This intermediate path teaches you how to use the Model Context Protocol (MCP) to connect AI agents with external tools, data sources, and services. You will work in TypeScript as you build servers that expose tools, resources, and prompts for agents to discover and use. You will start by learning how agents run, manage conversation flow, and call built-in or custom tools. From there, you will create MCP servers with different transport options, manage stateful sessions, and design practical capabilities such as calculators or task-specific assistants. The path also covers production-oriented integration patterns, including connecting one agent to multiple MCP servers, caching tool information, integrating servers with web frameworks, and adding authentication middleware. It is designed for developers who already have TypeScript experience and want to build more capable agent-based applications.
