AI visibility · 15 points
Training-crawler consent
One check, stated three ways: what we measure, what would make it wrong, and how to confirm it yourself.
What this measures
The same parse, for GPTBot, ClaudeBot and Google-Extended — the crawlers that collect material for future models.
Where this check can be wrong
Nothing, deliberately: this check is descriptive and always earns full credit, because declining to be training data is a normal decision. It can never appear in your fix list.
Verify it without taking our word for it
curl -s https://your.site/robots.txt and look for these three names. Decide once, on purpose, and write the reasoning down.