content-extraction
Installation
SKILL.md
This skill extracts ALL content from an existing website and outputs it as structured, reusable data files. It crawls every page, downloads every asset, and produces a complete content inventory.
The user provides: the URL of the site to extract from, and optionally the target format (TypeScript, JSON, Markdown).
What Gets Extracted
For each page on the site:
| Content type | Output |
|---|---|
| Text | Headings, paragraphs, lists, quotes — preserved with hierarchy |
| Metadata | <title>, <meta description>, OG tags, canonical URL, lang |
| Images | Downloaded to public/images/ with original filenames. Alt text cataloged |
| Links | Internal + external, with anchor text and destination URL |
| PDFs & assets | Downloaded to public/assets/. Filenames and original URLs cataloged |
| Forms | Field names, types, labels, validation rules, action URLs |
| Navigation | Menu structure, link hierarchy, active states |
| Structured data | JSON-LD, microdata, schema.org markup |