Simple MCP POC
A proof-of-concept MCP server that enables reading local files and performing basic arithmetic operations. It provides a simple foundation for understanding how tools are exposed to MCP clients.
README
Simple MCP POC
A simple Model Context Protocol (MCP) server that exposes tools for:
- Reading local files
- Performing basic arithmetic operations
This project is intended as a proof of concept (POC) to understand how MCP servers work and how tools are exposed to clients.
Prerequisites
- Node.js v18 or higher
- npm (comes with Node.js)
node -v npm -v
---
## Installation
1. Clone or copy this repository
2. Open a terminal in the project directory
3. Install dependencies:
```bash
npm install
Running the Server
Start the MCP server using standard I/O (stdio):
node server.js
The server will start and wait for JSON-RPC messages on stdin/stdout.
Testing with MCP Inspector
The easiest way to test the server is using the MCP Inspector tool.
Windows (PowerShell)
$env:DANGEROUSLY_OMIT_AUTH="true"
npx mcp-inspector node server.js
macOS / Linux
DANGEROUSLY_OMIT_AUTH=true npx mcp-inspector node server.js
After running, open your browser to http://localhost:6274 to interact with the tools.
Available Tools
The server exposes two tools:
1. read_file
Reads the content of a local text file.
Arguments:
path(string): The relative or absolute path to the file.
Example usage:
{
"jsonrpc": "2.0",
"id": 1,
"method": "tools/call",
"params": {
"name": "read_file",
"arguments": {
"path": "notes.txt"
}
}
}
2. calculator
Performs basic arithmetic operations.
Arguments:
a(number): First operandb(number): Second operandoperation(string): One ofadd,subtract,multiply,divide
Example usage:
{
"jsonrpc": "2.0",
"id": 2,
"method": "tools/call",
"params": {
"name": "calculator",
"arguments": {
"a": 10,
"b": 5,
"operation": "add"
}
}
}
Manual Testing (JSON-RPC)
You can also test the server by sending raw JSON-RPC messages to the running server process.
Example: Read notes.txt
# Start server
node server.js
# In another terminal, send JSON-RPC request
echo '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"read_file","arguments":{"path":"notes.txt"}}}' | node server.js
Project Structure
mcp-poc/
├── server.js # Main MCP server entry point
├── package.json # Project dependencies
└── README.md # This file
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。