<2-3 sentences summarising what the project addresses and solves, the target audience, and the benefit.> This project offers 14 AI Skills to ingest diverse Chinese and international online content—video, articles, social media—into a personal knowledge base. It targets users building a searchable digital second brain, automating information capture and organization across platforms.

How It Works

Modular "AI Skills" loadable by compatible Agents automate content acquisition from platforms like Douyin, Bilibili, Xiaohongshu, WeChat, X, and YouTube. A key advantage is "subtitle-first" transcription, minimizing GPU needs by prioritizing existing subtitles. Content is standardized into Markdown with frontmatter for Obsidian integration, and includes an MCP server for agent querying.

Quick Start & Requirements

Installation uses setup.sh or individual requirements.txt. Content enrichment and YouTube translation require DEEPSEEK_API_KEY; XHS_COOKIE is recommended for Xiaohongshu. System dependencies like ffmpeg and yt-dlp are needed for media processing.

Highlighted Details

• Broad Platform Support: Ingests from Douyin, Bilibili, TikTok, Weibo, Zhihu, YouTube, Podcasts, WeChat Official Accounts, Xiaohongshu, and X.
• GPU-Efficient Transcription: Leverages subtitles where available, reducing hardware demands.
• Unified Knowledge Format: Standardized Markdown frontmatter aids Obsidian/Dataview integration.
• Automated Enrichment: Adds summaries, key points, and tags using DeepSeek.
• Knowledge Base Management: Tools for Obsidian, including ingestion, health checks, and MCP server.
• Specialized Skills: Includes industry intelligence radar, learning notes automation (flashcards), and content processing.

Maintenance & Community

Maintained by @chubbyguan, the project acknowledges numerous open-source libraries. Community interaction occurs via linked social media (X/Twitter, Jike, Xiaohongshu) and a WeChat Official Account.

Licensing & Compatibility

MIT License permits free personal use, modification, and redistribution. Users must adhere to platform ToS, robots.txt, and legal regulations, avoiding commercial use or batch scraping. Content copyright remains with original authors.

Limitations & Caveats

Users must comply with platform ToS and rate limits; non-compliance may lead to blocking. DEEPSEEK_API_KEY is mandatory for certain enrichment features. X video transcription requires specific dependencies. Provided "as is," author disclaims responsibility for usage consequences.