The free health check

How the API key is handled

• The key is used for this one scan and then discarded.
• It is not written to logs and not stored, unless you later connect the instance to an account.
• The field is a password field, so the key is not visible on screen or kept by the browser.

If you are only evaluating OutcomeGuard, delete the key in n8n after the scan. It takes a moment and it is good practice with any third party.

What it reads

• Workflows: names, active flags, trigger schedules, node names and a definition hash.
• Recent executions within the scan window: status, timestamps, duration, item counts, the last node executed, and error text with secrets redacted.

Execution payloads are not downloaded. Nothing is written to your instance.

The shareable report

Each scan produces a private link to a read-only copy of the report, which is useful for sending to a colleague or a client. The link is unguessable, it is not indexed by search engines, and the report is deleted after 7 days.

Anyone with the link can open it, so treat it as you would any shared document. It contains workflow names and summary evidence, not your data.

Limits

• The instance must be reachable over HTTPS from the internet. Private addresses are rejected rather than probed.
• There is a limit on scans per instance per day, and a rate limit per network, to keep the scanner from becoming a nuisance to the instances it reads.
• The scan looks at a bounded window of recent history, so a workflow with very little history is reported as insufficient data.

Scan versus monitoring

The scan

One snapshot, judged against whatever history was available at that moment. Good for finding what is already broken.

Monitoring

A baseline that stays current, every new execution checked, missed runs detected at the time they should have happened, Outcome Contracts verified after each deadline, and an email when something breaks.

A scan tells you about the past fortnight. Monitoring is what tells you about this morning.