identify-page-structure

To identify page structure from scraped webpage data, you provide a screenshot, cleaned HTML, and metadata JSON. The skill maps section boundaries and analyzes content sequences within each section to support structured authoring.

The best way to map content sequences is through a two-level analysis: first identifying broad page sections, then breaking down the content sequences within each section to produce neutral descriptions for downstream authoring.

Yes, you can use this skill to prepare inputs for downstream authoring and governance tasks. It outputs section boundaries, per-section sequence descriptions, and block inventory context specifically designed for orchestration with related authoring tools.

You need three specific inputs to analyze webpage sections: a scraped page screenshot (screenshot.png), cleaned HTML (cleaned.html), and metadata.json. These files allow the skill to determine section boundaries and map the block inventory.

Page structure identification works with cleaned HTML from any web scrape. You must supply the cleaned HTML file alongside a screenshot and metadata JSON to successfully extract section boundaries and content sequences.

Page structure analysis separates sections from content sequences to provide a structured two-level map. This separation outputs neutral, descriptive boundaries for each block, enabling precise downstream orchestration with page decomposition and authoring tools.