Agentic Coding with Claude Code: Delegate Real Software Engineering Tasks to an AI Agent: Context, Planning, Testing, and Git Workflows That Ship Production-Ready Code - Tapa blanda

Sinopsis

Stop using Claude Code like a smarter autocomplete. Start using it like an engineering agent.

Agentic Coding with Claude Code is a practical guide for developers who want to delegate real software engineering tasks to Claude Code while maintaining control over quality, security, and what actually ships.

Instead of focusing on clever prompts, this book teaches the workflow behind reliable agentic coding: giving an AI coding agent the right context, defining work precisely, establishing guardrails, delegating implementation, testing results, reviewing changes, and safely shipping production-ready code.

Inside this book, you'll learn how to:

• Understand the difference between an AI assistant and an AI coding agent.


• Set up Claude Code and run your first complete agentic development session.


• Use CLAUDE.md and project memory to give Claude Code the context it needs.


• Turn vague requirements into clear, agent-ready development tasks.


• Configure permissions, guardrails, and appropriate levels of autonomy.


• Delegate real feature development instead of limiting AI to small code snippets.


• Use testing, TDD, debugging, and verification to validate AI-generated code.


• Use Git to review AI-generated diffs, control scope creep, and recover from mistakes.


• Scale development with subagents, parallel sessions, Git worktrees, and custom skills.


• Connect Claude Code to external tools and services using MCP.


• Protect agentic workflows against prompt injection, excessive permissions, and unsafe autonomy.


• Apply the complete workflow to a real feature from understanding and planning through implementation, testing, debugging, review, and shipping.

This book is for developers who already know how to code and want to become more effective at working with AI agents on real-world software projects. You will learn not simply how to make Claude Code generate code, but how to direct, supervise, verify, and ship work produced by an AI coding agent.

Whether you're building a personal project, maintaining a production codebase, or developing repeatable AI-powered software workflows, this book provides a practical system for turning agentic coding into a disciplined engineering practice.

Learn the workflow. Delegate the work. Verify the result. Ship with confidence.

"Sinopsis" puede pertenecer a otra edición de este libro.