What “cannot verify” means
Error means the check could not verify the scenario. It is not a pass, and it is not a verdict that your app is broken.
The run report badge says Error. The homepage and the reliability page call that ending "cannot verify." The CLI JSON status is error. Those are the same ending. The badge does not say "cannot verify."
An error is not a pass. It is also not a product verdict. The check did not establish that the scenario happened, and it did not establish that your app broke the scenario.
Why a run ends there
The check has to observe what you wrote. If the evidence is missing, the run's provenance does not match, or the platform fails (a time limit, the model provider, the harness), the run ends as an error. A local tunnel that never connects is the same kind of ending: the check did not start.
Read the cause on the report before you treat it as a bug:
- Monito's error — not charged. The failure is on Monito's side.
- Check didn’t start — not charged. Nothing ran.
- Site unreachable or access blocked. The browser got no answer, or the site blocked the run.
- Test needs updating. The scenario no longer matches the product.
- Cause unknown — see the evidence. The report is not willing to name a cause.
- App bug is not this ending. That label is for a Failed run whose verdict was verified. An unverified failure is labeled Check failed — see the evidence instead, so an unverified verdict is not presented as an app bug.
What you do with it
Open the report. If the cause is yours (the site was down, the sentence is stale), fix that and run again. If the cause is Monito's, you do not pay for that run, and a retry is a new check. Do not read Error as "ship it."