my-mcp-server
A playground MCP server that demonstrates exposing tools and resources to AI agents, enabling them to manage a simple user database with create and read operations.
README
my-mcp-server
A playground and experimentation project for building an MCP (Model Context Protocol) server using TypeScript. It demonstrates how to expose tools and resources to AI agents and language models via the stdio transport.
The server manages a simple user database (backed by a local JSON file) and exposes endpoints that an AI agent can call to read or create user records.
Features
- Tool —
my-create-tool: Creates a new user by acceptingname,address,email, andphonefields and persisting the record tosrc/users.json. - Resource —
users://all: Returns the full list of users from the mock database as JSON. - Resource —
users://{userId}/profile: Returns the profile of a single user identified byuserId.
Prerequisites
Installation
npm install
Running the Server
Development mode (using tsx, no build step)
npm run server:dev
Build & run compiled output
npm run server:build # compiles TypeScript → JavaScript
node dist/server.js # run the compiled server
Watch mode (auto-recompile on changes)
npm run server:build:watch
Inspect with MCP Inspector UI
The MCP Inspector lets you interactively call tools and browse resources in a browser UI:
npm run server:inspect
Note:
server:inspectsetsDANGEROUSLY_OMIT_AUTH=true— use this only for local development.
Loading the MCP Server in Agent Environments
Agent CLI (e.g. Claude Code CLI)
Existence of .mcp.json should be picked up by the agent and prompted to load the server. Otherwise it's a simple MCP server config that needs to be added to the configuration with other servers.
Project Structure
my-mcp-server/
├── src/
│ ├── server.ts # MCP server entry point (tools & resources)
│ └── users.json # Mock user database (read/written at runtime)
├── package.json
└── tsconfig.json
Tech Stack
| Package | Role |
|---|---|
@modelcontextprotocol/sdk |
MCP server SDK (tools, resources, stdio transport) |
zod |
Input schema validation for tools |
tsx |
Run TypeScript directly without a build step |
typescript |
TypeScript compiler |
@modelcontextprotocol/inspector |
Browser-based MCP debug UI |
License
ISC
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。