OG / Twitter cards check
Social-card metadata: inspect declarations and image status
TL;DR
Review optional social-sharing declarations where link sharing is intended. Their presence does not prove any named AI provider renders or cites them.
What the current check observes
Sampled static metadata is checked for nonempty og:title, og:description and og:image using property or name attributes. Title differences are advisory. Up to three distinct resolved image URLs receive status checks. The current auditor does not separately validate Twitter-card fields despite the historical category name.
Context and prerequisites
No rendered preview, image decoding, crop, accessibility quality, consumer adoption or citation effect is measured. A 200 verifies only a bounded status response, not an image. Different share and document titles may be intentional.
How to review the observation
For pages meant to be shared, review accurate copy and actual image content in the intended consumer. Inspect confirmed 404/410 destinations before updating. Do not impose social-card metadata on every private or utility page.
Illustrative example, not a universal prescription
<meta property="og:title" content="Reference"> <meta property="og:image" content="https://example.com/reference.png">
These are declarations only. An omitted description is an optional metadata observation, not proof of a failed AI preview or lost citation.
Measurement and experimental legacy calculation
Public measurement separates Confirmed issue, Advisory, Unknown, Not applicable and Observed pass. An observed pass applies only to an explicit bounded rule; absence of an emitted failure is not a pass. Missing prerequisites do not earn repair credit.
Start 100. Missing title deducts min(30, 10 × pages); description and image each deduct min(25, 8 × pages). Any sampled image 404/410 deducts 20 once. Unknown image checks and title divergence do not deduct. Optional omissions remain legacy point deductions, not calibrated defects.
Legacy values are clamped to 0..100 and rounded. They are not readiness certification, risk indicators or predicted citation outcomes. Method changes do not rewrite historical records.
Full methodology and fixed weightsSources and further manual validation
The observations above describe the local implementation, not guarantees from these sources. Protocol and provider documents help interpret purpose and perform checks this audit does not implement.
What observations does this check return?
LLMScore audits eleven readiness categories. Free, no signup.
Run this check on your site →