Preface
This book aims to go deep into MCP, taking you beyond the available documentation and giving you a deep understanding of the MCP architecture, the features the protocol and its software development kits (SDKs) provide, and most importantly, how to use that understanding to build your own MCP servers and clients and augment your own generative AI workflows with them. By doing this, you can transform standard generative AI chatbots into fully fledged agents, which can act, plan, and use tools on their own, and even work together to accomplish complex tasks.
This book was written amidst a sea-change in how the MCP protocol and all its official SDKs worked. The 2026-07-28 protocol version upended many models of how the protocol worked, completely revamped how the Streamable HTTP transport worked, and made the protocol easier to use and more mature. Naturally, this posed a problem for your humble writer, who had nearly finished the book when the new protocol crystallized. When big changes happen like this, they’re often slow to socialize, and even though nothing is slow in the world of AI, I did my best to cover the current version of the protocol in-depth while also calling back to previous versions so that you can master the protocol no matter which era you are working with.
In the opening chapter, you will learn more about AI agents, what they are, how they are used, and how MCP fits into the agentic AI landscape. The next chapter will cover MCP itself: its history, the ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access