zaizaizhao/mcp-swagger-server

Overview

This tool converts any REST API that follows the OpenAPI/Swagger specification into a Model Context Protocol (MCP) server. It parses the API spec and generates MCP tool definitions, allowing AI assistants to discover and call the API endpoints. The server is written in TypeScript and is open source.

Best for

Best for
Developers who want to quickly connect their REST APIs to AI assistants using the Model Context Protocol

Use cases

• Expose existing REST APIs to AI assistants for automated interactions
• Build AI agents that can call external services via natural language commands
• Prototype integrations between AI models and backend systems using OpenAPI specs

How to use

Install

npm i mcp-swagger-server -g

Tools exposed

• node
• pnpm

Tested with

Claude Desktop

Example client config

{\n  "mcpServers": {\n    "swagger-converter": {\n      "command": "mss",\n      "args": [\n        "--openapi", "https://petstore.swagger.io/v2/swagger.json",\n        "--transport", "stdio"\n      ]\n    },\n    "secured-api": {\n      "command": "mss",\n      "args": [\n        "--openapi", "https://api.example.com/openapi.json",\n        "--transport", "stdio",\n        "--auth-type", "bearer",\n        "--bearer-env", "API_TOKEN"\n      ],\n      "env": {\n        "API_TOKEN": "your-bearer-token-here"\n      }\n    }\n  }\n}

Notes

This tool converts any REST API that follows the OpenAPI/Swagger specification into a Model Context Protocol (MCP) server. It parses the API spec and generates MCP tool definitions, allowing AI assistants to discover and call the API endpoints. The server is written in TypeScript and is open source.

67 stars on GitHub. Last updated 2026-04-19. Licensed MIT.

Use cases

• Expose existing REST APIs to AI assistants for automated interactions
• Build AI agents that can call external services via natural language commands
• Prototype integrations between AI models and backend systems using OpenAPI specs

Pros

• Works with any OpenAPI-compliant REST API without manual configuration
• Open source and written in TypeScript, easy to extend or audit
• Reduces the effort needed to make APIs accessible to AI assistants

Cons

• Requires a valid OpenAPI specification for the target API
• Limited community adoption (67 stars) may mean fewer examples or support
• May not handle non-standard API behaviors or complex authentication flows

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.