How Nokto monitors your workflows.
Every two minutes, Nokto queries your n8n and Make instances, compares what it sees with what was expected, and opens or closes incidents.
From sign-up to first alert in 10 minutes.
Four steps, guided inside the application.
Connect an instance
Create an API key in n8n or a token in Make and paste it. Nokto tests the connection and lists the workflows.
Arm monitoring
Run the rhythm calculation, then approve the suggested thresholds for the workflows that matter.
Choose your alerts
Email works straight away. Add Telegram or Pushover and send a test alert.
Add a client
Attach the instance to a client to get its monthly report. Skipped in company mode.
What happens every two minutes.
Four stages, always in the same order.
Collection through the API, no agent
Nokto reads the list of workflows and their latest executions. It keeps metadata only: status, dates, duration, execution count.
Evaluation by rules
For each monitored workflow: did the last execution fail? Is it deactivated? No success within the expected delay? Today's volume far below expected? Duration above the threshold?
Caution with doubtful data
If the API call fails or the history is incomplete, the workflow is “unknown” for that pass. An instance that stops responding altogether becomes an “unreachable” incident.
Alert, reminder, escalation
One alert on opening, a button to take ownership, a reminder if nobody reacts, then escalation. A notification when the workflow recovers.
Every workflow has its own rhythm.
An hourly sync and a monthly reminder are not monitored the same way. Nokto suggests the values, you stay in control.
- Maximum silence: beyond this delay without a success, the workflow is silent.
- Expected volume per day: a sharp drop opens a volume incident.
- Maximum duration: above it, the workflow is flagged as slow.
- Error budget: the number of failures tolerated over the month.
- Monitored or not: only ticked workflows count towards your quota.
Where you already are.
Four ways to be warned, to combine as you like.
Active from sign-up, no setup.
Telegram
A message with action buttons, in a private chat or a team group.
Push notification
Through Pushover, at high priority for critical incidents, even at night.
Live console
The “To do” screen updates in real time.
The technical questions.
How often does Nokto check my workflows?
Every two minutes. The actual delay before a silence alert depends on each workflow's setting: one hour for a frequent sync, several days for a monthly reminder.
How does Nokto know a workflow's normal rhythm?
It works it out from the execution history and suggests a silence delay and an expected volume, with a confidence level. Nothing is applied without your approval, and you can set everything by hand.
What happens if the n8n or Make API does not respond?
On a single pass, the workflow is marked “unknown” and no incident is opened or closed: a failed call is never treated as silence. If the instance stays unreachable, Nokto opens an “unreachable” incident.
Can I avoid alerts during maintenance?
Yes. You pause alerts for an instance for a set duration. Collection continues and notifications resume on their own.
Does Nokto work with self-hosted n8n?
Yes, with self-hosted n8n as well as n8n Cloud, as long as the instance's public API is reachable from the internet over HTTPS.
Your first monitored workflow in ten minutes.
Connect an n8n instance or a Make team: Nokto discovers your workflows, learns their rhythm and warns you at the first anomaly.
14 days free · no credit card · cancel anytime