Structured Data Reviewer
Inspect only the HTML snapshot and JSON-LD you provide. This bounded reviewer lists graph nodes, flags unresolved same-page references and checks documented Article recommendations and BreadcrumbList shapes. It compares selected text fields with supplied visible HTML as candidates for manual review, never as a truth or rich-result verdict.
Key features
- Extract local JSON-LD script blocks without running or fetching the page
- List typed nodes and resolve same-page fragment @id references
- Apply Google-documented BreadcrumbList required shapes and Article recommendations separately
- Compare selected headline, author and breadcrumb strings with visible HTML snapshot
- Keep candidate matches in a manual confirmation checklist
- Export review JSON and CSV without raw HTML or JSON-LD
How to use
- Paste or import a saved HTML snapshot and enter its absolute page URL.
- Paste JSON-LD separately or use the script blocks already in the HTML.
- Choose which visible facts to review and run the analysis.
- Inspect typed nodes, reference/property diagnostics and candidate text matches.
- Manually confirm facts against the full live page, then download the review report.
Use cases
- Review a blog Article graph before publishing
- Spot a broken local author @id fragment
- Check breadcrumb list positions and required item shapes
- Compare markup text against a provided rendered HTML snapshot
Frequently asked questions
Does the tool fetch or execute the page?
No. It parses supplied text only. HTML scripts are never executed, remote contexts are not resolved and page URLs are used solely as a base for local reference comparison.
Does a clean report mean Google will show a rich result?
No. Google does not guarantee rich results. Its eligibility, crawlability, policy and live-page content checks remain outside this local review.
Are Article fields required?
Google currently lists no required Article properties. Applicable headline, author, publication date and image fields are shown as recommendations, while malformed supplied values are diagnosed separately.
Which breadcrumb rules are checked?
The bounded profile checks at least two ListItem entries, positive positions, names, and item URLs or identifiers except on the final entry. It also warns if positions do not follow display order.
Can a visible-text match be trusted automatically?
No. The candidate check only finds a text string in supplied HTML after basic hidden-element filtering. JavaScript-rendered content, CSS and page context may differ; a person must confirm the full page.
Privacy
HTML, JSON-LD and checklist state remain in this browser tab. Downloads contain node labels, paths, diagnostic text and selected fact strings, but omit raw input. Inspect sensitive paths and facts before sharing a report.
Comments & questions