Skip to main content
Glama
dragogargo

mcp-sysmon

by dragogargo
README.md
# mcp-sysmon

[![CI](https://github.com/dragogargo/mcp-sysmon/actions/workflows/ci.yml/badge.svg)](https://github.com/dragogargo/mcp-sysmon/actions/workflows/ci.yml)
[![PyPI](https://img.shields.io/pypi/v/mcp-sysmon)](https://pypi.org/project/mcp-sysmon/)
[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)

MCP server for system monitoring. Ask Claude about your system — CPU, memory, disk, network, and processes.

## Tools

| Tool | Description |
|------|-------------|
| `get_system_overview` | CPU, memory, swap, disk, uptime — full snapshot |
| `get_system_health` | Quick health check — only reports problems |
| `get_top_processes` | Top processes sorted by CPU or memory |
| `get_disk_usage` | Disk partitions with I/O stats |
| `get_network_info` | Network interfaces, IPs, speeds, traffic |
| `get_open_ports` | Listening TCP/UDP ports with owning process |
| `get_battery_status` | Battery level, power source, time remaining |
| `find_process` | Search processes by name |
| `kill_process` | Terminate a process by PID (SIGTERM or SIGKILL) |

## Example prompts

- "Why is my laptop slow right now?"
- "What's using the most memory?"
- "Is anything wrong with my system?"
- "What ports are open?"
- "How much battery do I have?"
- "Find all Chrome processes and kill the biggest one"

## Install

```bash
pip install mcp-sysmon
```

## Usage with Claude Desktop

Add to your `claude_desktop_config.json`:

```json
{
  "mcpServers": {
    "sysmon": {
      "command": "mcp-sysmon"
    }
  }
}
```

## Usage with Claude Code

```bash
claude mcp add sysmon -- mcp-sysmon
```

## Development

```bash
git clone https://github.com/dragogargo/mcp-sysmon.git
cd mcp-sysmon
uv sync
uv run pytest tests/ -v
uv run mcp-sysmon
```

## License

MIT

TDQS

A4.8/5.0

Scored across 9 tools

Disambiguation5/5

Each tool targets a distinct aspect of system monitoring with clear descriptions. Tools like get_system_overview and get_system_health are differentiated by full metrics vs. warnings-only, and process tools (find_process, get_top_processes, kill_process) have non-overlapping purposes.

Naming Consistency5/5

All tools use a consistent verb_noun pattern: get_ for read-only, find_ for search, kill_ for destructive operations. Naming is predictable and clear, e.g., get_disk_usage, get_network_info, kill_process.

Tool Count5/5

With 9 tools covering CPU, memory, disk, network, ports, battery, and processes, the count is well-scoped for a system monitoring server. No redundancy or missing core functionality.

Completeness5/5

The tool set provides thorough coverage for system diagnostics: process lifecycle (search, top, kill), resource overviews (system overview, health check), and specific subsystems (disk, network, ports, battery). Gaps like hardware sensors are niche and beyond typical scope.

Maintenance

ActivityInactive
ResponsivenessNo issues