@yamaru-eu/hardware-probe
v1.0.3
Published
Official Yamaru Hardware Probe — MCP local agent for hardware inventory and AI diagnostics
Readme
Yamaru Hardware Probe (MCP)
Official Model Context Protocol (MCP) server for hardware inventory and AI-assisted diagnostics. Part of the Yamaru ecosystem.
📦 Official Repository
The official standalone repository for this project is: github.com/yamaru-eu/hardware-probe.
🚀 Quick Start (for Claude Desktop)
Since this package is hosted on GitHub Packages, you may need to authenticate your local npm client first if you haven't already.
1. Installation via npx
Add this to your claude_desktop_config.json:
{
"mcpServers": {
"yamaru-probe": {
"command": "npx",
"args": ["-y", "@yamaru-eu/hardware-probe"]
}
}
}2. Manual Installation
If you prefer to install it globally:
# Create or edit ~/.npmrc to include the GitHub Packages registry
echo "@yamaru-eu:registry=https://npm.pkg.github.com" >> ~/.npmrc
# Install the probe
npm install -g @yamaru-eu/hardware-probe
# Then use 'yamaru-probe' as the command in your configKey Features
- Deep Hardware Analysis: Reports CPU, GPU (VRAM), detailed RAM (slots), Storage (SSD/HDD/NVMe), and Displays (Resolution/Hz).
- Native Anonymization: Automatically filters out serial numbers, UUIDs, and hostnames before sending data to the LLM.
- Performance Monitor: Real-time analysis of CPU load, memory usage, and identification of the top 5 most resource-heavy processes.
- Compatibility Checker: Test if the machine can run a specific software/game via the remote API.
Commands
npm install # Install dependencies
npm run build # Compile TypeScript
npm run dev # Fast launch via tsx
npm run inspector # Launch the interactive MCP inspector
npm run lint # Type checkingExposed MCP Tools
analyze_local_system: Returns a complete, anonymized hardware inventory.analyze_performance: Returns current system load and active processes.check_software_compatibility: Verifies compatibility with a given software (e.g., "Cyberpunk 2077").get_hardware_recommendation: Analyzes the system and suggests upgrades via the remote API.
Configuration
REMOTE_API_URL: Remote API URL (default:http://localhost:3000).
