🚀 MCP Server Generator from OpenAPI

Automatically generate production-ready Model Context Protocol (MCP) servers from OpenAPI/Swagger specifications.

Save hours of boilerplate coding. Just provide your API spec, and get a complete, type-safe MCP server ready to deploy.

Features

• ✅ OpenAPI 3.0 & Swagger 2.0 support
• ✅ TypeScript or Python output
• ✅ Type-safe tool definitions
• ✅ Authentication handling (API keys, OAuth, Bearer tokens)
• ✅ Automatic parameter validation
• ✅ Error handling built-in
• ✅ Documentation generated
• ✅ Ready to deploy - complete package with dependencies

Use Cases

Scenario	Benefit
API Integration	Turn any OpenAPI API into MCP tools in minutes
LLM Agents	Give AI agents access to your APIs instantly
Rapid Prototyping	Test MCP integrations without writing boilerplate
Multi-API Orchestration	Generate servers for multiple APIs and compose them

Quick Start

1. Find your API's OpenAPI spec (Swagger JSON/YAML URL)
2. Run this Actor with the spec URL
3. Download generated code from Key-Value Store
4. Deploy your MCP server

Input

{
  "openApiSource": "https://petstore.swagger.io/v2/swagger.json",
  "language": "typescript",
  "serverName": "petstore-mcp",
  "includeAllEndpoints": true,
  "includeAuth": true
}

Output

Complete MCP server with:

• server.ts or server.py - Main MCP server implementation
• package.json or requirements.txt - Dependencies
• README.md - Usage documentation

All files saved to Key-Value Store and dataset.

Example Generated Code

TypeScript:

server.setRequestHandler('tools/call', async (request) => {
    if (request.params.name === 'get_pet_by_id') {
        const response = await axios.get(`${API_BASE_URL}/pet/${request.params.arguments.petId}`);
        return { content: [{ type: 'text', text: JSON.stringify(response.data) }] };
    }
});

Python:

@server.call_tool()
async def get_pet_by_id(pet_id: int):
    async with httpx.AsyncClient() as client:
        response = await client.get(f"{API_BASE_URL}/pet/{pet_id}")
        return response.json()

Why MCP?

Model Context Protocol (MCP) is the standard protocol for connecting AI assistants to external tools and data sources. Adopted by Anthropic, Google, and OpenAI.

Building MCP servers manually requires:

• Understanding MCP protocol specifications
• Writing request/response handlers
• Type definitions for all endpoints
• Error handling
• Authentication logic

This Actor does all of that automatically.

Advanced Features

Tag Filtering

Only generate tools for specific API sections:

{
  "includeAllEndpoints": false,
  "filterTags": ["pets", "store"]
}

Authentication

Automatically handles:

• API Key authentication (header, query param)
• Bearer token authentication
• OAuth 2.0 flows (code generated, you add tokens)

Test Generation

{
  "generateTests": true
}

Includes unit test templates for all generated tools.

Technical Details

Parsing: Uses swagger-parser for robust OpenAPI validation and dereferencing

Code Generation: Handlebars templates for clean, maintainable output

Type Safety: Full TypeScript type definitions or Python type hints

Error Handling: Try-catch blocks with meaningful error messages

Standards Compliant: Follows MCP 1.0 specification exactly

Limitations

• Complex authentication flows may require manual enhancement
• Custom request/response transformations not supported
• Generated code is a starting point - customize for production

Built for Apify $1M Challenge

This Actor solves a real problem in the exploding MCP ecosystem. Every API with an OpenAPI spec can now become MCP-enabled in seconds.

---

Ready to generate your MCP server? Run now →