Skip to content

Quick Start

Generate a fully-functional MCP server from an OpenAPI spec in under a minute.

0. Install

pip install mcp-generator

1. Generate

# Default: reads ./openapi.json
generate-mcp

# Or specify a file
generate-mcp --file ./my-api-spec.yaml

generate-mcp --url https://petstore3.swagger.io/api/v3/openapi.json

What gets generated:

  • ✅ Python API client from the OpenAPI spec
  • ✅ Modular MCP server modules (one per API tag)
  • ✅ Authentication middleware
  • ✅ OAuth2 provider (if spec contains OAuth2 security schemes)
  • ✅ Package files, tests, and Docker configuration
  • ✅ Everything outputs to generated_mcp/

Optional Features

By default, the generator creates a minimal, production-ready server. Enable additional features as needed:

# Enable persistent storage (for OAuth tokens, session state)
generate-mcp --enable-storage

# Enable response caching (reduces backend API calls)
generate-mcp --enable-storage --enable-caching

# Enable MCP resources (expose GET endpoints as MCP resources)
generate-mcp --enable-resources

# Enable all features
generate-mcp --enable-storage --enable-caching --enable-resources
Flag Description When to Use
--enable-storage Persistent storage backend OAuth refresh tokens, session data
--enable-caching Response caching with TTL Rate-limited APIs, slow endpoints
--enable-resources MCP resource templates Expose API data for context/retrieval

Note

--enable-caching requires --enable-storage as it uses the storage backend for cache persistence.

2. Register

# Register the generated server
register-mcp ./generated_mcp

# Verify registration
run-mcp --list

This adds your server to the local registry at ~/.mcp-generator/servers.json.

3. Run

export API_TOKEN="your-api-token-here"
run-mcp swagger_petstore_openapi

run-mcp swagger_petstore_openapi --mode http --port 8000

cd generated_mcp
python swagger_petstore_openapi_mcp_generated.py --transport stdio

cd generated_mcp
uv run fastmcp run swagger_petstore_openapi_mcp_generated.py:create_server
# Or with config:
uv run fastmcp run fastmcp.json

4. Connect to AI Clients

Claude Desktop

Add to ~/.claude/claude_desktop_config.json:

{
  "mcpServers": {
    "my-api": {
      "command": "python",
      "args": ["/path/to/generated_mcp/swagger_petstore_openapi_mcp_generated.py"],
      "env": {
        "API_TOKEN": "your-api-token-here"
      }
    }
  }
}

HTTP Mode

Any MCP-compatible client can connect to:

http://localhost:8000/mcp

Next Steps