get_cpu_usageLABSCO SUMMARY
The Activity Monitor numbers, available to ask about.
Each tool takes no input and returns one line, so 'why is this slow' gets answered with figures — load per core, free memory, disk headroom, and a real speed test rather than a guess about the connection.
What it is
Six read-only system metrics gathered with systeminformation, plus an internet speed test taken across multiple sources.
What you get
- CPU load overall and per core, as percentages
- Memory total, used and free in GB, with the percentage used
- Disk usage for the largest drive, in GB and percent
- Network RX and TX rates in KB/s, plus totals since boot in MB
- Battery charge, charging state and minutes remaining — or a plain answer when there is no battery
- Download and upload speed in Mbps, reported as medians across several sources
Requirements
Node, with the project cloned and built — the build writes build/index.js, which is what the configuration runs, even though the README says dist/index.js. It reports on the machine it runs on, so run it where you want the readings from. No key, no account.
Setup effort
Build from source — git clone https://github.com/abhinav7895/system-mcp.git
USE CASES
Working out why the machine has slowed down
Per-core load and free memory together separate a pegged CPU from a machine that is swapping.
Checking whether a build will fit on disk
Usage on the largest drive, in GB as well as percent, so the headroom is a number you can compare against.
Diagnosing a connection that feels slow
A measured download and upload figure alongside the live RX/TX rate, rather than an impression.
Deciding whether to go and find the charger
Charge percentage, whether it is charging, and the minutes remaining.
Seeing what a long job is doing to the machine
Sample CPU and network while it runs to tell a busy process from a stuck one.
6 tools
Listed in project docs. Your client may show a different list.
get_memory_usageget_disk_spaceget_network_usageget_battery_statusget_internet_speedBEFORE YOU INSTALL
What to know first
Tools you get
6 tools named in the project's documentation.
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
You clone and build this one yourself.
Server setup
git clone https://github.com/abhinav7895/system-mcp.git
cd system-mcp
npm install
npm run buildMCP Clients
Claude Desktop
{
"mcpServers": {
"system-resource-monitor": {
"command": "node",
"args": [
"<YOUR_PATH_TO_SYSTEM_MCP>/build/index.js"
]
}
}
}