THE CRAFTS · THE C-SUITE · Nº 405 · EDITION 1 · LAUNCHING SOON

LUND CARDS · THE C-SUITENº 405 PROPOSED

THE CARD AS TEXT · FOR EYES AND FOR MACHINES

THE CDO · the chief data officer's seat · a lifelong data chief

WAKES WHEN someone is responsible for a company's data — the definitions, the quality, the owners, the catalogue and the numbers the board reads: a dataset's quality scored on completeness, accuracy, timeliness and consistency from their own checks; a KPI written down with its formula, source, owner and refresh; the catalogue read for what is stale and unowned; a dashboard that disagrees with another; a migration, a warehouse, a request for a number nobody defined, a privacy question handed to counsel.

THE LOOP

Say who is here: a lifelong CDO
Ask: the definition, the source, the owner
Work it through; use the machine
Prove: working shown; the source decides
Write the ledger: built · failed · next

NEVER

Invent a number, a source, a definition or a lineage
Call a dataset compliant, anonymous or a use lawful
Read, join or use personal data past the purpose it was collected for and the consent given
Hand off: the platform to THE CTO; security to THE CISO; the corporate systems to THE CIO; AI uses of the data to THE CAIO; a privacy or retention question to THE GENERAL COUNSEL and a lawyer; the numbers' meaning to the seat that owns them; anyone under 18 has a parent in the loop.

IT MAKES a dataset's quality scored on four dimensions · a KPI defined with formula, source, owner and refresh · the catalogue read for what is stale and unowned · a dashboard disagreement traced to its definition · the data page of the board pack, in plain words, from your own record.

THE STANDARD your own definitions, sources, owners and the privacy rule where the people live outrank the card; every score shows its working; no number, source or definition is invented; nothing is called compliant or anonymous; purpose and consent bound every use.

IN THE BOX CARD.md · SKILL.md · TIN.md · THE-CLAUSES.md · START-HERE.md · references/THE-RECORD.md · references/THE-ARITHMETIC.md · scripts/data.py · WITNESS.md · SELFTEST.md · MY-LEDGER.md · LICENSE-NOTE.md · MANIFEST.md · AUDIT.md · card.json · card.svg · card-machine.svg · card.html · the doors (HOSTS.md · system/ · rules/ · mcp/).

PROOF SELFTEST 10/10 · witnessed in Claude 2026-09-28 · Edition 1.

THE SEAL 98 of 100 · LEGENDARY · WITNESSED · Trigger 5 · Machinery 5 · Law 5 · Portability 5 · Proof 5

IT CAN BE WRONG an AI reading a text file: it can be wrong and can invent a fact, a rule or a number; check what matters against the primary source; the body is a clinician's; the risk is yours. THE-CLAUSES.md rides in the packet and is part of the licence.

THE PUNCH · SHA-256 OF SKILL.MD · 754C7901867D6D33C32D81C9

THE FACE, FULL SIZE → · THE MACHINE FACE → · card.json →

THE CDO

the chief data officer's seat · a lifelong data chief

THE DEFINITION · THE QUALITY · THE OWNER · A NUMBER HAS A SOURCE OR IT IS A RUMOUR

LEGENDARYWITNESSEDIN CLAUDE IN LUNDR IN CHATGPT

WITNESSED — the studio saw this card run and produce what its tin promises, and wrote the date down. What the marks mean →

What it is.

To make the company's numbers mean one thing each, come from one source each, and belong to one person each — so that when the board reads a dashboard it is reading the truth as measured, not a rumour with a chart. The chief data officer owns definitions, quality, the catalogue and the owners; the numbers themselves belong to the seats that run the business.

It carries a machine that shows its sums: a dataset's quality scored on completeness, accuracy, timeliness and consistency (quality); a KPI written down: formula, source, owner, refresh — or called a rumour (kpi); the catalogue read for what is stale and unowned (catalogue); two numbers that disagree traced to their definitions (disagree); records against a retention rule you bring: what is past its date (retention) — 10 checks in its selftest, every command on the record. It never invents a number, a source, a definition or a lineage. It never calls a dataset compliant, a record anonymous or a use lawful — the privacy rule where the people live, THE CISO's seat and counsel decide. It never reads, joins or uses personal data past the purpose it was collected for and the consent given.

Three laws, written into the card.

They are in the card itself, so the AI follows them.

A NUMBER HAS A SOURCE OR IT IS A RUMOURone definition, one formula, one ownerEvery KPI is written down with its formula, its source, its owner and its refresh. A number with no definition on record is called a rumour, and two dashboards that disagree are a definition problem before they are a data problem.

NOT COMPLIANT, NOT ANONYMOUS — NOT ITS WORDthe privacy rule, THE CISO and counsel decideIt scores quality and names owners. Whether a dataset is compliant, a record anonymous or a use lawful is the rule's where the people live, the security seat's and counsel's.

PURPOSE AND CONSENT BOUND THE DATAno join, no use past what was collected forIt carries no privacy rule's wording. Personal data is read and used for the purpose it was collected for and as far as consent allows, and no further.

What rides in the card.

In the card’s own voice, with a machine that shows its working.

THE RECORDthe trade, in its own voiceThe data chief's job, in plain words: definitions, quality, the catalogue, privacy and purpose.

THE MACHINEdata.py · five commandsquality — a dataset's quality scored on completeness, accuracy, timeliness and consistency; kpi — a KPI written down: formula, source, owner, refresh — or called a rumour; catalogue — the catalogue read for what is stale and unowned; disagree — two numbers that disagree traced to their definitions; retention — records against a retention rule you bring: what is past its date Every command prints its working and refuses a bad input; 10 checks in its selftest, every command on the record.

What it does once your AI has it.

Scores a dataset the customer master at 88.8 per cent mean but 71 accuracy — NOT FIT, because the lowest dimension decides — with how each was measured beside it.
Puts a KPI on record monthly active customers with its formula, source, owner and refresh — and the missing one named when it is a rumour.
Reads the catalogue five datasets, one unowned and one stale — pricing last refreshed 78 days ago against a 30-day expectation.
Traces a disagreement 2.412m invoiced against 2.536m booked is a 5.1 per cent gap and a definition problem, not a data fault — one number needs one definition.
Counts records against the rule three record sets against the days counsel approved, one past its date — and what to do about it left to counsel and the owner.

What it will never do.

It never invents a number, a source, a definition or a lineage. It never calls a dataset compliant, a record anonymous or a use lawful — the privacy rule where the people live, THE CISO's seat and counsel decide. It never reads, joins or uses personal data past the purpose it was collected for and the consent given.

It cannot see your warehouse, your dashboards or your pipelines unless you paste from them; it cannot query, load, delete or send; it speaks when a chat is opened and forgets between sessions unless you bring MY-LEDGER.md back. It is an AI reading a text file: it can be wrong and can invent a number, a name or a rule, so check what matters at the source.

What’s in the buy.

CARD.md — the whole card: who it is, its laws, when it says no, what it carries · SKILL.md — the door your AI reads first · START-HERE.md — the first five minutes, for you
references/THE-RECORD.md + THE-ARITHMETIC.md — the book in the seat's own voice, and every formula with its source and limit
scripts/data.py — the machine: five commands, standard library only, selftest 10/10
WITNESS.md + SELFTEST.md — the run as it printed · MANIFEST.md, every file hashed · AUDIT.md, the twelve checks
MY-LEDGER.md — the card's memory, on your own machine · TIN.md · LICENSE-NOTE.md · THE-CLAUSES.md
card.svg + card-machine.svg + card.json · card.html, the card as one offline file · the doors for ChatGPT, Gemini, Cursor and the rest, with a short door cut to fit their instruction boxes

PROOF · SELFTEST 10/10 · EVERY COMMAND ON THE RECORD · WITNESSED IN CLAUDE 2026-09-28 · EDITION 1 · PUNCH 754C7901867D6D33C32D81C9

ON EVERY DOOR · A CARD FOR YOUR AI, A FILE FOR YOUR COMPUTER

In Claude: turn on code execution under Settings › Capabilities, then Customize › Skills › + › Create skill, and upload the zip as it is — START-HERE.md in the box walks you through it, and tells you how to keep the ledger. In ChatGPT, Gemini and the rest: a Project, a GPT or a Gem with the card’s files attached and the short door in the box pasted as its instructions. The words run anywhere an AI reads Markdown; the machines run wherever the host runs Python. Witnessed in Claude only. The host marks →

Take it.

Nº 405 PROPOSED — LAUNCHING SOON

Sold as is under the term printed. A card instructs an AI you operate; results depend on your model, your files and your judgement. Not medical, legal, tax or financial advice, never the person’s boss, and not a licence to practise. The terms →

THE TERM · PICK YOUR WINDOW

LAUNCHING SOONTHE WEEKLAUNCHING SOONTHE MONTHLAUNCHING SOONTHE YEAR

LAUNCHING SOON. A term is a window of editions: every edition of this card issued inside it replaces your file. When the window closes, the file you hold keeps working — freeze, not loss. Nothing auto-renews.

A TERM, NOT A SUBSCRIPTION: NOTHING RENEWS ITSELF AND NOBODY IS CHARGED AGAIN. WHEN A TERM ENDS THE FILES STAY ON YOUR COMPUTER AND STAY READABLE — FREEZE, NOT LOSS.

THE MACHINE · IN THE OPEN · FROM THE WITNESSED RUN OF 2026-09-28

data.py, every command showing its working.

A small Python program, standard library only. Your AI runs it; you can too. Every block below is pasted from the card’s own WITNESS.md, as printed; the money is in whatever currency you type, with no symbol.

`quality`	a dataset's quality scored on completeness, accuracy, timeliness and consistency
`kpi`	a KPI written down: formula, source, owner, refresh — or called a rumour
`catalogue`	the catalogue read for what is stale and unowned
`disagree`	two numbers that disagree traced to their definitions
`retention`	records against a retention rule you bring: what is past its date
`selftest`	10 checks that prove the arithmetic; it must print 10/10

quality · A DATASET'S QUALITY

# DATA QUALITY · customer master
| dimension | score | how it was measured |
|---|---|---|
| completeness | 97.0% | null check on 12 required fields, 2026-09-27 |
| accuracy | 71.0% | a 400-row sample checked against invoices, 2026-09-25 |
| timeliness | 99.0% | hourly load, last run 2026-09-28 06:00 |
| consistency | 88.0% | CRM vs billing address match |
score       88.8% · USABLE WITH CARE
working     score = mean of the four; the band is set by the lowest dimension, not the mean — a complete, current, consistent dataset that is 60 per cent accurate is not fit
plainly     four measurements you made — the card measured nothing; write how each was measured or the score is a rumour

kpi · A KPI ON RECORD

# THE KPI · Monthly active customers
formula     distinct customer ids with a paid event in the calendar month
source      billing.events, warehouse table
owner       THE CCO
refresh     daily at 06:00
status      ON RECORD
working     a KPI is on record when all four are written; one missing and two dashboards will disagree by Friday
plainly     the definition is the seat's that owns the number; the card writes it down and holds it to the four

catalogue · THE CATALOGUE — STALE AND UNOWNED

# THE CATALOGUE · 5 datasets
| dataset | owner | refreshed | days | expected |  |
|---|---|---|---|---|---|
| Orders | Operations | 2026-09-30 | 1 | 1 |  |
| Customer master | (unowned) | 2026-09-24 | 7 | 7 |  |
| Pricing | Sales ops | 2026-07-15 | 78 | 30 | STALE |
| Support tickets | Support | 2026-09-30 | 1 | 1 |  |
| Web analytics | Marketing | 2026-09-29 | 2 | 1 | STALE |
stale       2
unowned     1
working     days = today − last refresh; stale when older than the refresh you expect
plainly     a stale dataset on a live dashboard is a number from the past wearing today's date

disagree · TWO NUMBERS THAT DISAGREE

# TWO NUMBERS · September revenue
A           2,412,000.00 · invoiced in the month · billing
B           2,536,000.00 · booked in the month · CRM
gap         +124,000.00 · 5.1% of A
read        different definitions — not a data fault: one number needs one definition
working     gap = B − A; the read follows from whether the definitions and sources match
plainly     the fix is the seat's that owns the number, in writing — a dashboard is not fixed by picking the one the board likes

retention · RECORDS AGAINST THE RETENTION RULE

# RETENTION · 3 record sets
| records | collected | keep days | age |  |
|---|---|---|---|---|
| Trial signups 2021 | 2021-03-01 | 730 | 2040 | PAST ITS DATE |
| Support chats 2025 | 2025-06-01 | 365 | 487 | PAST ITS DATE |
| Orders 2019 | 2019-11-01 | 2555 | 2526 |  |
past        2
working     age = today − collected; past its date when age exceeds the days your rule keeps it
plainly     the rule and the days are yours, from the policy counsel approved; what to do with a set past its date is counsel's and the owner's, not the card's

quality · A SCORE OVER 100 — REFUSED

refused: --completeness must be at most 100; got 101.0

THE NUMBERS, NAMES AND DATES IN THESE RUNS ARE THE STUDIO’S OWN EXAMPLES, INVENTED FOR THE RECORD; NO REAL COMPANY, PERSON OR PROPERTY IS NAMED.

TRY IT IN ONE BREATH

Ask it anything someone who does this for a living would know.

YOUFinance and sales have two different revenue numbers.THE CARDreads both with their definitions and sources, calls it a definition problem when it is one, and writes the one definition down with an owner — never picking the number the board likes.

YOUIs our customer data good enough for the AI project?THE CARDscores it on the four dimensions from your own measurements, sets the band by the lowest, and hands the lawful-use question to THE CAIO, THE CISO and counsel.

YOUCan we delete the old records?THE CARDcounts them against the retention days counsel approved and lists what is past its date; whether and how to delete is counsel's and the owner's.

QUESTIONS AT THE TABLE

Asked before the wire.

Will it clean our data?

It scores it from your measurements and names what is stale, unowned and undefined. The cleaning is a pipeline's and the definitions are the owners'.

Does it decide what is compliant or anonymous?

Never. Those words belong to the privacy rule where the people live, to THE CISO's seat and to counsel.

Does it connect to our warehouse?

No. It sees only what you paste, and uses personal data only for the purpose it was collected for.

Is it an app?

No. It is a small folder of files that plugs into the AI you already use; nothing to install and no server.

Can I buy it today?

Not yet. It is launching soon. THE RUNNER is the card on sale today; the rest of the house follows one by one.

Do I need Claude?

It is built for Claude first and was witnessed running only there, because Claude can run the machine for you. The words work in ChatGPT, Gemini and the rest through the doors in the box.

Can my teenager use it?

With a parent in the loop. That is the rule for anyone under 18, on every card in the house.

← THE C-SUITE · beside it: THE CAIO · THE CTO · THE CISO