fetch-youtube-transcript

YouTube Transcript Download Skill

Purpose

Download transcripts from YouTube videos as readable text files with timestamps. This skill wraps the youtube-transcript-api tool with a simple interface for quick transcript extraction.

Usage

/fetch-youtube-transcript <video-url-or-id> [--output DIR] [--language LANG]

Arguments

Required

• <video-url-or-id>: YouTube video URL or video ID
  • Full URLs: https://www.youtube.com/watch?v=VIDEO_ID
  • Short URLs: https://youtu.be/VIDEO_ID
  • Video ID only: VIDEO_ID (11 characters)
  • URLs with timestamps and query parameters are supported

Optional

• --output DIR: Output directory for transcript files

  • Default: .work/transcripts/
  • Directory will be created if it doesn't exist
  • Transcript saved as VIDEO_ID.txt

• --language LANG: Preferred transcript language code

  • Default: Auto-select (usually English if available)
  • Examples: en, es, fr, de, ja
  • Falls back to auto-generated if manual transcript unavailable

Execution Instructions for Claude Code

When the user invokes /fetch-youtube-transcript, follow these steps:

1. Parse Arguments

Extract the video URL/ID and optional parameters from the user's command:

# Example argument parsing logic
import re
from pathlib import Path

def parse_args(args_string):
    """Parse /fetch-youtube-transcript arguments."""
    parts = args_string.split()

    if not parts:
        return None, None, None, "Error: Video URL or ID required"

    video_url = parts[0]
    output_dir = ".work/transcripts"
    language = None

    # Parse optional flags
    i = 1
    while i < len(parts):
        if parts[i] == "--output" and i + 1 < len(parts):
            output_dir = parts[i + 1]
            i += 2
        elif parts[i] == "--language" and i + 1 < len(parts):
            language = parts[i + 1]
            i += 2
        else:
            i += 1

    return video_url, output_dir, language, None