Schema Extractor: JSON-LD, Microdata, RDFa

About

Rich results don't happen by accident. For SEOs who know valid ≠ correct.

Extract structured data from any web page in three formats: JSON-LD (from <script type="application/ld+json"> blocks), microdata (itemscope/itemtype/itemprop), and RDFa Lite (typeof/property).

Useful for SEO professionals, developers, and anyone auditing structured data on production sites, without view-source archaeology.

TOOLBAR BADGE WITH AUTO-DETECTION
A number on the extension icon shows how many structured-data items are present on the current page. No badge means no markup detected. The count updates automatically as you browse, including on single-page applications (Next.js, Nuxt, and similar) where schema gets injected after route changes.

TABBED VIEWER
Click the icon to open a viewer tab with three tabs, one per format. Each extracted item shows:
- The detected schema type (Product, BreadcrumbList, Organization, etc.)
- A summary band with five metrics: levels deep, schema types, unique properties, total objects, schema blocks
- The full JSON content with proper formatting
- Copy and Visualize buttons

Microdata and RDFa are normalized to JSON-LD shape (@context, @type, @id, nested objects) so the same view works for all three formats and downstream tools parse all three identically.

EXPORT OPTIONS
- Copy all: combined JSON document with all three formats, page URL, and timestamp
- Download .json: same content as a file
- Download .csv: one row per structured-data block, with columns for Item label, full schema JSON, format, and URL. Stack CSVs from multiple pages into one workbook and pivot to spot coverage gaps across a site.

CSV output is UTF-8 with BOM, so Excel handles international characters correctly.

VISUALIZER INTEGRATION
Each extracted item has a Visualize button that copies the JSON to clipboard and opens the Schema Markup Visualizer at magstags.com — paste to render the structured data as an interactive tree with type badges, property counts, and nesting depth.

PRIVACY
The extension runs locally. The content script counting items on each page reads the DOM and sends a numeric count to the service worker — nothing leaves your browser. No analytics, no telemetry, no external requests except when you click Visualize to open the magstags viewer in a new tab.

PERMISSIONS EXPLAINED
- activeTab: needed to read the page DOM when you click the icon
- scripting: to inject the extraction code at click time
- storage: to pass extracted data from the background script to the viewer tab
- access to all sites (via content scripts): required for the content script that counts items and updates the toolbar badge on every page