yt-dlp
lwmxiaobei/yt-dlp-skill
Download videos and extract audio from YouTube, TikTok, Twitter, Instagram, and 100+ other platforms.
What is yt-dlp?
yt-dlp is a video downloader skill that fetches videos and audio from major platforms like YouTube, Twitter/X, TikTok, Instagram, and many others. Use it when a user provides a video URL, asks to download content, or when video links appear in conversation.
- Download videos from 100+ platforms including YouTube, TikTok, Twitter/X, Instagram, Vimeo, and Facebook
- Extract audio in multiple formats (MP3, M4A, OPUS, FLAC, WAV)
- Auto-detect video URLs in conversation and prompt for download
- Support quality selection (best, 1080p, 720p, 480p, audio-only)
- Extract video metadata and information without downloading
- Handle multiple URLs and playlist downloads
How to install yt-dlp
npx skills add https://github.com/lwmxiaobei/yt-dlp-skill --skill yt-dlp- yt-dlp installed (pip install yt-dlp or brew install yt-dlp)
- ffmpeg installed for audio/video format conversion (brew install ffmpeg or apt install ffmpeg)
- Python 3 with standard library
How to use yt-dlp
- 1.Provide a video URL or ask to download a video from a supported platform
- 2.Specify desired quality (best, 1080p, 720p, 480p) or format (for audio: mp3, m4a, opus, flac, wav) if needed, or accept defaults
- 3.The skill extracts the URL, confirms the action, and executes the download
- 4.Retrieved files are saved to your specified output directory
Use cases
- User shares a YouTube link and asks you to download it for offline viewing
- Extract audio from a TikTok or Instagram video as an MP3 file
- Detect video URLs mentioned casually in conversation and offer to download them
- Download a Twitter/X video before it's deleted
- Batch download multiple videos from the same platform
- Content creators who need to archive or repurpose videos
- Users wanting to download videos for offline access
- Researchers collecting video data from public sources
- Anyone needing audio extraction from video content
yt-dlp FAQ
YouTube, Twitter/X, TikTok, Instagram, Facebook, Vimeo, Twitch, Dailymotion, Reddit, Streamable, and 100+ others supported by yt-dlp.
MP3 (default), M4A, OPUS, FLAC, and WAV. Default quality is 192kbps MP3.
Yes. The skill can handle multiple URLs and offers playlist download options if URLs are from the same source.
Some videos may need cookies or authentication. The skill will report if a video is unavailable or inaccessible.
Yes, ffmpeg is required for audio/video format conversion. Install via brew (macOS) or apt (Linux).
Full instructions (SKILL.md)
Source of truth, from lwmxiaobei/yt-dlp-skill.
name: yt-dlp description: Download videos and extract audio from various platforms using yt-dlp. Use when user provides a video URL, asks to download a video, or when conversation contains video links from YouTube, Twitter/X, Vimeo, TikTok, Instagram, etc.
yt-dlp Video Downloader Skill
This skill provides tools for downloading videos and extracting audio from various platforms using yt-dlp.
Features
- Download videos from multiple platforms (YouTube, Twitter/X, Vimeo, TikTok, Instagram, Facebook, etc.)
- Extract audio from videos
- Auto-detect video URLs in conversations
- Support for different quality settings and formats
Usage Patterns
1. Command-based Download
When user explicitly asks to download a video:
User: Download this video https://youtube.com/watch?v=...
Action: Extract URL and call download script
2. Auto-detection in Conversations
When conversation contains video URLs:
User: Check out this video https://twitter.com/... and let me know what you think
Action: Detect video URL, ask user if they want to download it
3. Audio Extraction
When user wants to extract audio only:
User: Extract the audio from https://youtu.be/...
Action: Use audio extraction script
Available Scripts
Note: Scripts are located in the scripts/ directory
download_video.py
Main video downloader with quality and format options.
Usage:
# Download video
scripts/download_video.py <url> -o <output_dir>
# Download with specific quality
scripts/download_video.py <url> --quality 720p
scripts/download_video.py <url> --quality audio # For audio only
# Custom format selector
scripts/download_video.py <url> --format "bestvideo[height<=1080]+bestaudio/best"
# Extract info only
scripts/download_video.py <url> --info-only
Quality options: best, 1080p, 720p, 480p, audio
extract_audio.py
Extract audio from videos in various formats.
Usage:
# Extract as MP3 (default)
/scripts/extract_audio.py <url> -o <output_dir>
# Extract as M4A
/scripts/extract_audio.py <url> --format m4a
# Custom quality
/scripts/extract_audio.py <url> --quality 320
Formats: mp3, m4a, opus, flac, wav
extract_urls.py
Extract video URLs from text or files.
Usage:
# Extract from text argument
/scripts/extract_urls.py "Check https://youtube.com/watch?v=..."
# Extract from file
/scripts/extract_urls.py <file_path>
# Read from stdin
cat file.txt | /scripts/extract_urls.py
Video Platform Support
The skill recognizes URLs from:
- YouTube (youtube.com, youtu.be)
- Twitter/X (twitter.com, x.com)
- Vimeo (vimeo.com)
- TikTok (tiktok.com)
- Instagram (instagram.com)
- Facebook (facebook.com, fb.watch)
- Twitch (twitch.tv, clips.twitch.tv)
- Dailymotion (dailymotion.com)
- Reddit (reddit.com)
- Streamable (streamable.com)
- And many more supported by yt-dlp
Workflow
When User Provides Video URL
- Extract URL from user's input using
extract_urls.py - Confirm with user what action to take:
- Download video
- Extract audio
- Show video info
- Execute appropriate script based on user's choice
- Notify user of success/failure and file location
When Auto-detecting URLs
- Scan conversation text with
extract_urls.py(can process stdin) - If video URLs found, ask user: "I found video URLs in this conversation. Would you like me to download them?"
- If yes, proceed with download workflow
- If no, continue with conversation
Handling Multiple URLs
- For single URL: Direct download
- For multiple URLs: Ask user if they want to download all or select specific ones
- Provide option to download as playlist if URLs are from the same source
Quality and Format Selection
When user doesn't specify preferences:
- Default to best available quality
- For audio: Default to MP3 at 192kbps
When options needed:
# Ask user for quality preference if not specified
# Options: best (default), 1080p, 720p, 480p, audio
# Ask for format if extracting audio
# Options: mp3 (default), m4a, opus, flac, wav
Error Handling
Common issues and solutions:
-
yt-dlp not installed:
- Check with
yt-dlp --version - Install with
pip install yt-dlporbrew install yt-dlp
- Check with
-
ffmpeg not installed (required for format conversion):
- Install with
brew install ffmpeg(macOS) - Or
apt install ffmpeg(Linux)
- Install with
-
Video not available:
- Check if URL is accessible
- Some videos may require authentication
- Age-restricted content may need cookies
-
Network errors:
- Retry download
- Check internet connection
Dependencies
yt-dlp: Main video downloaderffmpeg: Audio/video processing (required for format conversion)python3with standard library
All scripts are self-contained and use only built-in Python modules.
Related skills
More from lwmxiaobei/yt-dlp-skill and the wider catalog.

lynx-devtool
Debug Lynx apps on Android, iOS, and Desktop using DevTool CLI and programmatic APIs.

feishu-doc
Read, write, and manage comments in Feishu documents via structured API actions.

feishu-drive
|

betting
Odds conversion, edge detection, Kelly criterion, arbitrage, and parlay analysis—pure computation, no API calls.

football-data
World football data across major leagues—standings, schedules, xG, transfers, Elo ratings, and match forecasts. Zero config.

kalshi
|