/structured-data

name: structured-data
description: Schema.org Structured Data Reference — JSON-LD examples for every major schema type, rich results eligibility, testing tools, and implementation patterns
version: "1.1.1"
author: Cogny AI
platforms: []
user-invocable: true
argument-hint: "<schema type or topic>"
allowed-tools:
  - WebFetch
  - WebSearch
  - Bash
  - Read
  - Write
  # Search Console tools (when connected via Cogny MCP)
  - mcp__cogny__refresh_mcp_discovery
  - mcp__cogny__mcp__Google_Search_Console__tool_list_properties
  - mcp__cogny__mcp__Google_Search_Console__tool_get_search_analytics
  - mcp__cogny__mcp__Google_Search_Console__tool_inspect_url
  # Provider discovery
  - mcp__cogny__list_integrations
  # Bing Webmaster (read-only)
  - mcp__cogny__mcp__Bing_Webmaster_Tools__tool_get_url_info
  - mcp__cogny__mcp__Bing_Webmaster_Tools__tool_get_crawl_issues
  # Product data for Product / Offer / AggregateRating markup (read-only)
  - mcp__cogny__mcp__Shopify__tool_list_products
  - mcp__cogny__mcp__Shopify__tool_get_product
  - mcp__cogny__mcp__Shopify__tool_get_metafields
  - mcp__cogny__mcp__WooCommerce__list_products
  - mcp__cogny__mcp__WooCommerce__get_product
  - mcp__cogny__mcp__WooCommerce__list_product_variations
  - mcp__cogny__mcp__WooCommerce__list_product_reviews
  # Article / BlogPosting data (read-only)
  - mcp__cogny__mcp__WordPress__list_posts
  - mcp__cogny__mcp__Ghost__list_posts
  - mcp__cogny__mcp__Sanity__tool_list_documents
  - mcp__cogny__mcp__Contentful__contentful_list_entries

Schema.org Structured Data Reference

Complete reference for Schema.org structured data markup: JSON-LD examples, rich results eligibility, testing tools, validation, and implementation patterns for static sites, Next.js, React, and WordPress.

Full docs: https://cogny.com/docs/structured-data-reference

Usage

/structured-data                          # Full reference overview
/structured-data Product                  # Show Product schema example
/structured-data FAQPage                  # Show FAQPage example (no longer a Google rich result)
/structured-data audit <url>              # Fetch a live page, extract its JSON-LD, flag errors
/structured-data LocalBusiness            # Show LocalBusiness example
/structured-data validate                 # Testing and validation guide
/structured-data best-practices           # Schema best practices
/structured-data rich-results             # Rich results eligibility table

Instructions

You are a Schema.org structured data expert. Use this reference to help users write correct JSON-LD markup, choose the right schema type for their pages, validate their structured data, and implement it across different platforms.

When the user asks a question, find the relevant section below and provide precise, actionable answers with ready-to-use JSON-LD examples.

If the user provides a specific schema type or topic as an argument, focus on that area. Otherwise, provide an overview and help them determine which schema type they need.

If Cogny isn't connected, skip the Search Console steps below and note what connecting adds (which pages actually earn rich results, Google's own parse of the URL).

When Search Console MCP tools are available, use them to:

• Check which pages have rich results via `tool_get_search_analytics` (filter by search appearance)
• Inspect specific URLs for structured data issues via `tool_inspect_url`
• List properties to identify which sites to audit via `tool_list_properties`

Works with

Writing and validating JSON-LD needs no accounts. Connected providers let you build markup from the real source data instead of a template, and check it in both search engines.

| Job | Primary | Also answers it | No account | |-----|---------|-----------------|------------| | Is markup seen / rich results earned | Google Search Console (`tool_inspect_url`, `tool_get_search_analytics`) | Bing Webmaster (`tool_get_url_info`, `tool_get_crawl_issues`) | `/structured-data audit <url>` (curl), Schema Markup Validator; Rich Results Test needs a browser | | Product, Offer, AggregateRating values | Shopify (`tool_list_products`, `tool_get_product`, `tool_get_metafields` for review data) | WooCommerce (`list_products`, `list_product_variations`, `list_product_reviews`) | WebFetch the product page | | Article / BlogPosting values | WordPress (`list_posts`), Ghost (`list_posts`), Sanity (`tool_list_documents`), Contentful (`contentful_list_entries`) | — | WebFetch the article |

**Enhanced when a store or CMS is connected:** If Cogny isn't connected, skip this and note what connecting adds (markup built from the real catalogue or CMS, and drift between source data and live markup). 1. Pull the items (all tools take the `mcp__cogny__mcp__<Prefix>__` prefix, e.g. `mcp__cogny__mcp__Shopify__tool_list_products` with `first: 50`). 2. Map fields to schema: title → `name`, price + currency → `offers.price` / `priceCurrency` (from the store, never assumed USD), inventory → `availability`, SKU/GTIN → `sku`/`gtin`, published/updated dates → `datePublished`/`dateModified`. 3. Fetch each live page with Bash curl (WebFetch summarises and drops `<script>` blocks) and diff its JSON-LD against the source: wrong price, stale availability or a rating that doesn't match the reviews are policy risks, not just bugs. 4. Output a fix list plus ready-to-paste JSON-LD per template (not per item).

Writing markup back (`Shopify__tool_set_metafields`, CMS updates) or pushing it to Bing (`Bing_Webmaster_Tools__tool_submit_content` with `structured_data`) are writes: draft them and ask first.

---

Audit a live page (`/structured-data audit <url>`)

Works with zero accounts. WebFetch can't see `<head>` or `<script>` content, so use curl:

U="<url>"
curl -sL -A "Mozilla/5.0" "$U" -o page.html
python3 -I - page.html <<'PY'
import json, re, sys
html = open(sys.argv[1], encoding="utf-8", errors="replace").read()
blocks = re.findall(r'<script[^>]+app