Can an AI engine
parse your schema?
Paste your JSON-LD or HTML. Every schema validator checks syntax — this one checks citation-readiness: the headline, author, dates, answer pairs, offers, and sameAs links an engine needs to attribute and cite your content.
What an answer engine needs from your schema
Structured data is how an answer engine knows what your page is — an article, a product, a FAQ, an organisation. But knowing the type is only half the job. To actually cite you, the engine needs the fields that let it attribute, date, and reconcile the content: a headline to title the citation, an author to name the source, a publication date to assess freshness, and sameAs links to connect your entity to its profiles across the web.
Every schema validator on the web checks whether your JSON-LD parses. That is the floor. This tool checks the ceiling: does your schema give an engine everything it needs to lift your content into an answer and attribute it to you?
Per-type citation-readiness checks
The checker walks every node in your JSON-LD — including nested @graph entries — and runs the checks that matter for each type it finds. Articles are checked for headline, author, datePublished, and dateModified. FAQPage is checked for mainEntity question/answer pairs with acceptedAnswer. Product is checked for name, description, offers, and brand. Organization and Person are checked for name, url, and sameAs links. HowTo, Event, and Recipe each get their own field checks. Unrecognised types are surfaced but not penalised.
Each check reports what it found, not just pass or fail — so the fix is concrete: 'Missing dateModified — helps engines detect updated content' rather than a generic 'schema incomplete'.
What this does and does not measure
Schema citation-readiness is measured — the same checks every time, over the exact JSON-LD you paste, with no model in the loop. What it does not do is check whether your schema matches your visible content, or whether the entities you declare are the ones engines actually reconcile. That needs a live crawl and is part of the full evidence-readiness report.
Nothing you paste is stored, and no external API is called. The analysis runs entirely in the request and returns the checks and score.
Frequently asked questions
What makes this different from a standard schema validator?+
Every schema validator on the web checks whether your JSON-LD is syntactically valid. That is necessary but not enough. An answer engine needs specific fields to attribute, date, and reconcile your content — a headline and author for articles, question/answer pairs for FAQPage, a price and brand for Product, sameAs links for entity identity. This tool checks what an engine actually needs, not just whether the JSON parses.
What schema types does it check?+
Article, Report, BlogPosting, NewsArticle, TechArticle, ScholarlyArticle (headline, author, dates), FAQPage (mainEntity with acceptedAnswer), Product (name, description, offers, brand), Organization and Person (name, url, sameAs), HowTo (steps), Event (name, startDate, location), and Recipe (ingredients, instructions). Unrecognised types are surfaced but not penalised.
Can I paste HTML instead of raw JSON-LD?+
Yes. Paste either raw JSON-LD or the full HTML page — the tool extracts every <script type="application/ld+json"> block and validates each one. If a block fails to parse, the error is reported with a snippet so you can fix it.
What is the citation-readiness score?+
The percentage of citation-readiness checks that pass across all schema types found. A perfect score means every field an engine needs is present. Missing fields are listed as concrete recommendations, not a generic 'fix your schema' message.
Is anything stored?+
No. The analysis runs entirely in the request and returns the checks. Nothing you paste is stored or sent to a third party.