Linux server monitoring with one Bash agent
An HTTP check tells you the site stopped answering. The server agent tells you the disk hit 94% an hour before.
QuickStatus.dev runs a small open source Bash script on each Linux server. It reports CPU, memory, disk, network and the processes you name. You get an alert when a rule holds for as long as you set, and when servers stop reporting.
One check for the whole fleet
Every server that runs the install command joins the same Server check. A web tier behind an autoscaler is one check, not one per instance, and it counts as one check toward your plan.
You set how many servers must be reporting and a grace period (default 5 minutes). A reboot or a rolling deploy stays quiet. A server that shuts down cleanly says so and never alerts by itself.
Alerts that start sensible
A new Server check comes with three rules:
- CPU above 90% on any server for 10 minutes
- Memory above 90% on any server for 10 minutes
- Disk above 90% on any server, right away
Add rules for swap, inodes, load or a process that must be running (for example nginx). Each rule can look at any server, the fleet average, or at least N servers. Choose Alert only to get a warning and keep the check up, or Mark down to take it down.
Alerts go to the same places as your uptime checks: email and Telegram on Free; Slack, Discord, PagerDuty, Pushover and webhooks on Pro. Link the check to a status page component like any other check.
Install in one line
Run as root on each server:
curl -fsSL https://quickstatus.dev/agent/install.sh | sudo bash -s -- --token qse_...
- Ubuntu, Debian, RHEL, Rocky, Alma, Amazon Linux, Alpine and openSUSE, on x86_64 and arm64
- Runs as its own unprivileged user from a systemd timer (cron if there is no systemd)
- No daemon and no open ports. It only sends; we never connect to your server
- Never reads logs, files or application data
- Read the token from AWS SSM, AWS Secrets Manager or Google Secret Manager for launch templates and Terraform
- MIT licensed: github.com/QuickStatusHQ/quickstatus-agent
Free and Pro
| Free | Pro ($15/mo) | |
|---|---|---|
| Servers per Server check | 1 | 25 |
| Report interval | 5 minutes | 1 minute |
| Metric history | 7 days | 30 days |
Server checks share the plan's check limit with HTTP, heartbeat, MCP, DNS and TCP checks (Free: 10, Pro: 25). See pricing
Let your coding agent set it up
Over MCP, create_check with kind "server" returns the install command, so Claude Code, Cursor or Codex with a shell on the box can finish the setup. Ask it “which server is out of disk?” and it reads the answer from the check results.
Watching scheduled jobs on the same servers? See cron job monitoring.
FAQ
Is the server agent free?
Yes. The free plan covers 1 server per Server check with 5-minute reports and 7 days of metrics, inside the free plan's 10 checks. No card.
Do I need to open a port or run a daemon?
No. The agent runs from a systemd timer, only sends reports out, and opens no ports.
What happens when a server crashes?
After three missed reports the server is marked Lost. If fewer servers than you require are reporting for longer than the grace period, the check goes down and you get one alert.
Does it work on macOS or Windows?
Not yet. The agent is Linux only. GPU metrics are planned.