# YouTube Knowledge Base
Search educational YouTube content directly from Claude Code. Access insights from top creators on marketing, engineering, growth, strategy, and more.
## Overview
YouTube KB is a public, open-source MCP (Model Context Protocol) server that provides semantic search across thousands of indexed YouTube educational videos. It's designed to work seamlessly with Claude Code through a lightweight plugin.
## Architecture
```
┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐
│ Supabase │◄───►│ Vercel │◄───►│ Claude Code │
│ (pgvector) │ │ (MCP Server) │ │ (Plugin) │
└─────────────────┘ └─────────────────┘ └─────────────────┘
```
## Packages
- **packages/server** - MCP server (TypeScript, deployed to Vercel)
- **packages/plugin** - Claude Code plugin for easy installation
## Available Domains
| Domain | Topics |
|--------|--------|
| `growth` | Marketing, acquisition, retention |
| `strategy` | Business models, pricing, positioning |
| `engineering` | Architecture, patterns, best practices |
| `product` | PM skills, roadmaps, prioritization |
| `startup` | Fundraising, founder advice |
| `sales` | Sales techniques, closing, pipelines |
| `design` | UI/UX, visual design |
| `ai-tools` | AI products, prompting |
| `ai-builder` | Building with AI, LLMs |
## Quick Start
### For Users
```bash
# Install the Claude Code plugin
claude plugin install youtube-kb
```
### For Developers
```bash
# Clone the repository
git clone https://github.com/unisone/youtube-kb.git
cd youtube-kb
# Install dependencies
pnpm install
# Set up environment variables
cp .env.example .env.local
# Edit .env.local with your credentials
# Run locally
pnpm dev
```
## Documentation
- [Product Requirements](docs/PRD.md)
- [Architecture Overview](docs/architecture/OVERVIEW.md)
- [Database Schema](docs/architecture/DATABASE.md)
- [API Specification](docs/specs/API.md)
- [Plugin Specification](docs/specs/PLUGIN.md)
- [Implementation Roadmap](docs/implementation/ROADMAP.md)
## Tech Stack
- **Database**: Supabase (PostgreSQL + pgvector)
- **Server**: Vercel (Edge Functions)
- **Embeddings**: OpenAI text-embedding-3-small
- **Protocol**: MCP with StreamableHTTP transport
- **Language**: TypeScript
## Contributing
We welcome contributions! See [CONTRIBUTING.md](CONTRIBUTING.md) for guidelines.
## License
MIT License - see [LICENSE](LICENSE) for details.
## Author
Alex Zaytsev ([@unisone](https://github.com/unisone))
