pocket-joe-mcp-toys
Enables transcription of YouTube videos to extract video titles, full transcripts, thumbnail URLs, and video IDs. Built with pocket-joe and deployable to Railway for scalable MCP server hosting.
README
pocket-joe-mcp-toys
MCP servers built with pocket-joe - YouTube transcription and more.
What's Inside
YouTube Transcriber
MCP server that transcribes YouTube videos, extracting:
- Video title
- Full transcript
- Thumbnail URL
- Video ID
Built using pocket-joe policies and deployed to Railway.
Deploy to Railway
Prerequisites
- Railway account (Hobby plan recommended)
- GitHub account
Steps
- Push this repo to GitHub
cd pocket-joe-mcp-toys
git init
git add .
git commit -m "Initial commit"
git remote add origin https://github.com/YOUR_USERNAME/pocket-joe-mcp-toys.git
git push -u origin main
-
Deploy to Railway
- Go to Railway Dashboard
- Click "New Project" → "Deploy from GitHub repo"
- Select
pocket-joe-mcp-toys - Railway auto-detects Python and deploys
- Wait for deployment (takes ~2-3 minutes)
-
Get your endpoint
- Click on your deployment
- Copy the public URL:
https://your-app.railway.app
Use in Claude Desktop
Add to your Claude Desktop config at ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"youtube": {
"url": "https://your-app.railway.app/mcp"
}
}
}
Restart Claude Desktop and you'll see the YouTube transcription tool available.
Local Development
Setup with uv
# Install uv if you haven't
curl -LsSf https://astral.sh/uv/install.sh | sh
# Clone and setup
git clone https://github.com/YOUR_USERNAME/pocket-joe-mcp-toys.git
cd pocket-joe-mcp-toys
uv sync
# Run locally
uv run python main.py
Server runs at http://localhost:8000/mcp
Test locally with Claude Desktop
Use http://localhost:8000/mcp in your config during development.
Architecture
This uses pocket-joe's policy system:
- Policy:
TranscribeYouTubePolicy- handles video transcription - AppContext: Manages policy instances and bindings
- FastMCP: Wraps policies as MCP tools with
@mcp.tool()decorator - Transport: Streamable HTTP for Railway
- Deployment: Stateless (scales horizontally)
Railway Configuration
- Runtime: Python 3.12
- Package Manager: uv (auto-detected via
pyproject.toml) - Entry Point:
python main.py - Port: Auto-assigned via
$PORTenv var
Cost
Railway Hobby plan:
- $5/month credit
- Enough for ~500k requests/month
- Auto-sleep after inactivity
- Custom domains included
Adding More Tools
To add more pocket-joe policies as MCP tools:
- Create your policy class in
main.py - Add to
AppContextwithself._bind(YourPolicy) - Create a FastMCP tool with
@mcp.tool()that calls your policy - Push to GitHub (Railway auto-deploys)
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。