Skip to content
Coming Soon

From Vibes
to Agents

How a non-engineer built enterprise software with AI — and the governance system that made it actually work.

Cover Coming Soon

From Vibes
to Agents

Ethan Owen

The Problem

You've been through the wizard phase. Now what?

The first two weeks of building with AI are magic. Features appear in hours. You feel invincible.

Then the system starts breaking.

The AI forgets what you told it yesterday. It claims features are complete when they aren't. It changes one file and breaks fifteen others. It guesses at database columns instead of checking. It builds three features on an assumption — and the assumption is wrong.

You're not imagining it. These aren't bugs in the model. They're structural failure modes that every AI model has. Better prompts won't fix them. A better model won't fix them.

You need a system.

The Book

Fourteen months. Forty-five incidents. One system.

From Vibes to Agents is the story of building that system — from the inside, under real pressure, with a real business depending on it.

Ethan Owen is a COO who runs a luxury furniture and fixtures company. He studied anthropology at Cornell. He's never been a software engineer. And he built an enterprise platform with AI agents governed by a framework he designed from scratch.

The book is the journey from chaos to structure. Every rule exists because something broke. Every verification type exists because a claim turned out to be false. Every automated hook exists because a manual check was forgotten.

Not theory. Practice. Battle-tested in production.

By The Numbers

The system, by the numbers

0
Canonical Rules
Canonical Rules
0
Incident Post-Mortems
Incident Post-Mortems
0+
Verification Types
Verification Types
0K
Words
Words
Inside the Book

What you'll learn

The 7 Failure Modes

Why ungoverned AI breaks — not occasionally, but structurally. The patterns that repeat until you build defenses against them.

Rules That Work

How to write rules AI will actually follow. Not suggestions. Enforceable law. The anatomy of a good rule vs. a useless one.

Memory That Persists

A system that makes AI learn from its mistakes across sessions. Corrections that stick. Context that carries forward. Permanently.

The Verification Mindset

50+ verification types that replace "the AI says it works" with proof. Commands, expected outputs, mechanical certainty.

The Incident Loop

How every failure makes the system permanently stronger. The mechanism that turns bugs into investments. The centerpiece of the framework.

Automated Enforcement

Hooks that run without human intervention — catching pattern violations, secrets exposure, and blast radius issues before code leaves your machine.

Planning Before Building

Why AI needs a blueprint. The blast radius principle. The plan-audit loop. A golden path from requirements to production.

The Knowledge Graph

64 structured tools that give AI a map of your codebase — not just files, but how they connect. Impact analysis, dependency tracing, and schema validation.

Tech Intelligence

A platform that turns article reviews into competitive advantage. 150+ structured reviews scored for relevance, generating 78 concrete implementation plans.

The Playbook

A maturity model from zero governance to full agentic engineering. Week-by-week plan to build your own system. Start the day you finish reading.

Who It's For

This book is for you if...

You're a solo founder building with AI and wondering why everything keeps breaking

You're a technical leader watching vibe coding create debt across your team

You're a non-engineer with domain expertise who wants to build real software — not prototypes

You're a CTO evaluating AI agents in production and asking "what's the governance model?"

You've been through the wizard phase and you're looking for what comes next

Get The Book

Get the book

The Book

$14.99one-time

The complete book in DRM-free PDF and ePub. Read it on any device, annotate it, share it with your team. Yours forever.

  • Full 16-chapter manuscript
  • 4 appendices with starter templates
  • PDF + ePub formats
  • DRM-free — yours forever
Best Value

Builder's Bundle

$59one-time

The book tells you what to build. The bundle gives you the building blocks.

  • Everything in The Book
  • 1 year of Massu Pro access
  • 60+ workflow commands
  • Rule templates & verification types
  • Hook configurations & pattern scanner
  • Standalone appendices as ready-to-use docs
  • Future book updates included

Includes 1 year of Massu Pro ($29/mo value). Auto-renews at $249/yr unless cancelled. Book is yours forever.

Team License

$199one-time

Roll out AI governance across your team. $20 per person for a complete implementation kit.

  • Everything in Builder's Bundle
  • Share with up to 10 team members
  • 1 year of Massu Team access
  • Team onboarding guide (5-page condensed framework)
  • Priority access to workshops

Includes 1 year of Massu Team ($499/mo value). Auto-renews at $4,990/yr unless cancelled. Books are yours forever.

Also available on Amazon Kindle ($9.99) and Amazon Paperback ($29.99). Need the full Massu platform? See plans →

About the Author

About the Author

EO

Ethan Owen is co-founder and COO of Limn, a luxury furniture and fixtures company that coordinates design and production between California, New York and Milan with factories in Indonesia. He studied cultural anthropology at Cornell and international policy at the Monterey Institute of International Studies.

He has held UX leadership roles at eBay, PayPal, and Visa, and has lived and worked in Germany, Singapore, and China.

In 2025, he started building enterprise software with AI and discovered that intelligence without governance produces impressive inconsistency. So he built the governance. The system he created became the foundation for Massu AI, an open-source governance platform for AI development.

FAQ

Questions

No. The author isn't one. The book is written for people with technical literacy — you understand what databases, APIs, and frontend code are — but who don't write production software for a living. Engineers will also get value, especially from the governance framework and incident loop.

The principles are tool-agnostic. The book was written using Claude Code, and most examples reference that workflow, but the governance framework works with any AI development tool — Cursor, Copilot, ChatGPT, or whatever comes next. The principles are about structure, not tooling.

Massu AI is the open-source platform that implements the governance framework. The book teaches the principles; Massu provides one implementation of those principles. You can build your own system from scratch using only the ideas in the book, or start with Massu and customize from there.

About 92,000 words across 16 chapters and 4 appendices. That's roughly 350 printed pages. Dense but readable — no filler, no fluff, every chapter earned its place.

The book has everything conceptually. The bundle adds 1 year of Massu Pro — the governance platform with 60+ workflow commands, rule templates, verification types, hook configurations, and the pattern scanner. Think of it as the difference between reading about a recipe and getting the mise en place pre-measured.

Amazon handles returns for Kindle and print editions per their standard policy. For the Builder's Bundle purchased directly, full refund within 30 days, no questions asked.

The model isn't the variable. The system is.

Start building yours today.

30-Day Money-Back Guarantee