Getting Started
This guide walks you through creating and deploying your first MCP server on MCP Registry.
Prerequisites
Before you begin, make sure you have:
- An MCP Registry account (sign up here)
- An API you want to expose to AI assistants
- Basic understanding of REST APIs
Step 1: Create a New Server
After logging in, navigate to your dashboard and click Create Server.
You’ll need to provide:
- Server Name: A unique identifier for your server
- Description: What your server does (this helps AI understand when to use it)
- Base URL: The root URL of your API
Step 2: Define Your Tools
Tools are the individual functions that AI assistants can call. For each tool, you’ll define:
- Name: A clear, descriptive name (e.g.,
get_weather,send_email) - Description: What the tool does and when to use it
- Parameters: Input parameters with their types and descriptions
- Endpoint: The API endpoint to call
Example Tool Definition
{
"name": "get_user",
"description": "Retrieves user information by their ID",
"parameters": {
"user_id": {
"type": "string",
"description": "The unique identifier of the user",
"required": true
}
},
"endpoint": "/users/{user_id}",
"method": "GET"
}
Step 3: Configure Authentication
Set up how your API authenticates requests:
- API Key: Include an API key in headers
- Bearer Token: Use OAuth2 bearer tokens
- None: For public APIs (not recommended)
Step 4: Test Your Server
Use the built-in testing tool to verify your server works correctly:
- Select a tool from your server
- Enter test parameters
- Click Run Test
- Verify the response
Step 5: Publish
Once testing is complete, click Publish to make your server available. Your API is now accessible to AI assistants!
Next Steps
- Learn about authentication in detail
- Explore best practices for tool design
- Check out example servers for inspiration