI workloads - Set up monitoring and alerting - Explore GPIO control capabilities ### Development - Read the [Architecture Guide](ARCHITECTURE.md) - Check out [Contributing Guidelines](../CONTRIBUTING.md) - Run the test suite: `make test` ### Production Deployment - Set up SSH key authentication - Configure system monitoring - Set up automated backups - Review security settings ## ๐Ÿ“ž Support If you encounter issues: 1. **Check the logs**: Look for error messages in the console 2. **Test components**: Use `jetsonmcp test-connection` and `make test` 3. **Search issues**: Check [GitHub Issues](https://github.com/ajeetraina/jetsonMCP/issues) 4. **Ask for help**: Create a new issue with: - Your setup details - Error messages - Steps to reproduce --- **Congratulations! ๐ŸŽ‰ You're now ready to manage your Jetson Nano with AI assistance through JetsonMCP!** For more advanced usage, check out the complete [README](../README.md) and [documentation](/).

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/ajeetraina/jetsonMCP'

If you have feedback or need assistance with the MCP directory API, please join our Discord server