Visible evidence
Every claim used by the demonstration is supported by human-readable page content or a declared GeoVector identity fact.
Technical demonstration
The GeoVector AI Visibility Lab is a small public origin used to demonstrate an evidence-first structured-data workflow: inspect a page, approve an annotation, deliver it at the edge, and independently verify the response.
The demonstration
Good structured data should describe what a person can already find on a page. This origin separates the page itself, the approved annotation, and the delivery mechanism so each can be examined independently.
Every claim used by the demonstration is supported by human-readable page content or a declared GeoVector identity fact.
Structured data is treated as a versioned, reviewable artifact—not an on-request model output or an undisclosed rewrite.
A crawler can inspect the delivered HTML and confirm the exact annotation version, just as a search system would retrieve it.
Why this is useful
Technical markup can improve machine readability, but it must not invent qualifications, offers, prices, reviews, or relationships. The useful question is whether a narrowly scoped annotation makes an already accurate page easier to interpret.
Read the implementation principles