Yamaru Hardware Probe
About
Yamaru Hardware Probe gives Claude Desktop and other MCP clients real visibility into your machine. 10 expert tools: full hardware inventory, real-time performance & thermal monitoring, RAM pressure analysis, SMART storage health, antivirus impact diagnostics, and LLM compatibili
Details
- Author
- yamaru-eu
- Downloads
- 348
- Categories
- AI, Infrastructure
Jump to
- Deep hardware inventory — CPU, RAM, GPU, motherboard, OS
- Real-time performance monitoring with top processes
- Thermal and power diagnostics with throttling detection
- SMART storage health per-disk temperature and wear
- Antivirus and EDR impact detection on dev paths
- LLM compatibility prediction for quantization fit (BETA)
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
Yamaru Hardware ProbeCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Add the server configuration to your claude_desktop_config.json with the command npx -y @yamaru-eu/hardware-probe and restart Claude Desktop. The server then exposes 10 diagnostic tools queryable through any MCP client.
analyze_local_system
Reads the hardware specifications of the local machine: CPU, RAM, motherboard and OS. Returns a structured JSON object.
analyze_performance
Provides real-time system performance: CPU load, memory usage, and top processes.
analyze_inference_config
Performs a deep scan for LLM inference optimization: GPU details, real memory bandwidth benchmark, ML runtimes (Ollama, Docker, WSL), and environment variables.
get_inference_expert_knowledge
Returns the expert instructions and rules for optimizing LLM inference on this specific hardware. Call this to learn HOW to interpret analyze_inference_config results.
check_llm_compatibility
(BETA) Checks if a specific LLM model can run on this machine. Returns optimal quantization and estimated tokens per second. Requires remote API connection.
get_llm_recommendations
(BETA) Recommends the best LLM models that can run locally on this machine. Requires remote API connection.
analyze_ram_pressure
Reports current memory pressure: total/used/free/available memory, swap usage, and top processes by RSS.
check_storage_health
Reports per-disk health: type (NVMe/SSD/HDD), vendor, temperature, SMART status, and firmware.
thermal_profile
Reports current CPU and GPU thermal and frequency state: temperature, utilization, and fan speed.
diagnose_antivirus_impact
Detects running antivirus/EDR products and reads their exclusion rules. Checks dev hot paths coverage.
monitor_system_health
Monitors system metrics (CPU load, RAM usage, CPU temperature) over a configurable duration (up to 10 minutes) and returns min/max/avg statistics. Use duration_seconds to set the observation window and interval_seconds to control sampling granularity.
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"yamaru hardware probe": {
"yamaru-probe": {
"command": "npx",
"args": [
"-y",
"@yamaru-eu/hardware-probe"
]
}
}
}
}McpServers
{
"yamaru-probe": {
"command": "npx",
"args": [
"-y",
"@yamaru-eu/hardware-probe"
]
}
}Yamaru Hardware Probe
Give Claude (and any MCP client) real eyes on your machine. Yamaru Hardware Probe is an open-source MCP server that exposes 10 expert diagnostic tools to your LLM client — so you finally get grounded answers instead of "it depends on your configuration".Why this exists
LLMs are great at reasoning, but blind to the box they run on. Ask Claude "can my PC run Llama 3 70B in Q4_K_M?" and you'll get a generic answer. With this MCP server, Claude reads your actual CPU, RAM, VRAM, thermals, and storage health — then gives you a real, chiffré answer.Features
- 🔍 Deep Hardware Inventory — CPU, RAM slots, GPU/VRAM, motherboard, OS topology - ⚡ Real-time Performance — live CPU/memory load + top processes by RSS - 🧊 Thermal & Power Diagnostics — detects throttling, fan RPM, frequency clipping - 💾 SMART Storage Health — per-disk temperature, wear level, I/O bottlenecks - 🛡️ Antivirus / EDR Impact — detects exclusion gaps on dev hot paths - 🤖 LLM Compatibility (BETA) — predicts quantization fit and VRAM requirements - 📊 Time-bound monitoring — min/max/avg health reports over a duration - 🔒 Privacy-first — automatic anonymization of unique hardware IDs before any remote callQuick Install
Add this to yourclaude_desktop_config.json:
``json
{
"mcpServers": {
"yamaru-probe": {
"command": "npx",
"args": ["-y", "@yamaru-eu/hardware-probe"]
}
}
}
`
Restart Claude Desktop. Done.
Available Tools
| Tool | Purpose |
|---|---|
| analyze_local_system | Full CPU/RAM/GPU/MB/OS inventory |
| analyze_performance | Real-time load + top processes |
| analyze_ram_pressure | Memory + swap + top 10 by RSS |
| check_storage_health | SMART status, temperature, type |
| thermal_profile | CPU/GPU temps, fans, throttling |
| diagnose_antivirus_impact | AV/EDR conflicts on dev paths |
| monitor_system_health | Min/max/avg over a duration |
| check_llm_compatibility | VRAM fit prediction (BETA) |
| get_llm_recommendations | Best local models for your rig (BETA) |
| analyze_inference_config | Ollama/CUDA/Metal runtime tuning |
Example Prompts
- "Why is my PC slow during local inference? Diagnose."
- "Can I run Llama 3 70B Q4_K_M on this machine?"
- "Check if my antivirus is scanning my models folder."
- "Monitor my system health for 60 seconds and report anomalies."
Stack & License
TypeScript · MCP SDK · systeminformation · Apache 2.0
Built by the Yamaru project.
- 📦 npm: @yamaru-eu/hardware-probe`
- 🐙 GitHub: yamaru-eu/hardware-probe
- 🐛 Issues & feedback welcomeSign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
