Show the code an AI assistant reads
{
"@context": "https://schema.org",
"@graph": [
{
"@type": "WebPage",
"@id": "https://validated.directory/b/safire-wellness",
"url": "https://validated.directory/b/safire-wellness",
"isBasedOn": "https://safirewellness.com/",
"dateModified": "2026-11-13T09:17:20.812Z",
"mainEntity": {
"@id": "https://validated.directory/b/safire-wellness#business"
}
},
{
"@type": "LocalBusiness",
"@id": "https://validated.directory/b/safire-wellness#business",
"name": "Safire Wellness",
"url": "https://safirewellness.com/",
"identifier": {
"@type": "PropertyValue",
"propertyID": "validated.directory",
"value": "safire-wellness"
}
}
]
}The code has two parts. The page: isBasedOn is the site we read, dateModified when we read it, the first date; once the record is confirmed, lastReviewed and reviewedBy say when and by whom, the second date and the badge. The business: what its own site states, and once confirmed, the summary its confirmer stood behind. The same dates sit in ai.json and llms.txt, and in every answer from POST /mcp, no key, as verified_at and confirmed_at; get_checks for the last ten.