Your agents already run on schedules.
Know the moment one stops.
WatchCron is a dead-man's switch for scheduled AI agent jobs and cron workflows. Each run pings a heartbeat — the moment a run goes silent or misses its window, the right channel gets alerted.
One heartbeat URL per job. Alerts only when something breaks.
Everything you need to trust your schedules
Purpose-built liveness monitoring for AI agents, cron jobs, and any workflow that must keep beating on a schedule.
Heartbeat pings, three ways
Ping a unique URL via plain HTTP after every run, call the SDK one-liner inside your agent, or fire the MCP tool. One line of setup per job.
Dead-man's-switch alerts
If a scheduled run is late, silent, or failing, WatchCron notices the missing heartbeat — and alerts before the silence becomes an incident.
Multi-channel routing
Deliver alerts to email, Slack, Discord, or any webhook. Route each job to the channel that can actually act on it.
Cron-aware grace windows
WatchCron understands cron expressions and jittery schedules. Define a grace window so a slow run never pages you, but a dead one always does.
Failure + latency timeline
Every check is recorded: last seen, duration, streak, and failure history. See which jobs flake and which ones quietly died weeks ago.
Status API & public page
Query job health programmatically or publish a status page for your fleet. Your agents’ uptime, visible at a glance.
Live in five minutes
No agents to install, no dashboards to configure. One URL per job and you're covered.
Add a heartbeat
Create a job in WatchCron and ping its unique heartbeat URL after each run — from an agent hook, a cron line, GitHub Actions, or any HTTP client. Works with anything that can make a web request.
Define cadence + grace
Tell WatchCron the expected schedule (cron expression, interval, or custom) and a grace window. WatchCron learns what "on time" means for each job.
Get alerted on silence
The moment a run misses its window, WatchCron alerts email, Slack, Discord, or your webhook — with the job, the last seen beat, and the failure timeline.
Simple pricing
Start free for 14 days. Pay only when your fleet grows.
Starter
For indie builders with a handful of scheduled agents.
- 10 monitored jobs
- Heartbeat pings via HTTP, SDK, or MCP
- Email + Discord alerts
- 30-day check history
Team
For teams running agents across many cron schedules.
- 50 monitored jobs
- Cron-expression aware grace windows
- Slack + webhook routing
- Failure + latency timeline
- 90-day history, weekly digest
Business
For agent fleets where silence is a production incident.
- Unlimited monitored jobs
- Multi-channel escalation policies
- Status API + public status page
- 1-year history
- Priority support
Loved by agent fleets
Teams running scheduled agents use WatchCron to know their fleet is alive.
“WatchCron caught a dead Airflow agent three days before anyone noticed. The silence alert paid for itself in one incident.”
“We run 40+ scheduled agents for customer content. One heartbeat line each — and no more discovering failures in customer tickets.”
“Cron-aware grace windows mean we stopped paging on slow runs but still catch the truly dead ones. Exactly the signal we needed.”
Questions & answers
Everything else you might want to know before you start monitoring.