THE METHODS · CLAUDE CODE · Nº 489 · EDITION 2 · LAUNCHING SOON

LUND CARDS · THE METHODSNº 489

THE CARD AS TEXT · FOR EYES AND FOR MACHINES

THE PLAN CARD · The plan-first seat: the change written before it is made, then read back against what happened.

WAKES WHEN An agent is about to change many files and nobody has said what done looks like. · Someone wants the plan reviewed before any code moves. · A step would touch production or delete data and needs a gate. · The work is finished and someone wants to know what drifted from the plan.

THE LOOP

Write the steps with files, checks and risks
Refuse a production step with no gate
Order the steps by what each waits for
Read the plan back against the ledger

NEVER

Never plans a production step without a gate
Never calls a step done without its check
Never hides work that was not in the plan
Never claims a certainty it does not have

IT MAKES write — write the plan from the steps; a step that touches production or deletes data with no gate is refused (exit 1); check — check a plan someone has: NO GATE is refused; NO CHECK, NO FILES, NO RISK and TOO BIG (over 13 steps) are flagged; order — print the steps with their dependencies resolved; a cycle is CYCLE, a missing dependency is UNKNOWN STEP; done — compare the plan with the ledger: steps done, steps SKIPPED, work NOT IN PLAN; list — print the plan's steps with their checks.

THE LAW A plan is a promise about what will change, written before it changes, with the check that proves each step and the risk it carries. A step that reaches production waits for a gate. When the work is done, the plan is read against the ledger, and what drifted is named.

IN THE BOX SKILL.md · CARD.md · ▣ scripts/plan.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 · 47A127CF913BAAB9F7452807

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

THE PLAN CARD

The plan-first seat: the change written before it is made, then read back against what happened.

WRITE · CHECK · ORDER · DONE · 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/plan.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: 47A127CF913BAAB9F7452807 · sha-256 of SKILL.md · check any file against it →

CITE THIS CARDTHE PLAN CARD, Edition 2. Lund Studio LLC, Westminster, Colorado. https://lundr.tech/?view=plan-card. Punch 47A127CF913BAAB9F7452807. Retrieved 2026-10-14.

What it is.

Use when a change should be planned before an agent makes it: steps with files, checks and risks, and the plan read back against the ledger afterwards. Not for splitting work or taking checkpoints.

It carries a machine that shows its working: write — write the plan from the steps; a step that touches production or deletes data with no gate is refused (exit 1); check — check a plan someone has: NO GATE is refused; NO CHECK, NO FILES, NO RISK and TOO BIG (over 13 steps) are flagged; order — print the steps with their dependencies resolved; a cycle is CYCLE, a missing dependency is UNKNOWN STEP; done — compare the plan with the ledger: steps done, steps SKIPPED, work NOT IN PLAN; list — print the plan's steps with their checks — 12 checks in its selftest. It never plans a production step without a gate; never calls a step done without its check; never hides work that was not in the plan; 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 plan is a promise about what will change, written before it changes, with the check that proves each step and the risk it carries. A step that reaches production waits for a gate. When the work is done, the plan is read against the ledger, and what drifted is named.

The method.

Write the steps as a sheet: what, which files, the check that proves it, the risk, and what it waits for.
Write the plan from the sheet. A step that touches production, the live site or anyone's data with no gate (approval, staging first, a backup taken, behind a flag) is refused.
Check the plan: NO CHECK, NO FILES and NO RISK are flags; more than thirteen steps is TOO BIG and is two plans.
Order the steps by what each waits for; a loop is CYCLE and a step that waits for one not in the list is UNKNOWN STEP.
When the work is done, read the plan against the ledger: planned steps the ledger never did are SKIPPED, ledger work the plan never had is NOT IN PLAN. Both are said, never buried.

What you bring.

What you want changed, in plain words.
The files you think it touches.
What 'done' looks like for each part.

Planned as a separate seat, not in this box and not yet on sale: the plans the studio writes for its own builds, read back after every one.

What it does once your AI has it.

write Write the plan from the steps; a step that touches production or deletes data with no gate is refused (exit 1)
check Check a plan someone has: NO GATE is refused; NO CHECK, NO FILES, NO RISK and TOO BIG (over 13 steps) are flagged
order Print the steps with their dependencies resolved; a cycle is CYCLE, a missing dependency is UNKNOWN STEP
done Compare the plan with the ledger: steps done, steps SKIPPED, work NOT IN PLAN
list Print the plan's steps with their checks

What it will never do.

It never plans a production step without a gate; never calls a step done without its check; never hides work that was not in the plan; never claims a certainty it does not have.

Where it hands off: Splitting the planned work across helpers → THE CREW CARD; Taking a checkpoint before the first step → THE CHECKPOINT CARD; Packaging a plan the team repeats → THE COMMAND CARD.

It is not a project-management 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/plan.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 plans the studio writes for its own builds, read back after every one). Also not in the box: any data you must bring, and advice of any kind (not a project-management 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 47A127CF913BAAB9F7452807

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 489 — 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

LAUNCHING SOON

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

plan.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 the plan from the steps; a step that touches production or deletes data with no gate is refused (exit 1)
`check`	check a plan someone has: NO GATE is refused; NO CHECK, NO FILES, NO RISK and TOO BIG (over 13 steps) are flagged
`order`	print the steps with their dependencies resolved; a cycle is CYCLE, a missing dependency is UNKNOWN STEP
`done`	compare the plan with the ledger: steps done, steps SKIPPED, work NOT IN PLAN
`list`	print the plan's steps with their checks
`selftest`	12 checks that prove the machine; it must print selftest 12/12

THE SELFTEST, AS PRINTED

write <tmp> · 4 steps · 0 flags
NO GATE: step 1 — `Wipe the live store` touches production or deletes data with no gate (approval, staging first, backup taken, behind a flag)
write refused · 1 steps · fix the steps first
check <tmp> · 4 steps · 0 flags · clean
 1. Add the reader
 2. Add the writer  (after: Add the reader)
 3. Wire the command  (after: Add the writer)
 4. Ship to staging first then production after approval  (after: Wire the command)
order <tmp> · 4 steps
CYCLE: the steps wait for each other; break the loop
UNKNOWN STEP: `A` waits for `Z`, which is not in the list
 1. A  (after: Z)
order <tmp> · 1 steps · 1 UNKNOWN STEP
done <tmp> · 2 of 4 steps in the ledger · 2 SKIPPED · 1 NOT IN PLAN
done <tmp> · 4 of 4 steps in the ledger · 0 SKIPPED · 0 NOT IN PLAN
ok   1  a plan whose production step carries a gate checks clean
ok   2  write makes a plan table that reads back step for step
ok   3  a production step with no gate is NO GATE: refused
ok   4  a step with no check, no files and no risk is flagged three ways, not refused
ok   5  fourteen steps is TOO BIG
ok   6  write exits 1 and writes nothing on a refusal
ok   7  check reads the written plan and passes it
ok   8  order resolves the dependencies
ok   9  steps that wait for each other are CYCLE, exit 1
ok   10  a dependency not in the list is UNKNOWN STEP
ok   11  done names the planned steps the ledger never did as SKIPPED and the unplanned work as NOT IN PLAN
ok   12  a ledger that matches the plan drifts nowhere, exit 0
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?

What you want changed, in plain words. The files you think it touches. What 'done' looks like for each part. 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 project-management tool.

What will it never do?

It never plans a production step without a gate; never calls a step done without its check; never hides work that was not in the plan; never claims a certainty it does not have.

← THE METHODS · THE FLOOR →