Testing your first MCP server

Learn how to test your MCP server using the Explorer tool

Now that you’ve created your first MCP server, it’s time to test its functionality. This guide will walk you through the steps to interact with your server and debug its responses.

What you’ll learn:

• How to test an MCP server

Before you start:

• Complete the Introduction guide
• Complete the Deploying your first MCP server guide

Testing an MCP server

MCP servers in Metorial provide tools, resources, and capabilities to AI assistants. In this guide, we will pick up on the server deployed in the last step.

1. Open the Server Page

   We start at the end of the last guide. Navigate to the overview page for your deployed GitHub MCP server.

   
2. Open the Explorer

   Click the Open Explorer button to open an interactive tool that lists all the MCP server’s capabilities.

   
3. Test Server Capabilities

   Select the Repository Content resource template from the available options. Enter the repository owner, repository name, and the file path you want to target. These settings define the list and method information that your AI integrations will invoke automatically.

   
4. View Result

   View the resulting response from the MCP server to see the data your AI integrations will consume.

   
5. Explore Available Tool Calls

   Navigate to the Tools section to browse available tool calls. Select the tool call you want to try out—here, it’s the fork_repository call. With the GitHub MCP server and a properly scoped Personal Access Token, your AI integrations can invoke any of these calls.

   

What’s Next?

🚀 You’ve successfully deployed and tested your first MCP server. Next, you’ll learn how to grant your AI integrations secure access to your MCP servers.

Next Up: How to use an MCP server

Learn how to use a deployed MCP server using SDKs.