mcpField
O
OpenAPI to MCP server Generator

OpenAPI to MCP server Generator

Local
github.com/vincent-pliDeveloper Tools1Updated 77w ago

About OpenAPI to MCP server Generator

OpenAPI to MCP server Generator is a command-line tool that creates MCP server projects from OpenAPI specifications. It generates tools for API endpoints and the files needed to run the server.

Key features

  1. 01Generate a tool for each API endpoint
  2. 02Support nested $ref references
  3. 03Generate server configuration and project files
  4. 04Configure generated servers with environment variables
  5. 05Send log notifications to clients

Use cases

  1. 01Create an MCP server from a local OpenAPI specification
  2. 02Generate a server project from a remote OpenAPI URL

MCP support

Generates MCP server code that exposes OpenAPI endpoints as tools. Generated servers support stdio transport and can be configured for MCP clients.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

No install method listed yet

We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.

Repository

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

OpenAPI to MCP server Generator

A command-line tool that generates Model Context Protocol (MCP) server code from OpenAPI specifications. This tool helps you quickly create an MCP server that acts as a bridge between LLMs (Large Language Models) and your API.

npm version License: MIT

English | 简体中文

At the beginning

This repo is originally forked from openapi-mcp-generator, and add some additional features:

  • Support nested $ref in openapi specifications
  • Besides source code, generate MCP server configuration
  • Allow client to set log level and send log message to client as notification
  • When hit error, send message to stderr
  • Support build docker image and guide client to run in docker container (2025/5/8 updated)

Features

  • Automatic Tool Generation: Converts each API endpoint in your OpenAPI spec into an MCP tool
  • Transport Options: Only supports stdio, for sse you can leveral mcp-proxy
  • Complete Project Setup: Generates all necessary files to run an MCP server
  • Easy Configuration: Simple environment-based configuration for the generated server

Installation

# Install globally from npm
npm install -g openapi-mcpserver-generator

# Or with yarn
yarn global add openapi-mcpserver-generator

# Or with pnpm
pnpm add -g openapi-mcpserver-generator

Usage

Generate an MCP server from an OpenAPI specification:

openapi-mcpserver-generator --openapi path/to/openapi.json --output /Path/to/output
Command Line Options
OptionAliasDescriptionDefault
--openapi-oPath or URL to OpenAPI specification(required)
--output-dOutput directory for generated files./mcp-server
--name-nName for the MCP serveropenapi-mcp-server
--version-vVersion for the MCP server1.0.0
--transport-tTransport mechanism (stdio, websocket, http)stdio
--help-hShow help information
Examples

Generate from a local OpenAPI file:

openapi-mcpserver-generator --openapi ./specs/petstore.json --output ./petstore-mcp

Generate from a remote OpenAPI URL:

openapi-mcpserver-generator --openapi https://petstore3.swagger.io/api/v3/openapi.json --output ./petstore-mcp

Generated Files

The tool generates the following files in the output directory:

  • server.js - The main MCP server implementation
  • package.json - Dependencies and scripts
  • README.md - Documentation for the generated server
  • .env.example - Template for environment variables
  • types.d.ts - TypeScript type definitions for the API
  • tsconfig.json - TypeScript configuration
  • Dockerfile - Dockerfile
  • .dockerignore - Docker ignore file

Using the Generated Server

After generating your MCP server:

  1. Navigate to the generated directory:

    cd my-mcp-server
    
  2. Install dependencies:

    npm install
    
  3. Create an environment file:

    cp .env.example .env
    
  4. Edit .env to set your API base URL and any required headers:

    API_BASE_URL=https://api.example.com
    API_HEADERS=Authorization:Bearer your-token-here
    
  5. Start the server:

    npm start
    

Requirements

  • Node.js 16.x or higher
  • npm 7.x or higher

E2E example

Suggest use mcpclihost as MCP host to take a try. This tool(mcpclihost) could support both Azure Openai and deepseek

You can add generated MCP server congiguration like this:

{
  "mcpServers": {
    "petstore-mcp": {
      "command": "/usr/local/bin/node",
      "args": [
        "/Users/lipeng/workspaces/github.com/vincent-pli/openapi-mcpserver-generator/petstore-mcp/server.js",
        "run"
      ]
    }
  }
}

to the ~/.mcp.json(default mcp server configuration path of mcpclihost), then take a try

Security Schemes in Openapi

Openapi 3.0 support 4 security types:

  • apiKey: for example:
        "securitySchemes": {
            "my_api_key": {
                "type": "apiKey",
                "name": "api_key",
                "in": "header"
            }
        }

Expect a env param named upper cased MY_API_KEY_{securitySchemes.my_api_key.name}, in this case, it should be: MY_API_KEY_API_KEY defined in .env

  • http:
        "securitySchemes": {
            basicAuth: {
               type: "http",
               scheme: "basic"
            }
        }

it try to find BASICAUTH_USERNAME and BASICAUTH_PASSWORD in .env

        "securitySchemes": {
            basicAuth: {
               type: "http",
               scheme: "bearer"
            }
        }

it try to find BASICAUTH_BEARERTOKEN in .env

  • oauth2: Because of the complexity of oauth2, cannot handle it automaticly, we suggest manually get the access token, then set it to .env as this:
API_HEADERS=Authorization:Bearer your-access-token-here
  • openIdConnect Not support yet

License

Apache 2.0

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.