watchdog

name	watchdog
description	Monitor system health — check services, disk usage, processes, logs. Use when asked about system status or health.
category	system
maturity	stable
tags	["systemd","disk-usage","process-monitoring","journalctl","health-check"]

System Health Watchdog

Quick Health Check

# Disk usage
df -h / /home | tail -2

# Memory
free -h

# CPU load
uptime

# Running services
systemctl --user list-units --state=running --no-pager | head -20

# Check specific services
systemctl --user status openclaw-poe-gateway openclaw-argus-gateway 2>&1 | grep -E "●|Active:"

# Large files
dust -n 10 $HOME

# Network connectivity
curl -s -o /dev/null -w "%{http_code}" https://api.telegram.org

Process Monitoring

# Check if a process is running
pgrep -fa "PROCESS_NAME"

# Top CPU consumers
ps aux --sort=-%cpu | head -10

# Top memory consumers
ps aux --sort=-%mem | head -10

Log Checking

# Recent system errors
journalctl --user --since "1 hour ago" --priority err --no-pager | tail -20

# Check specific log
tail -50 /path/to/logfile | grep -i "error\|fail\|crash"

Key Services to Monitor

• openclaw main gateway (systemd)
• openclaw-poe-gateway (systemd user)
• openclaw-argus-gateway (systemd user)
• Claude Code session (this instance)