cod-api MCP Server
Exposes REST API endpoints defined in an OpenAPI Specification as MCP tools, allowing AI models to call them via the ModelContext Protocol.
README
ModelContext Protocol Server
This project demonstrates how to use the ModelContext Protocol (MCP) with an Express server.
The MCP server methods were generated based on an OpenAPI Specification (OAS) file and integrated into the src/index.ts file.
Installation and Setup
Prerequisites
Before running the project, ensure the following are installed on your system:
- Node.js (version >= 18.0.0)
- npm (comes with Node.js)
Steps to Install and Run
- Navigate to the project directory.
- Install the dependencies:
npm install - Start the development server:
npm run dev - The MCP server runs by default at:
http://localhost:4000/mcp
Communicating with the MCP Server
The MCP server is running as a Streamable HTTP server and listens for requests at the default endpoint:
http://localhost:4000/mcp
Sending Requests to the MCP Server
To communicate with the MCP server, you can send HTTP POST requests to the /mcp endpoint. The server expects the request body to follow the JSON-RPC 2.0 specification.
Example Request
Here is an example of a JSON-RPC 2.0 request to invoke a tool or resource:
{
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"name": "echo",
"arguments": {
"message": "Hello, MCP!"
}
},
"id": 1
}
- method: Specifies the tool or resource to invoke (e.g., tool/echo).
- params: Contains the parameters required by the tool or resource.
- id: A unique identifier for the request
The server will respond with a JSON-RPC 2.0-compliant response:
{
"jsonrpc": "2.0",
"result": {
"content": [ {"type":"text","text":"Tool echo: Hello, MCP!"} ]
},
"id": 1
}
Using cURL to Send Requests
You can use cURL to send requests to the MCP server. Here is an example command:
curl -X POST http://localhost:4000/mcp \
-H "Content-Type: application/json" \
-H "Accept: application/json, text/event-stream" \
-d '{
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"name": "echo",
"arguments": {
"message": "Hello, MCP!"
}
},
"id": 1
}'
Server customization
Port
If you want to run the MCP server on a different port, you can set the MCP_SERVER_PORT environment variable before starting the server. For example:
export MCP_SERVER_PORT=4002
API url
By default, MCP Server is using first server url defined in the OpenAPI Specification (OAS) file.
In order to customize API URL for the MCP server, please set the MCP_API_URL environment variable:
export MCP_API_URL=<desired_url>
MCP server methods
If you need to modify the MCP server methods, you can do so in the src/index.ts file.
The methods are generated based on the OpenAPI Specification (OAS) file, and you can adjust them as needed.
Troubleshooting
If the project is not running as expected:
- Ensure all dependencies are installed by running
npm install - Verify that you are using Node.js version 18 or higher:
node -v - Check for errors in the terminal output when running
npm run dev - Ensure the port (default: 4000) is not already in use by another application.
- Verify custom environment variables:
echo $MCP_SERVER_PORT echo $MCP_API_URL - If the issue persists, review the code in
src/index.tsfor potential misconfigurations or errors.
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。