Structured Data Extractor for AI Agents

Given a public page URL, extract its machine-readable metadata: JSON-LD, schema.org types, OpenGraph tags, and microdata.

What it extracts

• JSON-LD blocks
• schema.org types present on the page
• OpenGraph metadata
• Microdata

This is extraction and normalization, not generic web scraping, LLM summarization, or semantic interpretation. The Actor performs a live extraction through Sentinel Verify.

Pricing

This is a paid, pay-per-event Actor: $0.03 per completed extraction.

Exactly one structured-data-extract event is charged per run, only after a real result exists (success or a useful partial). There is no charge for invalid input, a target blocked by the public-network policy, or a run that fails before a result is produced.

Input

• url (required): a public HTTPS page URL, e.g. https://example.com/product/widget.

Example input:

{
  "url": "https://example.com/product/widget"
}

Output

The default dataset receives one result envelope per URL; RUN_SUMMARY in the default key-value store records processing, delivery, billing, and dependency totals.

Representative abbreviated output:

{
  "status": "success",
  "target": {"identifier": "https://example.com/product/widget"},
  "result": {
    "url": "https://example.com/product/widget",
    "schemaTypes": ["Product", "Organization"],
    "jsonLd": [{"@type": "Product", "name": "Widget"}],
    "microdata": [],
    "openGraph": {"og:title": "Widget"}
  }
}

restricted means API-IFY rejected the target locally before calling Verify. unreachable means the page could not be reached. dependency_unavailable means Verify was temporarily unavailable.

How it works

This Actor is a thin wrapper around Sentinel Verify's POST /v1/utilities/extract-structured-data endpoint. The page URL is checked locally against API-IFY's public-network policy before being sent to Verify.

Limitations

• Public HTTPS page URLs only; private, loopback, link-local, metadata, reserved, and CGNAT targets are rejected.
• Extracts existing structured metadata as published — it does not infer or generate metadata a page doesn't already declare.
• Extraction availability depends on Sentinel Verify.

Privacy and security

The target URL is sent to Sentinel's live Utility API (/v1/utilities/extract-structured-data) for analysis after local public-network validation. Results are stored in the customer's Apify run dataset and summary store. VERIFY_API_KEY is an operator-managed Actor secret — a scoped utilities:read Intelligence API key sent as a bearer token on every request, the same credential mechanism used by any other Utility API consumer; VERIFY_BASE_URL is an operator setting, not an Actor input.

Support

For a reproducible support request, provide the Apify run ID, result itemId, and status. Contact Sentinel Signal Systems through the support link on the Actor page.