The Agentic Coding Playbook
Transform any product idea into a production-ready MVP in 6 weeks.
This book is the definitive playbook for building software products in the age of AI. It democratizes product development, enabling anyone—from vibecoders to CTOs—to build and ship software using agentic coding.
About This Book
Rather than teaching programming, this book teaches product building in an era where AI agents handle the mechanical coding work. Learn to combine systems thinking, product vision, design sensibility, and strategic judgment while AI agents handle implementation details.
Key Promise: Use proven patterns from battle-tested CTO experience to go from idea to production in 6 weeks.
Table of Contents
Book Structure
Part 1: Foundations - Core concepts including the Renaissance Developer mindset, what agentic coding is, architecture principles for AI-friendly systems, digestible interfaces, and understanding new bottlenecks when coding accelerates 5-10x.
Part 2: The Playbook - Complete 6-week workflow from idea to production: brainstorming, product brief, requirements in EARS notation, technical design, implementation planning, testing strategies, deployment pipelines, environment progression, managing AI agents, and post-MVP iteration.
Part 3: Patterns & Tools - Reusable patterns and tools including architecture patterns, specification patterns (EARS, OpenAPI, AsyncAPI), testing strategies for AI-generated code, and the agentic development toolchain (MCP servers, Claude Code skills, code review patterns).
Part 4: Complete Example - Real 6-week project documented end-to-end demonstrating all patterns and validating the timeline with actual code, decisions, and lessons learned.
Status
This book is currently in development. The infrastructure and build system are complete, and content is being actively written.
- Online Version: You’re reading it!
- PDF Version: Available in GitHub Releases
- Source Code: Available on GitHub
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
This book is being built using its own methodology - the planning artifacts, requirements, design documents, and implementation tasks are all available in the repository’s planning/ directory.