It didn't crash. It just stopped.
Flatline watches every job you run unattended and pages you the second one goes quiet — before your customer notices, not after.
5 monitors free · no credit card · first alert in under 2 minutes
Three moves, then it runs itself.
No agent to install, no dashboard to babysit. One request, one rule, one alert path.
Add one line to your job.
It calls a unique URL when it runs — from cron, a script, GitHub Actions, or a Zapier/Make/n8n step.
Tell us the schedule.
Set how often it should check in, plus a grace period for the ordinary jitter of real infrastructure.
Miss it, and we page you.
Email, Slack, Discord, SMS, or a phone call — your choice. The next good ping auto-resolves it.
Everything a quiet failure needs to become a loud one.
Built for the shape of real automation, not just a single cron line.
Heartbeat monitors
A unique ping URL per job, live the moment you create it. No setup wizard.
Start / success / fail
Catch jobs that run too long, finish too fast, or exit with an error — not just ones that vanish.
Agent-aware pings
A heartbeat built for long-running AI agents, so a stalled reasoning loop reads differently than a crash.
Status pages
A clean page per project to show a client or teammate that everything's actually fine.
Weekly digest
One email a week showing uptime across every monitor — proof it's working, even in quiet weeks.
API & CLI
Provision monitors programmatically when your jobs are created dynamically, not by hand.
Your AI agent doesn't crash. It just goes quiet mid-thought.
Classic cron monitors know two states: ran, or didn't. A multi-step agent can be technically alive and still stuck — no error, no exit code, no cron entry to check. Flatline gives agents a third state.
- A "still working" heartbeat an agent sends between steps
- A stalled-run alert, distinct from never-started or crashed
- Same ping URL pattern — one line inside your agent loop
Free while you're integrating. Paid once we're load-bearing.
Every plan includes unlimited alert history search and auto-resolve. Annual billing saves two months on any paid plan.
Starter
- 25 monitors
- + Slack, Discord, webhooks
- 90-day history
- 1 status page
Pro
- 100 monitors
- SMS · 20 credits incl.
- API access, 3 seats
- Agent-aware pings
Business
- Unlimited monitors
- Phone-call escalation
- Unlimited seats
- White-label status pages
Before you ask support.
What exactly counts as a "ping"?
Any HTTP request to your monitor's unique URL — a plain GET from cron/curl, a webhook step in Zapier, Make, or n8n, or a call from inside your own code. No SDK required, though we ship small helpers for popular languages.
What if my job's schedule varies?
Set a grace period instead of a rigid window — most teams add 25–50% on top of the expected interval to absorb normal jitter without missing a real failure.
Do you monitor yourselves?
Yes — with a different, independent monitoring service, on principle. If Flatline goes down, we don't want to be the only ones who'd know.
Can I move off the free tier later without losing history?
Yes. Upgrading and downgrading only changes limits going forward — nothing is deleted, and history simply extends as your plan allows.
Stop finding out from a customer.
Five monitors, free, running in under two minutes.
No credit card. We'll email you one link to confirm before anything goes live.