web-to-markdown

Web-to-markdown batch-processes multiple URLs with a headless browser, extracts content rendered by JavaScript, converts HTML to Markdown using Turndown, and saves all results to a single timestamped file. Run `pnpm tsx scripts/scrape-and-convert.ts <url1> [url2] ...` from the skill directory.

Yes. Playwright handles JavaScript-rendered content by executing scripts before extraction, then Turndown converts the full DOM to clean Markdown. This handles dynamic pages that static HTML parsing cannot.

Pass all URLs as arguments to the scrape-and-convert script in one command. It processes them sequentially, consolidates output into a single timestamped Markdown file stored at `docs/web-captures/YYYYMMDD_HHMMSS.md` with deterministic formatting.

Turndown is ideal when you need readable, semantic Markdown output from HTML—useful for documentation capture, content archiving, and save-web-content workflows where preserving structure and links matters more than raw HTML.

The skill includes robust error handling with sequential processing, so failures on individual URLs do not halt the batch. Errors are logged, and successfully captured pages are still saved to the output file.