At a glance
- Why a green run can still be a failuren8n marks a run as successful when no step throws an error.
- The four checks that catch themEmpty runs: Pick the step that matters, such as the one that creates the invoice or upserts the order, and check how many items it put out.
- How the monitor worksThe free template below runs every 15 minutes.
- What you need to set it upStart with the two or three workflows that would hurt most if they quietly stopped: orders, invoices and new leads are the usual ones.
- Make the workflows easier to watchGive the step that does the real work a clear name, so the monitor and your team can find it
Why a green run can still be a failure
n8n marks a run as successful when no step throws an error. It does not know what the run was supposed to achieve. These all finish green:
- A filter or IF step that sends every item down the wrong branch, so the step that saves records never runs
- An API that answers with an empty list because a token lost access or a date range is wrong
- A search that matches nothing after someone renames a column or a field
- A schedule trigger that stops firing because the workflow was switched off or the instance restarted at the wrong moment
An error workflow will not catch any of these, because there is no error to catch.
The four checks that catch them
- Empty runs: Pick the step that matters, such as the one that creates the invoice or upserts the order, and check how many items it put out. Fewer than you expect means the run did nothing useful.
- Missed schedules: For anything on a timer, set the longest gap you will accept between runs. A nightly job that has not run for 25 hours has stopped, whatever its last status says.
- Missing records: Check the destination, not just the workflow. If your CRM, sheet or database has an API that can return today's records, count them.
- Switched-off workflows: An inactive workflow fires no triggers and produces no errors. Check that the ones you rely on are still active.
How the monitor works
The free template below runs every 15 minutes. For each workflow on your list it reads the workflow and its recent runs from the n8n API, counts the items at the step you named, measures the gap since the last run, and calls your record-check URL if you gave one.
Each problem is reported once, to Slack and by email, with a link straight to the run. Every finding is also added to a Google Sheet, so you can see patterns over a week. At 8am it posts a short summary, which also proves the monitor itself is still running. If it cannot reach the n8n API, Slack or email, its own failure alert says so.
What you need to set it up
- n8n with the public API turned on, and an API key from Settings
- A Slack incoming webhook, an email account for alerts, and a Google Sheet
- The id of each workflow you want to watch, from its URL, and the name of the step that should never come out empty
Start with the two or three workflows that would hurt most if they quietly stopped: orders, invoices and new leads are the usual ones.
Make the workflows easier to watch
- Give the step that does the real work a clear name, so the monitor and your team can find it
- Turn on saving successful runs in each workflow's settings, or the monitor has no run data to count
- Keep run history for at least a day, longer than the monitor's lookback window
- Add an error workflow as well, for the failures that do turn red
Updated 2026-10-27. General guidance, not advice for your situation; we scope each project in writing before any work starts.
Download the template
An n8n workflow that watches the workflows you list and alerts you on Slack and by email when a run succeeds but does nothing, a schedule is missed, an expected record is missing, or a workflow is switched off, logging every finding to a Google Sheet. Free, with setup steps. You add your own accounts and settings.
Questions people ask first
Why doesn't my error workflow catch these?
An error workflow only runs when a step throws an error. A run that filters out every item or gets an empty answer from an API ends without an error, so the error workflow never starts.
Does this work on n8n Cloud?
Yes, as long as the n8n API is available on your plan. The monitor only needs an API key and the address of your instance.
Will it alert me every 15 minutes about the same problem?
No. Each problem has a key, such as the run id or the date, and is reported once. The daily summary counts everything found in the last 24 hours.
What if the monitor itself stops?
The 8am summary arrives every day. If it does not, the monitor is the thing that stopped.
Tell us what you need built.
Tell us the process and the tools involved, or call. We reply with questions or a written scope.
hello@pushbuild.com