A check couldn't measure
Failure code cant_judge
The lab, not your site: never counted against it
What it means
One of the lab's checks couldn't measure on this device: a screenshot it couldn't read, so it couldn't tell a blank page from a drawn one; a log it couldn't get; a status nobody could get. A check that can't measure never passes, so the result is inconclusive and isn't counted either way. A real failure elsewhere still counts.
What to do
Words on a canvas
- Run this flow on a Mac lab, or have its step look for an element (css=…): only a Mac reads words drawn on a canvas.
Otherwise
- Retry: a check that couldn't measure usually can the next time.
- Happens again on this device? Report it to PocketQA with the test id.
For your coding agent
Inconclusive: don't change code. Test again; if it repeats on one device, report_to_pocket_lab with the test id (its cant_judge says which check and why).