universal-video-notes

You can convert YouTube and Bilibili videos into Obsidian Markdown notes by providing the video URL; the Skill extracts metadata, acquires subtitles or uses local ASR, and outputs structured Markdown files with YAML frontmatter directly into your vault.

Yes, local ASR models like SenseVoice are required for converting videos to Obsidian notes when closed captions are unavailable, ensuring accurate transcription generation directly on your machine without relying on external cloud APIs.

Yes, you can batch process multiple videos to generate Obsidian Markdown files because the workflow is designed for batch-friendly pipeline execution, enabling automated end-to-end metadata extraction and note generation for numerous video URLs simultaneously.

To setup video transcription to Obsidian, you need to install yt-dlp for video access, configure DEEPSEEK_API_KEY for AI-generated YAML frontmatter, and set OBSIDIAN_OUTPUT_DIR in your environment to define the local Obsidian vault output directory.

To upload generated Obsidian video notes to Feishu Wiki, the Skill supports optional Feishu Wiki uploads as a flexible delivery mechanism, automatically syncing the structured Markdown output and image previews for team distribution.

Video to Obsidian note conversion fails to get subtitles when CC subtitles are unavailable on the source platform, triggering the robust error handling pipeline to fallback to local ASR with SenseVoice to transcribe the audio locally.