Navin

Scraping actions - 12 cards

Each card sends /scrape with a precise specification; your brief is appended.

Collect

ActionDelivers
Scrape a pageClean text/markdown from one or more URLs via scrape action=fetch. Saves JSON + Markdown under scrape/, with a short quality summary.
Crawl a siteSame-domain BFS via scrape action=pipeline with depth/page caps. Exports XLSX + HTML report; flags empty/JS-gated pages for a browser pass.
From sitemap / listParallel fetch of a URL list or sitemap seeds. Deduplicated CSV + JSONL with ok/error counts.
JS-rendered pageUses the browser tool for client-rendered pages, normalizes into scrape records, then scrape action=export.

Clean & enrich

ActionDelivers
Clean a corpusStrip nav/chrome leftovers, normalize whitespace, drop near-duplicates; rewrite cleaned files with a removal note.
Enrich metadataAdd description/OG fields, language guess, word count, fetched_at; enriched JSON/CSV + field dictionary.
Extract tablesHTML tables → structured CSV/Excel rows with source URL and normalized column names.
Prepare for RAGOne Markdown file per page with YAML frontmatter (url, title, fetched_at) plus manifest.json.

Export

ActionDelivers
Export ExcelWorkbook (xlsx) with url, title, status, text, error; optional summary sheet.
Export CSV + JSONMachine-friendly dumps with matching row counts and a one-line schema note.
Export XMLWell-formed pages/page XML for legacy ingest.
HTML reportReadable report (format=report) plus executive summary: pages, success rate, paths, next crawl.

Source: content/docs/navin_scraping/en/actions.md