What this prompt does: Creates and validates JSON-LD structured data for different page types.
Best AI tool: Claude Opus 4.8, strong fit for technical, structured-output tasks where accuracy matters.
Best for: FAQ, Article, Product, Organization, Breadcrumb, and other schema implementation.
Prompt
[task]
Generate valid JSON-LD structured data for the webpage below.
Page type:
[ARTICLE / FAQ / PRODUCT / ORGANIZATION / BREADCRUMB / OTHER]
URL:
[URL]
Page content:
[PASTE CONTENT]
Business information:
[BUSINESS DETAILS]
The goal is to produce schema markup that accurately represents the visible content and follows current Schema.org and Google Search requirements.
[context]
Important rules:
– Only include information that is actually present or reliably provided.
– Never invent ratings, reviews, prices, authors, dates, or claims.
– Use the most appropriate Schema.org type.
– Use @context and @type correctly.
– Use canonical URLs where applicable.
– Keep properties relevant to the page.
– If a property is required but information is missing, flag it rather than inventing it.
– Do not add deprecated or unsupported properties simply because they appear in older examples.
[examples]
For an article, consider:
Article
headline
description
author
datePublished
dateModified
image
publisher
mainEntityOfPage
For FAQ content, only mark up questions and answers that are visibly present on the page.
[persona]
Act as a technical SEO specialist and structured-data engineer with expertise in Schema.org and Google Search documentation.
[format]
Return:
- Recommended schema type
- JSON-LD code
- Validation checklist
- Required/missing information
- Implementation notes
- Potential warnings/errors
Put the final JSON-LD in a single code block.
[tone]
Technical, precise, conservative, and implementation-ready. Accuracy is more important than adding more schema properties.