This review is about Build an AI Agent (From Scratch) by Jungjun Hur and Younghee Song from Manning. Facts 10 chapters $29.29 (at the time of this writing) 315 pages Chapters Building your first LLM agent What is an AI agent? The brain of AI agents: LLMs Enabling actions: Tool use Implementing a basic ReAct agent Developing advanced agent capabilities Building knowledge bases with RAG Adding memory to your agent Planning and reflection for complex tasks Empowering agents with code execution Orchestrating multi-agent systems Evaluating agents Pros and cons At Manning, I consider the following two books to be references: Kubernetes in Action by Marko Lukša. It's the book that taught me the foundations of Kubernetes. Fun fact: I got it from the author at a conference in exchange for Swiss chocolate. API Design Patterns. I read it while working on the Apache APISIX API gateway. At this time, I was deeply involved in (RESTful) APIs. I had read a lot on APIs; I had researched a lot; I held talks on APIs; I talked with lots of people about APIs. And yet, the book was a revelation, full of brilliant insights. I think that Build an AI Agent is as fundamental to me as the other two were. While there are few chances that you will actually build an AI agent, this book teaches you the basics (and more) of how they operate. It starts with the agent loop, then proceeds to tools, then to MCP, etc. The only thing I would like to see for the next edition is for the developers to beef up the evaluating agents part. Conclusion If you use AI agents and want to understand what happens when you use them, this is the book for you. I think it should belong in the library of any engineer worthy of the name. Originally published at A Java Geek on September 6th, 2026.