MANIFEST — the punch on the face is the SHA-256 of the packet’s own door; the card is scored on THE HUNDRED once the studio sees it run with its app inside. What the marks mean →
CITE THIS CARDNJORD · THE DESK, Edition 1. Lund Studio LLC, Westminster, Colorado. https://lundr.tech/?view=njord-desk-card. Punch CBB1BE5B0ED0ADC279149030. Retrieved 2026-10-10.
What it is.
An agentic trading desk built as the glass-box inversion of the black box: a written mandate sealed with a hash, six hard rails in deterministic code, autonomy modes from propose to sovereign, a one-click kill, a hash-chained decision ledger, outcome reviews on every closed position — paper first on a fictional exchange, a real-market mode on the operator's own key, and a paper-brokerage harbour through the operator's own Worker where live trading is impossible by construction. Not investment advice.
The app itself goes inside the packet: the desk, one file, and the harbour Worker, from its chats (Jul 11–12 2026). Until it is sealed in, app/README.md says what goes there; seal.py writes it into the manifest without touching the door, so the punch stays true.
Its laws, written into the card.
Read the source before you change it; never generate what the app already has.
Prove every change by running the app; report what you checked and what you did not.
No key, token or password is ever written into the app or this card.
A card instructs an AI you operate; results depend on your model, your files and your judgement.
How it works, once your AI has it.
NAME — `This is NJORD · THE DESK.`
Open the desk; run its self-test.
Write and seal a mandate.
Sail the paper water for a week.
Read the ledger and the reviews.
Write the ledger: built · failed · next.
When it says no.
Trade real money.
Give investment advice.
Change a rail or a mandate on its own.
What’s in the buy.
CARD.md — the whole card · SKILL.md — the door your AI reads first · START-HERE.md — the first five minutes
app/ — the app itself, sealed in with seal.py; app/README.md until then
PROOF · UNSCORED UNTIL WITNESSED WITH ITS APP · EDITION 1 · PUNCH CBB1BE5B0ED0ADC279149030
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 984 PROPOSED · ONE PAYMENT
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 legal, tax, medical or financial advice. 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.
ONE PAYMENT — NOTHING RENEWS ITSELF. CHECKOUT IS SHOPIFY’S OWN; YOUR CARD COMES BY EMAIL AFTER IT.
