YouTube Research Agent

A powerful tool to analyze YouTube channels and generate actionable content insights.

Features

• Fetch & Analyze: Retrieves 20 most recent videos from any YouTube channel
• Smart Insights: Identifies performance patterns in content type, duration, and titles
• AI Suggestions: Generates data-driven video ideas based on what works
• Zero API Keys: Uses yt-dlp - no YouTube API key required

Installation

Prerequisites

Install yt-dlp:

# macOS
brew install yt-dlp

# Linux
pip install yt-dlp
# or
sudo apt install yt-dlp

# Windows
pip install yt-dlp

Setup

npm install

Usage

Slash Command

/youtube @channelname
/youtube https://youtube.com/@channelname

Direct Script

node scripts/youtube-research.js @channelname

Output

Generates youtube-research.md with:

🔑 Key Insights

• Performance patterns (content type, duration, title analysis)
• Audience preferences
• Data-driven observations

💡 Your Next Video

• 3 specific video ideas based on successful patterns
• Suggested duration and reasoning for each

📊 Top 10 Videos by Views

• Title
• View count
• Duration
• Clickable YouTube link

📈 Quick Stats

• Total and average views
• Average duration
• Most common duration range

Example

/youtube @mkbhd

Output:

🎬 YouTube Channel Research Tool
================================

📺 Channel: https://youtube.com/@mkbhd

🔍 Fetching channel data...
✅ Found 20 videos
📊 Fetching detailed statistics...
✅ Fetched details for 20 videos
🧠 Analyzing performance patterns...
✨ Analysis complete!

📄 Report saved to: youtube-research.md
📊 Top video: "iPhone 16 Review: The Best iPhone Yet?"
   1,234,567 views

How It Works

1. Fetch: Uses yt-dlp to extract video metadata (no API key needed)
2. Analyze: Processes 20 recent videos for patterns in:
   • Content types (tutorial, review, list, etc.)
   • Duration sweet spots
   • Title structures and hooks
   • View count trends
3. Generate: Creates insights and suggests next video ideas
4. Export: Writes comprehensive markdown report

Technical Details

• Language: Node.js (ES Modules)
• Data Source: yt-dlp
• Analysis: Pattern recognition, statistical averaging
• Output: Markdown format

Error Handling

• ✅ Missing yt-dlp installation detection
• ✅ Invalid channel name validation
• ✅ Network error handling
• ✅ Insufficient data warnings

License

MIT