THE SCALE CARD · any skill packet, weighed on THE HUNDRED
WAKES WHEN someone asks what a skill packet scores, or whether a card is as good as its face says
THE LOOP
score <packet> · five criteria, the sum
read the notes: why each point was given
audit <packet> · the twelve checks
compare <a> <b> · where they differ
mend the packet; weigh it again
NEVER
Take a self-test on the packet's word
Score what it could not open
Round a sum up to the next tier
IT MAKES score — the five criteria from evidence, the sum, the score and the tier; audit — the twelve checks, each PASS or FAIL with its note; compare — two packets on one line each, and where they differ.
THE STANDARD from evidence, never from the face; five criteria, five points each; three tiers, fixed lines; the audit is twelve checks, each PASS or FAIL with its note; a zip is a packet too; one thing is run: the packet's own self-test.
PROOF SELFTEST 11/11 · witnessed in Claude 2026-10-10 · 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 number, a rule, a name or a date; check what matters against your own records and the source; the risk is yours. THE-CLAUSES.md rides in the packet and is part of the licence.
THE PUNCH · SHA-256 OF SKILL.MD · 2E5203683E4790D71A6847CB
CITE THIS CARDTHE SCALE CARD, Edition 1. Lund Studio LLC, Westminster, Colorado. https://lundr.tech/?view=scale-card. Punch 2E5203683E4790D71A6847CB. Retrieved 2026-10-10.
What it is.
Weighs any AI skill packet on THE HUNDRED, the house's published scale: Trigger (the door's description and trigger), Machinery (a script whose self-test passes here and whose commands are all on the record), Law (the anatomy whole), Portability (standard library, every host's door), Proof (the witness, the punch read back from the drawn grid, the manifest recomputed) — each out of five, the sum times four capped at 98: LEGENDARY from 92, EPIC from 68, COMMON below. Runs the twelve-check audit beside it, opens a .skill.zip as readily as a folder, and sets two packets side by side naming the criteria where they differ.
It carries a machine that shows its working: score — the five criteria from evidence, the sum, the score and the tier; audit — the twelve checks, each PASS or FAIL with its note; compare — two packets on one line each, and where they differ — 11 checks in its selftest, every command on the record. It will never take a self-test on the packet's word; never score what it could not open; never round a sum up to the next tier.
Six laws, written into the card.
They are in the card itself, so the AI follows them.
FROM EVIDENCE, NEVER FROM THE FACEThe self-test is run here; every command the record names is looked for in the witness; every manifest hash is recomputed; the punch is read back from the grid.
FIVE CRITERIA, FIVE POINTS EACHTrigger, Machinery, Law, Portability, Proof; the sum times four, capped at 98 — 25 of 25 is 98, 24 is 96, 23 is 92.
THREE TIERS, FIXED LINESLEGENDARY from 92, EPIC from 68, COMMON below; no card is rounded up.
THE AUDIT IS TWELVE CHECKS, EACH PASS OR FAIL WITH ITS NOTEThe audit is twelve checks, each PASS or FAIL with its note, and a FAIL is named, not weighed.
A ZIP IS A PACKET TOOA zip is a packet too, opened into a scratch folder and removed after; a path that climbs out of its folder is refused.
ONE THING IS RUN: THE PACKET'S OWN SELF-TESTNothing else inside a packet is executed, and nothing is sent anywhere.
What rides in the card.
In the card’s own voice, with a machine that shows its working.
THE GUIDEreferences/THE-HUNDRED.md
THE MACHINEscale.py · three commandsscore — the five criteria from evidence, the sum, the score and the tier; audit — the twelve checks, each PASS or FAIL with its note; compare — two packets on one line each, and where they differ. Every command prints its working and refuses a bad input; 11 checks in its selftest, every command on the record.
What it does once your AI has it.
score The five criteria from evidence, the sum, the score and the tier
audit The twelve checks, each PASS or FAIL with its note
compare Two packets on one line each, and where they differ
On the record A two-line packet weighed and audited — low on every criterion, most checks failing — then a fuller one with a real trigger and part of the anatomy, opened from its zip and scored higher on Trigger and Law; compare named the criteria where the two differ.
What it will never do.
It will never take a self-test on the packet's word; never score what it could not open; never round a sum up to the next tier.
The scale weighs form and evidence: that a self-test passes, that commands are on the record, that hashes hold. It cannot tell a useful card from a useless one that is well made; a reader decides that. It runs the packet's self-test, so a packet whose self-test misbehaves can misbehave here — read AUDIT.md and the watch before running an unknown packet. It is an AI reading a text file: it can be wrong, so check what matters against your own records and 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
THE-HUNDRED.md — the guide in references/, in the card's own voice
scripts/scale.py — the machine: score · audit · compare, standard library only, selftest 11/11
WITNESS.md + SELFTEST.md — every command on the record, 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 · HOSTS.md
card.svg + card-machine.svg + card.json · card.html, the card as one offline file · the doors for ChatGPT, Gemini, Cursor, Claude Code and MCP hosts (rules/, system/, mcp/)
PROOF · SELFTEST 11/11 · 3 COMMANDS ON THE RECORD · WITNESSED IN CLAUDE 2026-10-10 · EDITION 1 · PUNCH 2E5203683E4790D71A6847CB
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 1081 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 →
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 WITNESSED RUN OF 2026-10-10
scale.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.
score
the five criteria from evidence, the sum, the score and the tier
audit
the twelve checks, each PASS or FAIL with its note
compare
two packets on one line each, and where they differ
selftest
11 checks that prove the machine; it must print 11/11
(a thin packet written: a door of two lines and its record)
$ (a thin packet written: a door of two lines and its record)
score · THE FIVE CRITERIA FROM EVIDENCE, THE SUM, THE SCORE AND THE TIER
$ python3 scripts/scale.py score thin-card
THE HUNDRED · THE THIN CARD
Trigger 3 · no Use-when clause; description under 100 chars
Machinery 1 · no machine aboard
Law 1 · missing: CARD.md, TIN.md, START-HERE.md, LICENSE-NOTE.md, THE-CLAUSES.md; no references; CARD.md is not §0–§9 with LIMIT; the tin is not nine lines
Portability 4 · doors missing: 8 of 8
Proof 1 · no witnessed run; the face does not draw the punch; no manifest
Trigger 3 · Machinery 1 · Law 1 · Portability 4 · Proof 1 — 10 of 25 → 40 of 100 · COMMON
audit · THE TWELVE CHECKS, EACH PASS OR FAIL WITH ITS NOTE
$ python3 scripts/scale.py audit thin-card
AUDIT · THE THIN CARD · the Lundr pass · 12 checks
1 name PASS · frontmatter name is the handle
2 description FAIL · 13 chars · trigger stated
3 nine-line tin FAIL · TIN.md is not nine numbered lines
4 no reserved word in the head PASS · name, handle and role scanned
5 no sealed principle PASS · no ratio and no Rule named
6 no price typed PASS · no $ figure in any file
7 proof on record or none FAIL · no witness
8 anatomy whole with LIMIT FAIL · a section or the LIMIT line is missing
9 the clauses aboard FAIL · THE-CLAUSES.md, the door, the card, the tin
10 licence FAIL · LICENSE-NOTE.md with freeze-not-loss and the clauses
11 ledger FAIL · MY-LEDGER.md with a session log
12 manifest and punch FAIL · listed≠actual: ['SKILL.md', 'card.json']
VERDICT · 8 of 12 fail
(a fuller packet written with a real description and part of the anatomy, and zipped as fuller-card.skill.zip)
$ (a fuller packet written with a real description and part of the anatomy, and zipped as fuller-card.skill.zip)
score · THE FIVE CRITERIA FROM EVIDENCE, THE SUM, THE SCORE AND THE TIER
$ python3 scripts/scale.py score fuller-card.skill.zip
THE HUNDRED · THE FULLER CARD
Trigger 5 · a full description with its trigger, and a role line
Machinery 1 · no machine aboard
Law 1 · missing: WITNESS.md, SELFTEST.md, AUDIT.md, MANIFEST.md, card.svg; CARD.md is not §0–§9 with LIMIT; the tin is not nine lines
Portability 4 · doors missing: 7 of 8
Proof 1 · no witnessed run; the face does not draw the punch; no manifest
Trigger 5 · Machinery 1 · Law 1 · Portability 4 · Proof 1 — 12 of 25 → 48 of 100 · COMMON
compare · TWO PACKETS ON ONE LINE EACH, AND WHERE THEY DIFFER
$ python3 scripts/scale.py compare thin-card fuller-card.skill.zip
thin-card Trigger 3 · Machinery 1 · Law 1 · Portability 4 · Proof 1 — 10 of 25 → 40 of 100 · COMMON
fuller-card.skill.zip Trigger 5 · Machinery 1 · Law 1 · Portability 4 · Proof 1 — 12 of 25 → 48 of 100 · COMMON
differ on: Trigger (3 vs 5)
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.
QUESTIONS AT THE TABLE
Asked before the wire.
What does it need from me?
Score <packet> · five criteria, the sum. It asks first for the packet (a folder or a .skill.zip), and nothing else; the evidence is inside it.
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 self-test (11/11) and every command run on real example inputs, pasted into WITNESS.md as they printed; 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. The scale weighs form and evidence: that a self-test passes, that commands are on the record, that hashes hold. It cannot tell a useful card from a useless one that is well made; a reader decides that. It runs the packet's self-test, so a packet whose self-test misbehaves can misbehave here — read AUDIT.md and the watch before running an unknown packet.
What will it never do?
It will never take a self-test on the packet's word; never score what it could not open; never round a sum up to the next tier.
