system-monitor - SKILL.md Agent Skill

name: system-monitor description: "Monitor system health: CPU, memory, disk, processes, and network. Alert on thresholds via any channel." version: 1.0.0 metadata: echo: tags: [Monitor, System, CPU, Memory, Disk, DevOps]

System Monitor

System health monitoring with configurable alert thresholds.

Quick Status

# macOS
top -l 1 | head -10
df -h | grep -v tmpfs

# Linux
free -h && uptime
df -h | grep -v tmpfs

Script (psutil-based)

pip install psutil

python3 scripts/system_check.py status       # full report
python3 scripts/system_check.py cpu           # CPU only
python3 scripts/system_check.py memory        # Memory only
python3 scripts/system_check.py disk          # Disk only
python3 scripts/system_check.py processes     # top consumers
python3 scripts/system_check.py alerts        # check thresholds

Output Example

System Status Report — 2026-06-13 14:30
CPU: 23.5% (8 cores)
Memory: 12.4/16.0 GB (77.5%)
Disk /: 180/500 GB (36.0%)
Load: 2.1 / 1.8 / 1.5
Uptime: 14 days, 3:22:10

Top Processes (CPU):
  1. python3 (pid 1234) — 15.2% CPU, 1.2 GB RAM
  2. node (pid 5678) — 8.1% CPU, 800 MB RAM

Alert Configuration

~/.echo-agent/monitor.yaml:

thresholds:
  cpu_percent: 90
  memory_percent: 85
  disk_percent: 90
  load_average_5m: 8.0
check_interval: "*/5 * * * *"
alert_channel: telegram

Integration

Schedule with cron channel (*/5 * * * *) to check every 5 minutes. When thresholds are exceeded, alert is sent through configured channel.

Monitoring Stack (Optional)

For persistent monitoring with dashboards, consider:

Prometheus + Grafana (heavy but powerful)
Glances (pip install glances) — terminal dashboard
psutil + SQLite for historical data