seo-schema extracts structured data from a URL, including JSON-LD, Microdata and RDFa, and validates it against required properties, data types, absolute URLs, date formats and @context and @type rules, with correct handling of @graph containers.
It keeps a dated list of which types are active, which no longer give rich results (FAQPage, HowTo and others) and which are retired, and it generates truthful JSON-LD with marked placeholders. Output is a SCHEMA-REPORT.md and a generated-schema.json file.
What You Get
- SCHEMA-REPORT.md with a validation table per schema block
- generated-schema.json with ready JSON-LD snippets
- Templates for Organization, LocalBusiness and Article
- Retirement dates and active replacements for deprecated types
Common Use Cases
- Check a page’s structured data before launch
- Fix invalid JSON-LD syntax and missing required properties
- Replace retired markup such as HowTo with an active alternative
- Generate Organization or Article markup for a new site
What Makes It Unique
It carries a dated rich result status list and a reference file on retired types, so it does not recommend markup that no longer earns a result. It also rejects fake or incentivized reviews and understands @graph inheritance, which avoids false errors on valid documents.
Under The Hood
The skill is instructions with one reference file on deprecated types and a pointer to shared schema type lists in the plugin. For detection it runs the plugin’s parse_html.py through a launcher, which fetches the page over the network and extracts the JSON-LD. It writes the report and the snippets to files. The frontmatter declares none of this access. Fetched pages are to be treated as untrusted data.