Back to hands-on classes
Coming upIntermediateOnline

Learn how to work with MCP server

Master the Model Context Protocol (MCP) to connect LLMs to your local tools and data safely.

Schedule

August 8, 2026

Duration

2 Hours

Project

Hands-on capstone

Detailed Curriculum

3 practical sections built around live exercises.

01

MCP concepts and architecture

Understand MCP as a standard bridge between AI assistants, tools, and data.

Topics covered

  • Clients, servers, tools, and resources
  • When MCP is useful
  • Local vs remote servers
  • Common integration patterns

Hands-on lab

Diagram an MCP integration for a local project, database, or internal tool.

02

Building a custom MCP server

Create a small server that exposes useful functionality to an AI client.

Topics covered

  • Tool definitions
  • Resource discovery
  • Input validation
  • Error responses

Hands-on lab

Build a simple MCP server with one tool and one read-only resource.

03

Security and permissions

Connect tools to AI without giving away the whole machine.

Topics covered

  • Least privilege
  • RBAC patterns
  • Secrets handling
  • Auditability and approval flows

Hands-on lab

Add permission boundaries and a safety checklist to the example MCP server.

What You Get Out Of It

Concrete capabilities you should leave with.

Explain MCP and where it fits

Build a minimal MCP server

Expose tools and resources safely

Think clearly about AI tool permissions