Claude skill that reads paywalled news so you don’t have to

What it does This Claude Code Skill ingests content from over 15 sources—WeChat articles, YouTube videos, podcasts, PDFs, EPUBs, Office documents, images, and audio—and uploads it to Google NotebookLM. It handles the entire fetch-convert-upload chain, including OCR, audio transcription, and format normalization, so you can request a podcast, PPT, mind map, or quiz in plain English. Think of it as a dedicated ingestion layer that does the tedious scraping and file-wrangling NotebookLM won’t do for you.

The interesting bit The standout feature is a six-level paywall bypass cascade that tries proxy services, search-engine bot user agents, AMP pages, and archive.today before falling back to a local fetch tool. The README explicitly credits techniques borrowed from the Bypass Paywalls Clean browser extension, giving the project a pragmatic, slightly grey-hat flavor. It also uses an MCP browser simulation to grab WeChat articles, which is a nice touch for a Chinese-centric workflow.

Key highlights

• Ingests 15+ source types, including X/Twitter threads, ZIP archives, and scanned documents.
• Automatically bypasses paywalls on 300+ news sites via a tiered fallback strategy.
• Generates podcasts, PPTs, mind maps, quizzes, and reports by piping cleaned content into NotebookLM.
• Supports mixed-source synthesis, letting you combine a paywalled article, a video transcript, and an EPUB into one deliverable.

Verdict If you live in NotebookLM and constantly hit paywalls or need to process Chinese social media content, this skill saves you from manual drudgery. If you already have clean, open text files and don’t need automated fetching, it is mostly elaborate glue.