How the gate works
This tool grades a single product listing across 7 machine-readability and trust signals AI shopping agents check before recommending it. Each signal is scored 0–1; the verdict is the weakest one — then a gate that can only make it worse.
The 7 signals
Identifier completeness at 0, or feed-vs-page consistency at 1-or-below, forces UNREADABLE even when every other signal reads clean - an agent distrusts a self-contradicting or unidentifiable item rather than just finding it thin.
Why: Six of seven signals read fully clean. But the feed and the live product page disagree - feed-vs-page consistency marks 1, tripping its gate threshold. On the raw minimum this would read GAPS, but the contradiction gate overrides it: an agent that catches a feed lying about price or stock distrusts the whole listing, not just this field.
A clean sheet didn’t save it. The verdict is the weakest signal — and a top trigger is dispositive on its own — so it names the one thing to fix first. Grades feed and listing readiness for AI shopping agents from marks you set - not a guarantee of placement, ranking, or purchase by any specific agent or platform.