Testing MCP Servers With a Python MCP Client (Summary)

Testing MCP Servers With a Python MCP Client Rodrigo Girão Serrão 01:36

You’ve learned how to build a basic command-line Model Context Protocol (MCP) client in Python. Your client can connect to MCP servers, discover their capabilities, and interact with their tools, prompts, and resources directly from your terminal.

You started by setting up your project environment and structuring a robust client class using asynchronous context management and the MCP Python SDK. Next, you added a command-line interface that lets you point the client at any server, along with subcommands to list a server’s capabilities and to call tools, get prompts, and fetch resources.

In this video course, you’ve learned how to:

Set up and structure a Python project using uv and pyproject.toml
Implement an asynchronous MCP client with the MCP Python SDK
Build a command-line interface (CLI) for your client
List and inspect MCP server capabilities, including tools, prompts, and resources
Call tools, get prompts, and fetch resources from an MCP server with dedicated subcommands

With these skills, you can now quickly test MCP servers directly from your terminal. You can also extend your client app with additional features as needed.

Locked learning resources

Join us and get access to thousands of tutorials and a community of expert Pythonistas.

Unlock This Lesson

Already a member? Sign-In

Locked learning resources

The full lesson is for members only. Join us and get access to thousands of tutorials and a community of expert Pythonistas.

Unlock This Lesson

Already a member? Sign-In

00:00 Congratulations for making it this far. After working through this course, now you can tell the difference between MCP servers and clients, and you understand which is which.

00:10 You can also write MCP clients that interact with servers, and in particular, you can test your own servers deterministically by using your MCP client that provides a CLI that allows you to,

00:29 among other things, get prompts by name, fetch resources by URI, and call tools, passing in arbitrary inputs. This is a pretty big technical feat, so congratulations, pat yourself on the back, you deserve it, and if required, you’re also invited to take another look at the code, work through it, tear it apart, play with it, to make sure you understand it as much as possible.

00:50 If you would like to keep exploring these topics, Real Python has a tutorial on building your own MCP server, and this is the tutorial that the MCP data server you’ve been using is based off of, and there’s also a tutorial on building a Python MCP client.

01:07 It’s also the basis for this course, but the written tutorial takes a fairly different direction in how to test the servers, so it’s a very interesting read that will complement this course very nicely.

01:21 Real Python provides a Python Coding With AI learning path that will take you through the skills and tools that you need to master in this day and age. Thank you so much for watching.

01:32 My name is Rodrigo, and I’ll catch you on the next one.

Become a Member to join the conversation.