Marketing Sections

№ 126

Alpha Honesty Band

Scene STA

Alpha Honest about it

We are in Alpha.
We will not pretend otherwise.

Three things we refuse to dress up — and the one place to tell us where it hurts.

01 Admission

No SLA. None.

Nothing here is on a uptime guarantee, a support contract, or a roadmap commitment. If a tool breaks, you might be the first to find out. We will say so.
02 Admission

Breaking changes are likely.

The framework is being shaped in public. Spec shapes, tool flags, and folder layouts will move. We will document the move; we will not pretend it did not happen.
03 Admission

Tell us what is wrong.

The feedback channel is the Discord — that is where bugs are triaged, friction is named, and the next change is decided. Lurking is fine. Speaking up is better.

№ 127

Footer CTA Combo

Scene STA

Do·Co·De·Go

An open framework for teams accountable for what AI produces in their name.

DOcument COmpose DEmonstrate GOvern

// quiet updates

One short note a month.

// real-time

Where the framework gets sharper.

Specs are debated, decisions land in writing, the conversation is the artefact.

Join the Discord

Framework

Community

In Alpha — evolving in public.

№ 128

Newsletter Signup Hero

Scene STA

Quiet updates // no funnel

One short note a month — only when something material ships.

We email when a new statute lands, when a case study goes public, or when the framework version bumps. That is the entire promise.

What to expect

▸ A short note when a new statute lands.
▸ A short note when an external case study is published.
▸ A short note when the framework version bumps.
▸ About once a month. Often less. Never more.

What we will not do

✕ No drip sequence.
✕ No tracking pixels in the emails.
✕ No "we noticed you opened" follow-ups.

// the form

Drop your email. Unsubscribe in one click.

№ 129

Social Proof Tiles

Scene STA

Voices // early adopters

Six teams. Six honest sentences.

What changed when they replaced ceremony with specifications.

We replaced our Jira-driven sprint ceremony with a four-spec backlog. Cycle time dropped, blame conversations vanished.

L. Marchetti

Engineering Director · Fintrace
The Intent Completion Score is the first quality gate that has survived a quarterly re-org. It cannot be gamed by adding bullet points.

K. Owusu

Head of Platform · NorthSignal
Our LLM agents now refuse to ship when the spec lacks failure modes. We did not build that — DoCoDeGo did.

R. Aslan

Staff Engineer · Helios Robotics
I have run agile for 12 years. This is the first framework that names the cost it creates, in writing, on day one.

P. Caldwell

VP Engineering · Outpost Health
AAR-GTR turned a six-week post-incident silence into a three-day written debrief that the whole company actually reads.

M. Devi

SRE Lead · Cargolane
We hired two engineers on the Residency track. Both shipped a spec in week two. We knew more about them than after six normal interviews.

J. Kobayashi

Founder · Strand Labs

№ 130

Hiring Poster

Scene STA

Three doors in Careers · DoCoDeGo

Pick the commitment that fits.

Start in any track. People move between them — Contribute first, Residency later, Founding when both sides know what we are building together.

Async · Your pace

Contribute

Open a PR, write a spec, file a pattern, add a glossary entry. Anything that compounds.

Best for

You have an hour, an opinion, and an internet connection.

Open in Discord →

~6 weeks · Time-boxed

Residency

Ship a defined piece of the framework under guidance. Credited on docs and Discord. Stipend if and when funding lands.

Best for

You can dedicate one focused chunk and want a written track record.

Apply in Discord →

Open-ended · Eventually paid

Founding Hire

Co-author v1 of the framework's direction. First to convert to a compensated role when funding lands.

Best for

You want to build the company, not just the codebase.

Apply in Discord →

№ 1201

Why Now Section

Scene STA do

Q2 2026 Why now

The conditions changed.
The framework should too.

Three forces converged in the last twenty-four months. None of them are reversing.

01 Force

AI labour collapsed in cost.

A senior engineer-hour now buys an order of magnitude more code than it did in 2022. Implementation is no longer the binding constraint — investigation, specification, and judgement are.
02 Force

Human attention got more expensive.

Calendars are denser, hiring is harder, and the cost of context-switching never came down. The scarce resource is no longer keystrokes; it is durable, written intent.
03 Force

Agile stopped repaying its overhead.

Standups, ceremonies, certification mills — the apparatus survived its original purpose. Teams pay for it; nobody knows what it returns. The framework that comes next has to be honest about that.

№ 1202

Who It's For Section

Scene STA co

Who it's for 3 lenses

Three roles. Three questions.
One framework that answers all of them.

Different vantage points, different scarce resources, different definitions of "done". DoCoDeGo is built to read coherently from each of them.

Individual contributor 01 / 03

Brought you here

"How do I stop writing the same code three times because nobody could agree what we wanted?"

What you get

A spec format that survives a planning meeting and a Friday refactor — written before the keyboard work begins.

Team lead 02 / 03

Brought you here

"How do I see whether the team is converging or just being busy?"

What you get

Two signals — Intent Completion Score and Green-To-Red — that read the same in every standup, every quarter.

Executive sponsor 03 / 03

Brought you here

"What stops the agents from doing something I cannot defend in writing?"

What you get

A governance ladder with explicit autonomy bands — and an audit trail that maps every shipped change back to an approved intent.

№ 1203

Before / After Section

Scene STA de

Before → After 4 rituals

Same teams. Same week.
Two different days of work.

Topic

Before

After

Cadence

Before Two-week sprint with a fixed scope and a retro that never quite lands.

After Continuous convergence — the next batch ships when the spec is approved, not when a calendar tick fires.
Definition of done

Before Acceptance criteria written after the code, then quietly relaxed under deadline pressure.

After Acceptance criteria are part of the spec, scored by ICS, and gate the merge.
Change requests

Before A change-control board, a JIRA epic, and a Tuesday status meeting that everyone attends and nobody owns.

After Re-issue the spec. Diff the intent. The board reviews the diff, not the meeting.
AI autonomy

Before A team Slack thread that argues about whether the agent should be allowed to merge anything.

After A documented autonomy band per task class, with an audit trail that survives a regulator.

№ 1204

Testimonial Stack

Scene STA go

Field reports 4 anonymised voices

No names. Just roles.
The quotes do the work.

" Practice: Failure modes

The first quarter we tried this, our on-call burden dropped because the spec told the agent what NOT to do. That was the thing nobody else was writing down.

Staff engineer Industrial robotics
" Practice: Convergence cadence

We replaced four standing meetings with one weekly spec review. The meetings were the cost; the review is the signal.

Engineering director Payments
" Practice: Audit trail

I have shipped under regulator scrutiny for nine years. The audit trail this framework produces is the first one I have been able to hand over without redaction.

Head of platform Health insurance
" Practice: Alpha honesty

The honesty about what the framework does not yet solve is what made it possible to adopt. We are not used to that from methodology vendors.

VP engineering B2B SaaS

№ 1205

Proof Points Strip

Scene STA co

Proof points Measured · 2026

Four numbers.
Each cited, each falsifiable.

62 %

fewer "what did we agree?" reverts

Quarterly rolling avg across 3 design-partner teams, 2026.

4.1 ×

spec-to-merge cycle compression

Median measured against the same teams' Agile baseline.

0

agents merging outside their autonomy band

Across 18,400 agent merges audited under the governance ladder.

78 /100

median Intent Completion Score on shipped specs

Threshold is 60. The score is calculated by the open-source ics-scorer.

№ 1206

Fundamentals Section

Scene STA do

Fundamentals 4 axioms

The load-bearing four.
Everything else is consequence.

I Pillar · DO

Intent is the primary artefact. Code is a compiled derivative.

Corollary
Regenerate the code if you must; never lose the intent.
II Pillar · CO

Humans decide what and why. Agents decide how, under boundaries.

Corollary
The cheaper side of the labour does the larger share of the typing.
III Pillar · DE

Convergence is continuous. Delivery is what convergence looks like in public.

Corollary
Small batches, frequent integration, no release theatre.
IV Pillar · GO

Governance enables autonomy. It is not the friction layer; it is the runway.

Corollary
Without a band, autonomy is a liability — and a regulator's headline.

№ 1207

Open Letter Section

Scene STA de

Open letter No NDA

To anyone still shipping software.

Filed Q2 2026

DoCoDeGo working group

If you have been doing this work for any length of time, you already know that the bottleneck moved. Code is cheap. Coordination is cheap. The expensive thing is now figuring out — and writing down — what was actually meant.

This framework is a working answer to that. It is not finished. It is not for every team. And it costs something to adopt — most visibly, the discipline of writing the intent before the code.

We are publishing it in public because the methodology cannot survive being privately owned. If you adopt any part of it, the bargain is that you tell us where it broke. That is how the next version gets written.

Signed,

The DoCoDeGo working group

// no individual signatures — that would imply ownership

№ 1208

FAQ Strip

Scene STA go

FAQ 4 questions

The questions that arrive first.

+ Is this just Agile with a new vocabulary? DO Skeptical practitioner

No. Agile optimised iteration cycles when the binding constraint was investigation cost. DoCoDeGo optimises intent capture now that AI made implementation cheap. The vocabulary is downstream of that shift.
+ How long does adoption actually take? CO Engineering lead

A small team can ship its first spec-gated change inside a week. Reaching steady-state — where the ICS gate is non-negotiable and the AAR loop is automatic — takes about a quarter, in the teams we have measured.
+ What does the governance layer actually block? DE Risk / compliance owner

Anything that lacks an approved intent, an autonomy band, and a failure-mode declaration. Concretely: agent merges outside their band, undocumented schema migrations, and any change that closes an open AAR finding.
+ What if we already use a different framework? GO Platform engineer

The pillar tools are stand-alone. ics-scorer reads any markdown spec; rc-runner runs any JSON checklist; the spec-linter only cares about a YAML frontmatter shape. Adopt one tool to start.

№ 1209

Founding Story Section

Scene STA co

Founding story 3 beats

How a question became a framework.

Dated waypoints, not personalities. The story is about a moving constraint — not the people who happened to notice it.

01

Late 2023 Provocation

Coding agents got good before our processes did.

An internal experiment shipped a working production change overnight — and then nobody could reconstruct what the change was supposed to do. The artefact existed; the intent did not. That gap became the question worth answering.
02

Mid 2024 First attempt

A four-section spec format outlived three sprints.

A small group started writing intent in a fixed shape: intent, acceptance criteria, constraints, failure modes. The format survived a re-org. The retros stopped being about blame. Something was working.
03

Q1 2026 Codification

The practice became a framework, with tools and a public name.

Four pillars, eight tools, a website, and an Alpha label that refuses to come off. The point of publishing was never to monopolise the practice — it was to make the practice quotable.

№ 1210

Principles Section

Scene STA go

Principles 6 of 14 excerpts

Manifesto excerpts.
Numbered, quotable, cite-able.

01 DO

Intent is the primary artefact. Implementation is a compiled derivative.

02 DO

A spec without failure modes is a wish, not a specification.

03 CO

The cheapest labour does the most typing; the most expensive labour does the most deciding.

04 DE

Convergence is continuous. Releases are receipts.

05 GO

Autonomy without a band is liability dressed as velocity.

06 GO

Governance is the runway, not the friction layer.

№ 1211

Team Statement Section

Scene STA do

Team statement No org chart

Sixteen people.
One working group.
Zero permanent seats.

7 Practitioners

3 Researchers

2 Regulators

4 Sceptics

01 Commitment

We will name what we do not know.

Every gap in the framework is documented in the open. If a question cannot be answered yet, the docs say so — explicitly — instead of finessing the answer.
02 Commitment

We will not charge for the framework.

Documentation, tools, and reference implementations stay open. Workshops, consulting, and certifications are not on our roadmap, and they are not what funds the next version.
03 Commitment

We will keep at least one sceptic on the group.

A working group of believers turns into a marketing department. There is a standing seat for someone whose explicit job is to argue the framework is wrong.
04 Commitment

We will replace ourselves on a fixed cadence.

Working-group membership rotates every twelve months. No founders' clause, no permanent seats. The methodology has to survive its current authors.

No SLA. None.

Breaking changes are likely.

Tell us what is wrong.

One short note a month.

Where the framework gets sharper.

Framework

Community

One short note a month — only when something material ships.

Drop your email. Unsubscribe in one click.

Pick the commitment that fits.

Contribute

Residency

Founding Hire

AI labour collapsed in cost.

Human attention got more expensive.

Agile stopped repaying its overhead.

Coding agents got good before our processes did.

A four-section spec format outlived three sprints.

The practice became a framework, with tools and a public name.

We will name what we do not know.

We will not charge for the framework.

We will keep at least one sceptic on the group.

We will replace ourselves on a fixed cadence.

The framework is real.The community is forming now.

The framework is real.
The community is forming now.