
MCP Workshop Starter
A starter kit for building Model Context Protocol servers that enables AI tools to access external data and functionalities like checking holidays, disk space, timezones, RSS feeds, code diffs, and web performance metrics.
README
Model Context Protocol — Workshop starter
Ideas
- Get public holidays: I have 15 remaining vacation days, use public holidays and optimize when to get them.
- Get HDD free space: How many copies of War and Peace I can store in the remaining space on my HDD?
- Convert timezones: What time in Valencia is 9am NY time? Here’s the agenda for the conference, show me the schedule in Sofia time.
- Get latest news from RSS: What are the positive news from CNN? Summarize what’s happening in the USA right now in the style of Bill Hicks.
- Get diff of pull request: Perform code review of this open source PR https://github.com/rails/rails/pull/55072 (hint if you append .diff to the URL you get the git diff patch)
- Get Core Web Vitals for an URL: How can I improve the performance of this website https://www.flywire.com
Resources
- https://modelcontextprotocol.io/
- https://github.com/yjacquin/fast-mcp
- https://github.com/geelen/mcp-remote
Getting started with plain Ruby
Install dependencies with:
gem install fast-mcp
Getting started with Sinatra
Install dependencies with:
bundle install
Start the server with:
ruby sinatra.rb
Getting started with Node
Install dependencies with:
npm install
Testing with MCP Inspector
Start the official MCP inspector:
npx @modelcontextprotocol/inspector
Open http://127.0.0.1:6274/ and connect to you server:
Plain Ruby
- transport: stdio
- command: ruby
- arguments: /path/to/server.rb
Sinatra
- transport: sse
- URL: http://127.0.0.1:4567/mcp/sse
Node
- transport: stdio
- command: node
- arguments: /path/to/build/index.js
Go to Tools and explore.
Testing with AI agent
You can use the following clients:
- Claude Desktop
- Visual Studio Code
- Cursor
- Zed
Follow its configuration to add the MCP server to the list of tools.
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。