Web & Page Intelligence API
Independent public-page reads for agents: Markdown conversion, schema extraction, change detection, metadata, displayed contacts, and sitemap indexes. Pages are fetched directly over HTTP. No browser session is used.
Fetch one public webpage and convert its readable HTML into Markdown. Scripts are not run, and missing text is not filled in.
Extract caller-defined fields from one public webpage. A field is returned only when a selector, JSON-LD property, or matching meta tag contains it. Otherwise the value is null.
Fetch a public page and compare its Markdown with a caller-supplied previous snapshot. The snapshot is not stored. The result lists added and removed lines and does not rank their importance.
Read title, description, canonical URL, Open Graph tags, language, and robots directives from one public page. Absent tags are null. robots.txt is not fetched.
Read publicly displayed emails, tel: links, address elements, and structured contact data from one public page. Linked pages are not crawled, and missing contacts are not guessed.
Fetch one public sitemap or sitemap index and return the listed URLs with lastmod, changefreq, and priority when those fields are present. Page URLs are not fetched.