| name | vps-monitor |
| description | Check remote server health over SSH and surface actionable operational risks. |
VPS Monitor
Use this skill when inspecting a VPS, cloud VM, or remote Linux host over SSH.
- Confirm host, user, authentication method, and whether elevated access is available.
- Start with low-risk read-only checks before considering restarts or config changes.
- Check uptime, load average, CPU saturation, memory pressure, disk usage, and inode exhaustion.
- Review the status of critical services with
systemctl, docker ps, or the host's process supervisor.
- Verify the application's own health endpoint, not just whether the process exists.
- Inspect recent logs for crash loops, OOM kills, certificate issues, and upstream dependency failures.
- Check network reachability, bound ports, and firewall symptoms when services appear healthy but unavailable.
- Confirm backup, cron, or scheduled jobs are still running if they are part of the host's role.
- Compare findings to the last known good state if prior metrics or notes exist.
Useful commands usually include:
uptime
df -h
free -m
systemctl --failed
docker ps
journalctl -u <service> --since "1 hour ago"
curl -fsS http://127.0.0.1:<port>/health
Summarize results as critical, warning, and informational items.
Do not reboot, restart services, or edit config unless the user asks for intervention instead of monitoring.