Desarrollo de agentes de IA con Claude Code

Programming has entered a new era. It's no longer enough to simply use artificial intelligence to generate code: the next step is creating agents capable of understanding a project, automating tasks, coordinating tools, and collaborating with the developer throughout the entire software lifecycle. In Developing AI Agents with Claude Code, you'll discover how to transform Claude Code into a true agent-based development platform: scalable, automated, and fully under your control. This book will teach you to go far beyond isolated prompts to design professional, reusable, and scalable workflows, both from the terminal and your development environment. Throughout the book, you'll learn how to customize Claude Code's behavior using slash commands, manage persistent memory and context, automate processes with hooks, and create assistants capable of maintaining project knowledge throughout its entire lifecycle. The book also delves into the Model Context Protocol (MCP), the standard that is transforming communication between agents, tools, and information sources. You will discover how its architecture works, how to configure MCP servers, and when it is most appropriate to use MCP, skills, or sub-agents depending on the needs of each project. Furthermore, you will learn to design multi-agent systems using specialized sub-agents, hierarchical delegation, parallel execution, and orchestration patterns, applying all these concepts in a real-world project developed with Next.js. The result is a methodology that allows you to scale AI-assisted programming without losing control over the code, software quality, or project maintainability.