Notion Content Ingestion Skill — Multi-Tool Reference
Purpose
This skill provides machine-readable ingestion of Notion self-study guide databases into local Markdown: recursive crawler → sitemap.yaml, then an enricher (yt-dlp / crw / Scrapling / webfetch) → manifest.json. Any startup with a Notion "Biblioteca de Guías" database can produce a local, versioned, agent-searchable content dump.
Capabilities
- Recursive crawl: walk page/database tree → sitemap.yaml (children + resources)
- Bookmark resolution: verbatim
bookmark.urlvia Notion REST API (not the native MCP<unknown>blind spot) - Enrichment: classify resources by kind (youtube/article/pdf/image/notion_page); fetch via yt-dlp / crw / Scrapling / webfetch
- Resilience: unresolvable bookmarks recorded as
resolved: false(never crash) - Manifest: per-resource status (dumped/enriched/pending/failed) + per-guide hub status
Usage
This skill enables Notion-content ingestion for agents operating across Claude Code, Cursor, and OpenCode. Runs the shared pipeline (notion-batch MCP + scripts/notion-enrich.py); protocol SSOT is references/logistics-import/protocol.md.
References (APA)
- Notion. (2026). Notion API. https://developers.notion.com/
- yt-dlp. (2026). yt-dlp — youtube-dl fork with additional features. https://github.com/yt-dlp/yt-dlp
- Scrapling. (2026). Scrapling — stealth web scraping. https://github.com/D4Vinci/Scrapling