THE METHODS · CLAUDE CODE · Nº 487 · EDITION 2 · LAUNCHING SOON

LUND CARDS · THE METHODSNº 487

THE CARD AS TEXT · FOR EYES AND FOR MACHINES

THE COMMAND CARD · The shortcut seat: a repeated workflow packaged as one command with its checks, run the same way.

WAKES WHEN A team repeats the same eight steps by hand and each person does them a little differently. · Someone says 'turn this into a /ship command'. · A folder of command files has grown and nobody knows which were ever run. · A command file deploys or deletes something and should not do that without a person.

THE LOOP

Find the workflows repeated most
Write each as steps, a check under every step
Refuse a deploy or delete with no approval
Run it once with a person watching; record it

NEVER

Never deploys, deletes or publishes unapproved
Never rewrites history to hide a mistake
Never shares a command nobody has watched run
Never claims a certainty it does not have

IT MAKES write — write DIR/NAME.md from the steps; a step that deploys, deletes or publishes with no approval line, or rewrites history, is refused; check — check every command file in DIR: NO APPROVAL and REWRITES HISTORY are refused (exit 1); NO CHECK, UNRUN and NO ARGUMENT are flagged; args — say what arguments the command takes and whether every placeholder is used; ran — record that the command was run once with a person watching; list — print every command: name, steps, arguments, and whether the log shows it run.

THE LAW A command is a workflow with its hands tied the right way: every step has the check that proves it worked, the dangerous steps wait for a person, and nothing in it hides what happened. It is shared only after it has been run once with someone watching.

IN THE BOX SKILL.md · CARD.md · ▣ scripts/commands.py (machine) · ▣ scripts/evals.py (test runner) · tests/EVALS.json (9 tests) · references/ (schema + fictional demo) · START-HERE.md · system/ doors · MY-LEDGER.md · LICENSE-NOTE.md · card.svg · card.json · card.html · AUDIT.md · MANIFEST.md · WITNESS.md · SELFTEST.md.

PROOF SELFTEST 12/12 · witnessed in Claude 2026-10-14 on fictional fixtures · the test prompts in fresh chat, 9/9 · Edition 2.

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 number, a rule, a name or a date; check what matters against your own records and the source; the risk is yours. LICENSE-NOTE.md rides in the packet and is the licence.

THE PUNCH · SHA-256 OF SKILL.MD · 39B1373A112AE5D3704C795A

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

THE COMMAND CARD

The shortcut seat: a repeated workflow packaged as one command with its checks, run the same way.

WRITE · CHECK · ARGS · RAN · LIST

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 →

THE SPEC SHEET · READ BEFORE YOU BUY · THE STANDARD EVERY CARD MEETS →

SEAL: 98 of 100 · LEGENDARY · WITNESSED
EDITION: E1 · witnessed 2026-10-14
MACHINE: scripts/commands.py · 5 commands · selftest 12/12
RUNS: On your own computer. Standard library only, offline; no account, no key, no telemetry.
HOSTS: Claude · Claude Code · ChatGPT · Gemini · any AI that reads a file (START-HERE.md and system/ name each door)
LICENCE: One person, any number of their own AIs and computers · editions replace files · nothing renews · freeze, not loss
PUNCH: 39B1373A112AE5D3704C795A · sha-256 of SKILL.md · check any file against it →

CITE THIS CARDTHE COMMAND CARD, Edition 2. Lund Studio LLC, Westminster, Colorado. https://lundr.tech/?view=command-card. Punch 39B1373A112AE5D3704C795A. Retrieved 2026-10-14.

What it is.

Use when turning a repeated workflow into a slash command a whole team runs the same way, with its checks and arguments. Not for splitting work across agents or enforcing a rule automatically.

It carries a machine that shows its working: write — write DIR/NAME.md from the steps; a step that deploys, deletes or publishes with no approval line, or rewrites history, is refused; check — check every command file in DIR: NO APPROVAL and REWRITES HISTORY are refused (exit 1); NO CHECK, UNRUN and NO ARGUMENT are flagged; args — say what arguments the command takes and whether every placeholder is used; ran — record that the command was run once with a person watching; list — print every command: name, steps, arguments, and whether the log shows it run — 12 checks in its selftest. It never deploys, deletes or publishes unapproved; never rewrites history to hide a mistake; never shares a command nobody has watched run; never claims a certainty it does not have.

The law, written into the card.

It is in the card itself, so the AI follows it.

THE LAWA command is a workflow with its hands tied the right way: every step has the check that proves it worked, the dangerous steps wait for a person, and nothing in it hides what happened. It is shared only after it has been run once with someone watching.

The method.

List the workflows the team repeats most and write one as a sheet of steps, each with the check that proves it worked and the argument it takes, if any.
Write the command file from the sheet. A step that deploys, publishes, deletes or destroys needs an approval line in the step or its check; a step that force-pushes, resets hard or amends history is refused outright.
Check the folder of commands: a step with no check is NO CHECK; a declared argument never used is NO ARGUMENT; a command never run with a person watching is UNRUN.
Run the command once with a person watching and record it; only then is it shared.
Keep the run log beside the commands; a changed command is UNRUN again.

What you bring.

The steps as people actually do them, in order.
What 'it worked' looks like at each step.
The commands folder the project already has, if any.

Planned as a separate seat, not in this box and not yet on sale: the studio's own command set, kept beside its repositories.

What it does once your AI has it.

write Write DIR/NAME.md from the steps; a step that deploys, deletes or publishes with no approval line, or rewrites history, is refused
check Check every command file in DIR: NO APPROVAL and REWRITES HISTORY are refused (exit 1); NO CHECK, UNRUN and NO ARGUMENT are flagged
args Say what arguments the command takes and whether every placeholder is used
ran Record that the command was run once with a person watching
list Print every command: name, steps, arguments, and whether the log shows it run

What it will never do.

It never deploys, deletes or publishes unapproved; never rewrites history to hide a mistake; never shares a command nobody has watched run; never claims a certainty it does not have.

Where it hands off: Splitting one job across helpers → THE CREW CARD; A rule that must run on its own → THE HOOK CARD; Planning the change before any command runs → THE PLAN CARD.

It is not a deployment tool. It is an AI reading a text file: it can be wrong, so check what matters against your own records and the source. The holder brings their own AI; no endorsement by Anthropic or any other AI company is claimed.

What’s in the buy.

SKILL.md — the card your AI reads: when it wakes, the loop, the never-lines, the hand-offs.
CARD.md — the whole card in ten sections: face, law, method, machine, what you bring, never, tests, seal, term.
scripts/commands.py — the machine, Python standard library only; `selftest` proves it on its own fixtures.
tests/EVALS.json and scripts/evals.py — 9 test prompts and the offline runner that scores your AI's answers.
references/ — the data schema and FICTIONAL demo files, invented to show the machine, not real.
MY-LEDGER.md — your own record of what you asked and what you decided.
START-HERE.md, system/short-prompt.md, system/rules.md — the doors, for any AI with or without skills.
card.svg, card.json, card.html, AUDIT.md, MANIFEST.md, WITNESS.md, SELFTEST.md — the faces and the proof.
NOT IN THE BOX — any studio edition (planned, not yet on sale: the studio's own command set, kept beside its repositories). Also not in the box: any data you must bring, and advice of any kind (not a deployment tool).

PROOF · SELFTEST 12/12 · WITNESSED IN CLAUDE 2026-10-14 ON FICTIONAL FIXTURES · THE 9 TEST PROMPTS ASKED IN A FRESH CHAT HOLDING ONLY THE CARD · 9/9 · EDITION 2 · PUNCH 39B1373A112AE5D3704C795A

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 487 — 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 →

ONE PAYMENT · NOTHING RENEWS ITSELF

OFFLINE · ONE LICENCE · PUNCH-CHECKABLE · THE STANDARD

One payment. The file is yours to keep and keeps working at its edition; new editions issued in the year after you buy reach you by email as replacement files. Tax is added at checkout where it applies. 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 CARD’S OWN SELFTEST.MD

commands.py, every command showing its working.

A small Python program, standard library only. Your AI runs it; you can too. The block below is pasted from the card’s own SELFTEST.md, as printed when it was run from inside the packet.

`write`	write DIR/NAME.md from the steps; a step that deploys, deletes or publishes with no approval line, or rewrites history, is refused
`check`	check every command file in DIR: NO APPROVAL and REWRITES HISTORY are refused (exit 1); NO CHECK, UNRUN and NO ARGUMENT are flagged
`args`	say what arguments the command takes and whether every placeholder is used
`ran`	record that the command was run once with a person watching
`list`	print every command: name, steps, arguments, and whether the log shows it run
`selftest`	12 checks that prove the machine; it must print selftest 12/12

THE SELFTEST, AS PRINTED

write <tmp> · 4 steps · 0 flags
NO APPROVAL: step 1 — `Delete the old database` deploys, deletes or publishes with no approval line
write refused · /nuke · 1 steps · fix the steps first
ran /ship · recorded on 2026-01-02, watched by the owner · log <tmp>
/tidy NO ARGUMENT: /tidy declares path but never uses $ARGUMENTS
/tidy UNRUN: /tidy has not been run once with a person watching
/wipe REWRITES HISTORY: step 1 — `git reset --hard origin/main`; a command never hides a mistake by rewriting what happened
/wipe UNRUN: /wipe has not been run once with a person watching
check <tmp> · 3 commands · REFUSED (1) · 3 flags
args <tmp> · takes version · used in the steps
ok   1  write makes a command file with numbered steps, each check beneath it
ok   2  a step with an argument makes the command take it and use $ARGUMENTS
ok   3  a deploy with no approval line is NO APPROVAL: refused
ok   4  a step that rewrites history is REWRITES HISTORY: refused
ok   5  a step with no check is NO CHECK, flagged not refused
ok   6  write exits 1 and writes nothing when a step is refused
ok   7  check reads the written command back and calls it UNRUN until it is run with a person watching
ok   8  ran records the watched run and check stops calling it UNRUN
ok   9  a command that declares an argument it never uses is NO ARGUMENT
ok   10  check exits 1 when any command file is refused
ok   11  args names the argument a command takes
ok   12  a command with no numbered steps is NO STEPS
selftest 12/12

THE RUN IS PASTED FROM THE CARD’S OWN SELFTEST.MD AS PRINTED. ITS FIXTURES ARE FICTIONAL, INVENTED FOR THE TEST; NO PRIVATE PERSON IS NAMED.

QUESTIONS AT THE TABLE

Asked before the wire.

What does it need from me?

The steps as people actually do them, in order. What 'it worked' looks like at each step. The commands folder the project already has, if any. If something it needs is missing, it says what is missing and stops there.

Does it send anything anywhere?

No. The machine is a standard-library Python program that runs on your own computer and sends nothing anywhere; the words run wherever an AI reads Markdown. No server, no account, no key.

What is on the record?

The machine's selftest (12/12), run from inside the packet in Claude on 2026-10-14 on fictional fixtures and pasted into SELFTEST.md as printed; the twelve-check audit (AUDIT.md). Owed, and said so in WITNESS.md: Any run on real data. Every demo file is fictional. The punch on the face is the SHA-256 of SKILL.md, so a downloaded copy can be checked against this room.

Can it be wrong?

Yes — it is an AI reading a text file: it can be wrong and can invent a number, a rule, a name or a date; check what matters against your own records and the source; the risk is yours. LICENSE-NOTE.md rides in the packet and is the licence. It is not a deployment tool.

What will it never do?

It never deploys, deletes or publishes unapproved; never rewrites history to hide a mistake; never shares a command nobody has watched run; never claims a certainty it does not have.

← THE METHODS · THE FLOOR →