video-processing

What I do

• Download Videos - Fetch videos from YouTube, Facebook, and other platforms
• Extract Audio - Pull audio tracks from video files in various formats (MP3, WAV, AAC, etc.)
• Transcribe Audio - Convert speech to text with word-level timestamps using faster-whisper (supports Bengali and English)
• Analyze Videos - Detect scenes, extract keyframes, generate thumbnails and subtitles

Commands

Download a video

video download <url> [--output <path>] [--format <mp4|webm>] [--quality <quality>]

Extract audio from video

video extract-audio <video> [--output <path>] [--format <mp3|wav|aac>] [--bitrate <kbps>]

Transcribe video to text

video transcribe <video> [--output <path>] [--format <json|srt|vtt>] [--language <code>]

Analyze video comprehensively

video analyze <video> [--output <path>] [--threshold <0.0-1.0>] [--interval <seconds>]

Requirements

• Node.js 18+
• FFmpeg (system installation required)
• faster-whisper (CTranslate2-based Whisper implementation for faster local transcription)

Usage examples

# Download YouTube video
video download "https://youtube.com/watch?v=..." --output ./videos

# Extract MP3 from video
video extract-audio ./video.mp4 --format mp3 --bitrate 320k

# Generate subtitles
video transcribe ./video.mp4 --format srt --language en

# Analyze video with scene detection
video analyze ./video.mp4 --threshold 0.3

When to use me

Use this skill when you need to:

• Save online videos for offline viewing
• Convert video content to audio (podcasts, music)
• Create subtitles or transcripts from video content
• Analyze video structure for editing or highlights