Quickstart Guide

Prerequisites

Step 1: Install MCPify CLI

Install the MCPify CLI tool to manage your integrations from the command line.

npm install -g @mcpify/cli

yarn global add @mcpify/cli

Step 2: Initialize Your Integration

Create a new MCPify configuration for your API. The CLI will guide you through the setup.

mcpify init my-api

This creates a mcpify.json configuration file:

{
  "name": "my-api",
  "version": "1.0.0",
  "description": "My API integration for AI agents",
  "gateway": {
    "url": "https://gateway.mcpify.org",
    "region": "us-east-1"
  },
  "source": {
    "type": "openapi",
    "url": "https://api.example.com/openapi.json"
  },
  "auth": {
    "type": "oauth2",
    "flow": "client_credentials",
    "tokenUrl": "https://api.example.com/oauth/token",
    "scopes": ["read", "write"]
  },
  "tools": []
}

Step 3: Import Your API

MCPify can automatically import your API from various sources:

mcpify import openapi https://api.example.com/openapi.json

mcpify import graphql https://api.example.com/graphql

mcpify add-tool --name "list_users" --endpoint "/users" --method GET

Step 4: Configure Authentication

Securely store your API credentials in MCPify's encrypted vault:

mcpify auth configure

The CLI will prompt for your credentials based on the auth type:

• • API Key: Enter your API key
• • OAuth2: Client ID and Client Secret
• • Basic Auth: Username and Password

Step 5: Deploy to MCPify Gateway

Deploy your configuration to make it available to AI agents:

mcpify deploy

Step 6: Test Your Integration

Test your MCP service with the built-in testing tool:

mcpify test list_users

Or connect it to Claude Desktop for interactive testing:

# Add to Claude Desktop config
{
  "mcpServers": {
    "my-api": {
      "url": "https://gateway.mcpify.org/services/my-api",
      "apiKey": "YOUR_MCPIFY_API_KEY"
    }
  }
}

Using with AI Agents

from anthropic import Anthropic

client = Anthropic()
response = client.messages.create(
    model="claude-3-opus",
    mcp_servers=[{
        "url": "https://gateway.mcpify.org/services/my-api"
    }],
    messages=[{"role": "user", "content": "List all users"}]
)

from openai import OpenAI

client = OpenAI()
response = client.responses.create(
    model="gpt-5",
    tools=[{
        "type": "mcp",
        "server_url": "https://gateway.mcpify.org/services/my-api"
    }],
    input="Get the list of active users"
)

Next Steps