Model Context Protocol (MCP): Hands-On with Agentic AI

30mGeneral2025-03-24

Authors

Morten Rand-Hendriksen

Senior Staff Instructor, Speaker, Web Designer, and Software Developer

Course details

The Model Context Protocol (MCP) allows developers to add agent behavior to LLMs by providing a universal protocol providing context to language models so they can interface with data and applications in a consistent way. MCP servers expose resources (data), tools (actions), and prompts (instructions) for the LLM and the user to use in performing more complex operations. In this course you’ll explore how the MCP works in Claude Desktop to extend its functionality, and you’ll build your own MCP servers using Python and TypeScript to give LLMs new capabilities to do things on the computer, connect with external APIs, and perform advanced multi-step actions.

Learning objectives
Define and understand Model Context Protocol (MCP).
Discover available MCP servers.
Install and use MCP servers in Claude Desktop.
Install and use MCP servers in Cursor.
Build custom MCP servers in Python and Typescript.

Skills covered

ClaudeAnthropicAPIsAI Productivity ToolsGenerative AIArtificial Intelligence for BusinessArtificial Intelligence (AI)Business Software and ToolsSoftware DevelopmentOne-Off

Concepts

0. Introduction

01 - MCP - Connecting AI agents to data, apps, and more

1. MCP Explained

02 - Using MCP servers in Claude Desktop
03 - Model Context Protocol (MCP) explained
04 - Exploring avaialble MCP servers and clients
05 - Limiting the blast radius of AI agents
06 - Leveraging the power of MCP servers
07 - Using MCP servers in Cursor

2. Building MCP Servers

08 - Building your own MCP servers
09 - Testing with the MCP inspector
10 - Troubleshooting MCP servers

3. Conclusion

11 - Building AI agents with MCP