transcribe-video

Transcribe YouTube, podcast, or audio URLs into verbatim Markdown transcripts with metadata.

16|9|Updated Apr 4, 2026
One-click install
npx skills add https://github.com/sonthanh/brain-os-plugin --skill transcribe-video
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: transcribe-video
Source: https://github.com/sonthanh/brain-os-plugin/tree/main/skills/transcribe-video
Command: npx skills add https://github.com/sonthanh/brain-os-plugin --skill transcribe-video

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Transcribes YouTube/podcast/audio URLs into clean, citation-ready text, ensuring verbatim transcripts when quotes matter and providing reliable fallback paths for higher quality output.

Core Features & Use Cases

  • Auto-caption path by default for fast, complete transcripts, tolerant of minor caption typos.
  • Whisper path (--whisper) for higher quality transcripts in challenging audio or when proper nouns need accurate spelling.
  • Outputs two artifacts: <out>/_transcript-verbatim.md (verbatim transcript with metadata) and <out>/_transcription-metadata.json (build metadata).
  • Stores results in the specified output directory, enabling seamless integration with research findings workflows.
  • Storage context and wiring: designed to work when /research invokes transcription before finding quotes, ensuring quoted content is traceable to a primary source.

Quick Start

Provide a video or audio URL to /transcribe-video and an optional --out path to save the transcript and metadata.

Frequently Asked Questions about transcribe-video

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I transcribe a YouTube video into a verbatim text transcript?

To transcribe a YouTube video into a verbatim text transcript, provide the video URL to automatically extract captions or use the high-quality Whisper fallback, outputting a clean markdown transcript and metadata JSON file.

What is the best way to get accurate quotes from a podcast audio URL?

Getting accurate quotes from a podcast audio URL involves using an automated transcription process that extracts verbatim text, applying a high-quality fallback path for challenging audio to ensure proper noun spelling and citation-ready accuracy.

Do I need yt-dlp and ffmpeg installed to transcribe video URLs?

Yes, you need yt-dlp and ffmpeg installed to transcribe video URLs, as these dependencies handle the underlying audio extraction required to process the media and generate verbatim markdown transcripts.

Can I use Whisper for transcribing audio with challenging background noise?

Yes, you can use the Whisper path for transcribing audio with challenging background noise, which provides higher quality transcription results specifically when proper nouns need accurate spelling or when default captions fail.

How does automatic video captioning handle minor typos in YouTube transcripts?

Automatic video captioning handles minor typos in YouTube transcripts by using an auto-caption path by default that tolerates minor caption errors to deliver fast, complete, and verbatim markdown transcripts.

What output files are generated when transcribing an accessible audio URL?

Transcribing an accessible audio URL generates two output files: a verbatim markdown transcript containing the text with metadata, and a separate JSON file capturing the transcription build metadata.