Something the parser had to say aloud about the file itself: it was read as cp1251 rather than UTF-8, a few bytes were not UTF-8, the front matter was opened and never closed, a canonical could not be parsed as an address. None of these is a ranking problem by itself, and each is a reason the other findings on the page may be wrong: a page read in the wrong encoding has the wrong words, the wrong title and the wrong duplicates. That is why it is a warning and why it comes first on the page's work order — fix the file, and the rest of the report becomes trustworthy.
The finding reads:
warning source-note what the parser has to say aloud: the file was read as cp1251, the front matter was not recognised, a canonical cannot be parsedFix the file itself — its encoding or its front matter. While it is read wrongly, every other finding on it is unreliable.
Collapsed when it fires template-wide. If it appears on 90% or more of the pages — or of one language version — of a site with at least 20 pages, it is a property of the template and is reported once, with the count, instead of once per page.
pipx install indexgap
cd your-project
indexgap init
indexgap checkThen indexgap brief turns
every finding on this page into a work order in indexgap-briefs/,
with the thresholds the result has to satisfy and the command to check it.