Colophon · v1.3.1

About

A living handbook for founders and product engineers building when AI changes every step — not another tool tutorial, but decisions that survive the next model release.

Start here

You do not need to read eighteen chapters in order. Pick an entry point:

Not sure? Try Solo founder if you are validating alone, orProduct engineer if you already own a codebase.

What each chapter is

Every chapter runs the same loop — judgment with a trail, not a hot take:

  1. Principle — the durable claim
  2. Decision — when to do X vs Y (also in the registry)
  3. Workflow — steps you can run this week
  4. Artifact — template, checklist, or test that makes the decision re-readable

Tool names appear last and are allowed to age. The frameworks are what you keep.

Maturity

Chapters are labeled so progress means depth, not just existence. Right now: 3 refined, 15 shipped,18 total.

shipped
Readable v1 — full loop present, ready to use.
refined
Depth pass — composite scenario, falsifiers, tighter artifacts.
battle-tested
Used on real bets with revision after signal (reserved for later).

Bumps are tracked in the changelog. This is software-shaped writing: ship, gather signal, revise.

Author

Written by Angel Guirao — product-minded engineer who has shipped at venture scale and cared about backends that do not wake you at 3am.

This handbook sits in the public stack besidemesh (work and writing) andfootnotes (distilled principles). Sharp ideas from chapters may spin out to footnotes when they deserve a shorter essay.

Source & license

Chapters live as MDX in GitHub. MIT — fork, cite, teach. Keep the decision frameworks; replace the tooling with whatever ships next month.

git clone https://github.com/Angelguirao/ai-native-product-building.git

Templates for artifacts live in templates/ in the repo (validation experiments, migration checklists, agent boundaries, ADRs).

Influences

Essay rhythm from Making Software, catalog spine from AI Engineering from Scratch, warm voice from footnotes — adapted for people who own product outcomes, not just prompts.