GumboGumbo

Gumbo curriculum + package system

AI Engineering Evolution

A Gumbo-built operating system for AI-native engineering teams: curriculum, templates, starter skills, workflow patterns, downloadable assets, and the judgment layer that helps teams move from scattered AI wins to a durable way of building.

6

Curriculum modules covering bottlenecks, planning, memory, risk, org design, and compounding follow-through.

10+

Named package artifacts across templates, starter-pack installables, workflow examples, and presentation surfaces.

3

Download surfaces to ship now: guide preview, machine-readable manifest, and a preview bundle.

1

Finder-style library teaser that makes the package feel like a real product instead of abstract bullets.

What it is

AI changes the bottleneck. This changes the operating model.

When AI starts working, code gets cheaper first. The next pain moves into planning quality, review burden, release confidence, role clarity, and shared memory. AI Engineering Evolution gives teams a way to name that shift, teach it, and install better habits around it.

Curriculum

A six-module Gumbo curriculum for teaching teams how AI changes engineering bottlenecks and what has to change in response.

Starter operating kit

Governance files, templates, starter skills, workflows, and config examples that teams can actually adapt and use.

Upgrade path

The package stands on its own, and it also creates a clean path into Gumbo's guided and evolved engineering engagements.

What's inside

The box has ideas, installables, and story.

A lot of AI packages fall apart because they only ship one layer. This system ships four: the operating thesis, the working files, the long-form guide, and the polished surfaces that make it easy to understand and share.

01

Thinking layer

Thesis note, curriculum, module logic, and the follow-through model that explains how teams move from AI wins to AI operations.

02

Operating layer

Templates, starter governance files, repo-local skills, workflow templates, and risk/reviewer config examples.

03

Library layer

A previewable file system for curriculum, templates, workflows, and assets — designed to support future digital downloads and gated products too.

04

Presentation layer

A Gumbo-branded marketing site and long-form guide surface that makes the package feel like a real product, not a zip file.

Field guide

A real walkthrough for how teams move from AI wins to AI operations.

The current field guide should become a substantive companion, not just an overview page. This surface previews the larger book: a plainspoken walkthrough of the bottleneck shift, the control plane, the package, and the operating changes teams actually have to make.

Chapter 1

Why AI Engineering Evolution exists

Explains why faster code generation is only the first effect, and why the real operating challenge shows up in planning, review, release, context, and ownership.

Frames the core thesis: AI changes the bottleneck before it changes the org chart.

Chapter 2

The bottleneck migration

Walks through the move from typing-cost constraints to planning drag, review burden, release anxiety, and context loss.

Shows what teams usually misdiagnose when AI output rises faster than their control plane matures.

Chapter 3

The failure modes

Names the recurring traps: review drag, private wizardry, tool sprawl, exception-driven shipping, and hidden blast radius.

Grounded in the same live-team patterns already captured in Obsidian.

Chapter 4

The control plane

Covers constitutions, business context packs, durable intent artifacts, risk tiers, reviewer routing, and writeback loops.

This is the operating layer that turns scattered AI usage into a durable system.

Chapter 5

How to install the package well

Explains what ships in the box, what gets customized first, and where the package ends and live judgment begins.

Designed to make the package feel like something a serious team can actually adopt.

Chapter 6

When to move from package to engagement

Helps buyers see the line between useful self-serve assets and the higher-touch Gumbo work that hardens the system in real environments.

The package creates motion. The engagement creates compounding.

Open the box

A finder-style preview of the package library.

The package should not feel like vague promises about templates and workflows. It should feel like a real product made of real files. This teaser previews the kinds of artifacts buyers actually get — with room later for curriculum, downloads, and paywalled digital products beyond this one offer.

AI Engineering Evolution.pkg

Selected folder

thinking/

Thesis, framing, and the long-form guide.

NameTypeAccessSize

Offer ladder

Buy the system. Buy help using the system. Or bring Gumbo in to install it.

The downloadable package is real. It is also not the same thing as live diagnosis, repo-specific policy design, and the follow-through needed when an engineering system bends under real load.

Raw package

Self-serve

Buy the operating system and install it yourself.

  • Curriculum + long-form guide
  • Templates, starter skills, workflows
  • Best for capable teams who want motion now

Guided package

Light advisory

Buy the package plus a little Gumbo help interpreting and configuring it.

  • Setup review and office hours
  • Help tuning first reviewer/risk choices
  • Best for teams that want confidence, not consulting yet

Growth package

Stronger defaults

Buy a richer installable system with more examples, scaffolding, and integration support.

  • More opinionated config examples
  • Better integration and sync scaffolding
  • Best for teams that want a closer-to-install bundle

Evolved engagement

Live install

Bring Gumbo in to diagnose, tailor, and harden the system in real work.

  • Live diagnosis and repo-specific policy
  • Workshops, tailoring, and follow-through
  • Best for teams with real complexity and real pressure

Downloads

Ship the product like a product.

The package needs real downloadable surfaces now, and the content model should be reusable later for other curricula, digital downloads, and gated products. These download cards are the first pass at that system.

HTMLFree

Download guide preview

Open or save the current browser-readable guide preview.

Open download
JSONFree

Download package manifest

See the current file inventory and access states in one machine-readable artifact.

Open download
ZIPEmail gate

Download preview bundle

A small downloadable bundle proving the package can ship as files, not just page copy.

Open download

Why Gumbo

The package is useful. The engagement is where the real compounding happens.

The package gives teams the language, artifacts, install kit, guide, and download surfaces to start moving. Gumbo's higher-touch work turns those materials into a living operating system tuned to the actual team, repo, release path, and pressure profile.