Claude Code Projects

Master Claude Code through 11 hands-on projects, with a companion GitHub repo containing demos and code, downloadable assets, and reusable SKILL.md files for professional AI-assisted development.

Key Features:

- Build 11 working projects using a repeatable plan, implement, test, review, commit loop

- Create reusable Claude Code workflows with CLAUDE.md, Skills, hooks, MCP, and subagents

- Test, refactor, review, and ship AI-built software with Git and production-readiness checks

Book Description:

Claude Code can generate code quickly, but professional development demands more than accepting the first answer. You need a workflow for planning, testing, reviewing, constraining, and shipping AI-assisted changes with confidence.

Claude Code Projects teaches that workflow through 11 hands-on projects built around one repeatable loop: plan, implement, test, review, commit. You'll start with effective prompting and CLAUDE.md project context, then build and compare API implementations, generate persistent tests, debug AI-written code, and use Git as a safety net for every change.

As the projects progress, you'll turn wireframes into working interfaces, refactor legacy code without changing behavior, and package repeatable workflows as Claude Code Skills. You'll extend Claude Code with hooks, MCP, subagents, worktrees, and plugins while applying least-privilege access and safe automation.

Finally, you'll assess AI-built software for production readiness using evidence, guardrails, rollback thinking, and decisive go/no-go criteria. A capstone brings the full workflow together as you build a new API from scratch, test it, review it, assess readiness, and ship it through a pull request.

What You Will Learn:

- Use the plan, implement, test, review, commit development loop

- Write precise prompts with goals, constraints, outputs, and examples

- Create lean CLAUDE.md files that guide project behavior

- Compare AI-generated solutions with tests and scoring rubrics

- Test, debug, refactor, and review AI-generated code safely

- Build reusable Skills with scripts, inputs, and permissions

- Extend Claude Code with hooks, MCP, subagents, and plugins

- Run evidence-based production-readiness reviews before shipping

Who this book is for:

This book is for developers, vibe coders, tech leads, DevOps engineers, and automation engineers who want to use Claude Code in professional software development. It is especially useful for readers moving beyond ad hoc prompting into repeatable workflows for testing, Git, refactoring, Skills, MCP, multi-agent development, and production of readiness. Basic programming and Git familiarity are helpful; the book explains unfamiliar commands and concepts as they appear.

Table of Contents

- Getting Started with Claude Code and the AI Coding Loop

- Prompting Like a Tech Lead with GCOE

- Giving Claude Project Context with CLAUDE.md

- Building Faster with Best-of-N

- Testing, Debugging and Self-Review of AI-Generated Code

- Safe Git Workflows for AI-Assisted Development

- Multimodal Development: From Wireframe to Working UI

- Constrained Refactoring and Documentation from the Diff

- Packaging Your Workflows as Reusable Skills

- Extending Claude Code with Hooks, MCP and Multi-Agent Workflows

- Production Readiness Reviews for AI-Built Software

- From Workshop to Workplace: Habits, Anti-Patterns and a Capstone Project