R Runwake

Runwake tells you when a scheduled job goes quiet.

Use cases for Runwake

Runwake watches the background jobs you only notice when they stop: cron scripts, backups, scheduled workflows, scrapers, reports, and maintenance tasks.

Start with one job See how it works

The pattern

Most monitoring checks whether something is up right now. Runwake checks whether something checked in when it was supposed to.

Your job does the work. At the end, it calls its Runwake ping URL. If Runwake does not hear from it on schedule, you get an email.

curl -fsS https://app.runwake.dev/ping/YOUR_MONITOR_TOKEN
MonCheck-in
TueCheck-in
WedCheck-in
ThuMissed, alert sent

Use cases

Runwake is for the job you assume is still running.

Cron

Cron jobs

What goes quietScheduled scripts can stop running, exit early, or never reach the final step.

Check-in patternAdd a ping at the end. If the check-in is late, Runwake emails you.

0 2 * * * /srv/jobs/nightly-task.sh && curl -fsS https://app.runwake.dev/ping/YOUR_MONITOR_TOKEN
Backups

Nightly backups

What goes quietA backup that fails loudly is easy to notice. A backup that never ran is harder.

Check-in patternPing Runwake after the backup command finishes. Runwake verifies the backup job checked in, not that the backup contents are restorable.

backup-command && curl -fsS https://app.runwake.dev/ping/YOUR_MONITOR_TOKEN
GitHub Actions

GitHub Actions schedules

What goes quietScheduled workflows can fail, be delayed, be disabled, or stop reaching the part of the workflow you care about.

Check-in patternAdd a ping step at the end of the workflow. Runwake confirms the scheduled workflow reached that point.

- name: Ping Runwake
  run: curl -fsS https://app.runwake.dev/ping/YOUR_MONITOR_TOKEN
Scrapers

Scrapers and data jobs

What goes quietData can go stale quietly. A scraper may stop collecting data long before anyone notices.

Check-in patternPing Runwake after a successful scrape or import cycle. If the job goes quiet, you know the dataset may be stale.

node scraper.js && curl -fsS https://app.runwake.dev/ping/YOUR_MONITOR_TOKEN
Reports

Scheduled reports

What goes quietDaily summaries, exports, and customer reports can fail silently.

Check-in patternPing Runwake after the report is generated or sent. If the report job misses its expected check-in, you get an alert.

python send_daily_report.py && curl -fsS https://app.runwake.dev/ping/YOUR_MONITOR_TOKEN
Maintenance

Server maintenance scripts

What goes quietCleanup jobs, cache warmers, rotation scripts, and certificate helpers often run quietly in the background.

Check-in patternAdd a ping after the maintenance task completes. Runwake tells you when the routine stops checking in.

/usr/local/bin/cleanup-temp-files && curl -fsS https://app.runwake.dev/ping/YOUR_MONITOR_TOKEN
Automations

Small business automations

What goes quietOwner-operator scripts, no-code scheduled jobs, exports, imports, and sync tasks can fail without a dedicated ops team watching.

Check-in patternUse Runwake as a simple "I ran" signal for background automations you would rather not discover late.

After the export, import, or sync finishes, have the automation call its Runwake ping URL.

Good fit

When Runwake fits

  • A scheduled thing should finish on a predictable rhythm.
  • It can make an outbound HTTPS request.
  • You want an email when the check-in is late.
  • You want setup to be one curl command, not a full monitoring stack.
  • You care about absence: the job did not say "I'm done."

Not the right tool

When Runwake is not the right tool

  • You need logs, traces, metrics, dashboards, or full observability.
  • You need Slack, PagerDuty, on-call escalation, or incident management today.
  • You need Runwake to store secrets.
  • You need proof that a backup is restorable, not just that the backup job ran.
  • You need internal network checks from inside a private LAN unless the job can make outbound HTTPS requests.
  • You need high-volume event ingestion.

Start small

Start with one important job.

Pick one script, backup, workflow, or report you would hate to discover failed three days late. Add a Runwake ping at the end.

Start with one job