# System Monitor MCP Server — Real-Time Infrastructure Monitoring for AI Agents

**Give your AI agent real-time access to system metrics, process management, and container orchestration.**

## 🚀 Quick Install

### Claude Desktop
```json
{
  "mcpServers": {
    "system-monitor": {
      "command": "python",
      "args": ["-m", "server"],
      "env": {
        "PYTHONPATH": "${workspaceFolder}/src"
      }
    }
  }
}
```

### VS Code (Continue)
```json
{
  "experimental": {
    "mcpServers": {
      "system-monitor": {
        "command": "python",
        "args": ["server.py"]
      }
    }
  }
}
```

### Docker
```bash
docker build -t system-monitor-mcp .
docker run -v /var/run/docker.sock:/var/run/docker.sock system-monitor-mcp
```

## 🛠️ Tools

| Tool | Description | Returns |
|------|-------------|---------|
| `get_system_load()` | CPU, memory, swap usage | CPU%, Memory%, Swap%, Load avg |
| `get_process_list()` | Running processes sorted by CPU/memory | PID, name, CPU%, memory%, status |
| `get_disk_usage()` | Disk usage per mount point | Mount, total, used, available, usage% |
| `get_docker_status()` | Docker container status | Container name, status, uptime, ports, image |
| `tail_log(path, lines)` | Tail log files | Last N lines with timestamps |

## 💡 Use Cases

- **Self-monitoring** — AI agents that monitor their own infrastructure
- **DevOps automation** — Deploy, restart, and manage services autonomously
- **Incident response** — Detect and diagnose issues in real-time
- **Resource optimization** — Analyze usage patterns and suggest cost savings

## 📋 Requirements

- Python 3.10+
- Linux (most features) / macOS (partial) / Windows (limited)
- Docker (optional — for container monitoring)
- psutil, mcp, docker-py

## 🔐 Security

- **Read-only by design** — no kill/rm/create operations exposed
- **No root required** — runs with user-level permissions
- **Sandboxed log access** — path traversal protection built in

---

*Part of the Hermes Agent ecosystem*