Can I get a clean text file instead of a VTT file with timestamps?

Absolutely. The skill includes a post-processing script that removes metadata, timestamps, and the 'overlapping' duplicates common in YouTube auto-captions to produce a clean .txt file.

What happens if a YouTube video doesn't have any subtitles?

The skill will detect the lack of subtitles and offer to download the video's audio to transcribe it locally using OpenAI's Whisper model.

Does it support non-English transcripts?

Yes, it can list all available subtitle tracks for a video, allowing you to download manual or auto-generated captions in any available language.

Does this skill require external software to be installed?

Yes, it relies on yt-dlp for the core extraction. The skill is designed to check for this dependency and can guide you through the installation via Homebrew, apt, or pip.

YouTube Transcript Downloader

Name: YouTube Transcript Downloader
Author: Jst-Well-Dan

byJst-Well-Dan

ウェブスクレイピングとデータ収集

Extracts and converts YouTube transcripts or auto-generated captions into clean, readable text files.

概要

This skill empowers Claude to fetch subtitles and closed captions directly from YouTube videos using yt-dlp. It intelligently prioritizes high-quality manual transcripts, falls back to auto-generated captions when necessary, and can even utilize OpenAI Whisper for local audio-to-text transcription as a last resort. Beyond simple downloading, it includes advanced post-processing capabilities to deduplicate overlapping text and convert VTT files into clean, readable plain text, making it an essential tool for content research, video summarization, and documentation workflows.

主な機能

Multi-source extraction: priorities manual subtitles, then auto-captions, then AI transcription
Whisper AI integration: generates transcripts for videos that have no captions available
Flexible output formats: supports both raw VTT with timestamps and cleaned plain text files
Automated environment setup: checks for and installs yt-dlp and Whisper dependencies
Intelligent post-processing: removes VTT timestamps and deduplicates overlapping lines
0 GitHub stars

ユースケース

Analyzing video content for research by extracting full text for sentiment or keyword analysis
Converting educational video tutorials into searchable written documentation or notes
Creating blog posts, summaries, or articles based on existing YouTube video content

概要

主な機能

Multi-source extraction: priorities manual subtitles, then auto-captions, then AI transcription
Whisper AI integration: generates transcripts for videos that have no captions available
Flexible output formats: supports both raw VTT with timestamps and cleaned plain text files
Automated environment setup: checks for and installs yt-dlp and Whisper dependencies
Intelligent post-processing: removes VTT timestamps and deduplicates overlapping lines
0 GitHub stars

ユースケース

Analyzing video content for research by extracting full text for sentiment or keyword analysis
Converting educational video tutorials into searchable written documentation or notes
Creating blog posts, summaries, or articles based on existing YouTube video content