mofa-youtube

Generate timestamped transcripts and metadata.json from YouTube or local audio.

11|12|Updated Feb 28, 2026
One-click install
npx skills add https://github.com/mofa-org/mofa-skills --skill mofa-youtube
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mofa-youtube
Source: https://github.com/mofa-org/mofa-skills/tree/main/mofa-youtube
Command: npx skills add https://github.com/mofa-org/mofa-skills --skill mofa-youtube

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill automates the extraction of transcripts and structured metadata from YouTube videos or local audio sources, saving time and enabling easy repurposing for accessibility, indexing, and content creation.

Core Features & Use Cases

  • Download and ingest: fetch YouTube videos or process local audio inputs.
  • Transcript and metadata output: produce a markdown transcript with timestamps and a JSON metadata package (metadata.json) for downstream tooling.
  • Use Case: create searchable video archives, generate subtitles, and summarize long videos for quick previews.

Quick Start

Provide a YouTube URL or local audio path to generate a timestamped transcript and metadata package.

Frequently Asked Questions about mofa-youtube

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

FAQPage Schema
How do I transcribe a YouTube video and export the transcript?

To transcribe a YouTube video, provide a YouTube URL or local audio path. The Skill downloads the video, processes audio chunks of up to 4 minutes each using a local ASR service, and exports a markdown transcript with timestamps.

What metadata is included when transcribing YouTube videos?

Transcribing YouTube videos produces a metadata.json package containing duration, language, and chapters. This structured output enables downstream tooling for indexing, search, and content repurposing.

Do I need ffmpeg and yt-dlp to download and transcribe YouTube audio?

Yes, you need yt-dlp to fetch YouTube videos or local audio, and ffmpeg or ffprobe to process the media. The Skill also relies on curl and jq to handle downloads and structure the metadata output.

Can I use local audio files instead of YouTube URLs for transcription?

Yes, you can process local audio inputs instead of YouTube URLs. The Skill ingests local audio paths, transcribes the audio in chunks, and generates a timestamped markdown transcript and metadata.json file.

What is the maximum audio chunk length for ASR transcription?

The maximum audio chunk length for ASR transcription is 4 minutes. The Skill splits longer audio into these segments to ensure accurate speech recognition and generate a complete transcript.

What's the best way to create searchable video archives from YouTube?

The best way to create searchable archives is using this Skill to extract timestamped transcripts and structured metadata from YouTube videos. The markdown and JSON outputs enable accessibility, search indexing, and content summarization.