🦀 Rusty Sentinel MCP

A high-performance Model Context Protocol (MCP) server written in Rust for real-time system monitoring.

This agent allows LLMs (like Claude) to inspect the host machine's resources, memory usage, and running processes efficiently, with zero runtime overhead.

🚀 Why Rust?

• Zero-cost abstractions: Monitor your system without affecting host performance or eating up RAM (unlike Python/Node-based agents).
• Memory Safety: Rust's borrow checker ensures memory-safe interactions with low-level system resources.
• Single Binary: Compiles to a standalone executable. No need to install Python, Node.js, uv, or manage complex virtual environments on the target machine.

🛠️ Included Tools

The server exposes the following tools to the MCP Client:

1. get_system_stats

Provides an instant snapshot of the system's overall health.

• Inputs: None.
• Returns:
  • CPU usage percentage (global).
  • Memory usage (Used/Total RAM and Swap).
  • OS Name and Kernel Version.
• Example Prompt: "Check if the server is under heavy load right now."

2. find_heavy_processes

Identifies the specific processes consuming the most resources.

• Inputs:
  • limit (optional, integer): Number of processes to list (default: 5).
• Returns: A sorted list of processes with their PIDs and CPU usage.
• Example Prompt: "What is slowing down my computer? Find the top 3 heaviest processes."

📦 Getting Started

1. Prerequisites

You need Rust and Cargo installed to compile the project.

• MacOS/Linux:

  curl --proto '=https' --tlsv1.2 -sSf [https://sh.rustup.rs](https://sh.rustup.rs) | sh
  

• Windows: Download and run rustup-init.exe from rustup.rs.

2. Compilation

Clone this repository and build the project in release mode (optimized for speed):

cargo build --release

Where is my binary? After compiling, the executable will be located at:

• Windows: target\release\mcp-server-rust-sentinel.exe
• MacOS/Linux: target/release/mcp-server-rust-sentinel

🧪 Testing (MCP Inspector)

You can test the server tools visually in your browser without connecting to Claude yet.

Requirements:

• Node.js installed (to run npx).

Run the Inspector: Point the inspector to your compiled binary:

npx @modelcontextprotocol/inspector target/release/mcp-server-rust-sentinel

(Windows users: Remember to add .exe at the end)

This will open a dashboard at http://localhost:5173 where you can manually click "Run" on the tools and see the JSON results.

🔌 Configuration (Claude Desktop)

To use this server with Claude Desktop, you need to add it to your configuration file.

1. Locate the Config File

• MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

2. Add the Server

Add the rust-sentinel entry to the mcpServers object.

⚠️ IMPORTANT: You must use the ABSOLUTE PATH to the binary you just built.

{
  "mcpServers": {
    "rust-sentinel": {
      "command": "/Users/YOUR_USER/path/to/repo/target/release/mcp-server-rust-sentinel",
      "args": []
    }
  }
}

(Windows users: Remember to escape backslashes, e.g., C:\\Users\\Name\\...)

🏃 Usage

1. Restart Claude Desktop completely.
2. Look for the 🔌 (plug icon) in the top right corner. It should have a green dot indicating "Connected".
3. Start chatting! You can now ask questions about your system.

Example Conversation:

You: "How is my system performing?" Claude: (Calls get_system_stats) "Your CPU is at 15% and you are using 8GB of 16GB RAM. Everything looks normal." You: "Actually, it feels sluggish. Who is using the CPU?" Claude: (Calls find_heavy_processes with limit=3) "I found these top consumers: 1.

... View full README on GitHub