Creating an MCP tool and using it in Claude

Please RSVP ahead of time and bring a government-issued ID for check-in.

In this session, we'll learn the basic steps to build a custom Model Context Protocol (MCP) tool in Python and connect it to Claude, enabling us to extend the LLM’s capabilities with our own logic and applications. We’ll go step by step through defining a simple MCP server, implementing tool functions, and integrating them into Claude’s context so you can call your own APIs or workflows directly within the chat. By the end, you’ll understand how to make Claude interact seamlessly with external systems, turning static conversations into dynamic applications. We'll explore these concepts using the example of resume customization for job descriptions. Whether you’re experimenting with automation, custom agents, or enhanced chat experiences, this guide will give you a practical foundation for connecting your tools to an LLM through MCP.

About the Speaker
Nazar Mammedov: Web Specialist at Northeastern University. Full-stack software developer. AWS Certified Solutions Architect.
Venue Sponsor:
CIC Cambridge
https://cic.com/about/