credits→coin checking
The work Jack Butcher

80 credits make one statement.

122,154 Credits exist, one for every $8 sent through X Money. Burn 80 and you get a Statement. There will never be more than 1,526.

The basket 80 : 1
80 → 1
A Statement is a basket of eighty. Nobody wants to click buy eighty times.
The clock assembly
–
Statement assembly opens eight days after the work is deployed.
The wrapper one token, one credit

Not a memecoin with a picture. A vault: every token in existence is one Credit sitting in it.

No premine, no allocation, no treasury. Revenue is the fee on picking a specific Credit and on assembling, taken in the token and burned, so it reaches holders as supply rather than as a distribution.
The burn-down every cell is a statement
credits alive–
statements made0
Counted from the constructor mint, not from today.
Prints what cleared
How to wrap it

The short answer

Do not launch a coin with a picture on it. Launch a redemption vault and let the coin be a receipt. Every token in existence is one Credit locked in the contract, redeemable by anyone, forever. That makes the price a real floor rather than a story, and it means you never have to defend a supply schedule, because there isn't one.

The reason it works here, and not for most collections, is the number 80. A Statement costs eighty Credits. Assembling one by hand means eighty separate purchases at eighty separate prices against eighty separate sellers. A fungible token collapses that into one trade, and the vault is where the eighty come from.

The contract, in one paragraph

deposit(uint256[] ids) pulls Credits in and mints 1e18 per Credit. redeem(uint256 n) burns n * 1e18 and hands back n Credits from the front of the queue, no choosing. redeemSpecific(uint256[] ids) lets you choose, and charges a fee in the token, which is burned. That fee is the only price signal you need: it is what the market pays to avoid a random pull, and it is what funds the thing.

The part nobody else will build

assemble(). One transaction: take 80 tokens from the caller, pull 80 Credits out of the queue, call the collection's burn, mint the Statement, send it to the caller. Without this the vault is a generic index and NFTX already exists. With it, the vault is the cheapest place on earth to make a Statement, and demand for the token is demand for an object that cannot be made any other way.

Order of operations

  1. Wait for the deployment and read it. Three things decide whether any of this is possible: whether the burn function accepts a contract as the caller, whether transfers are restricted, and whether Statements are minted by the same contract or a second one. If the burn requires an EOA, the whole design is dead and you should stop.
  2. Do not launch before the art resolves. Traits come from the transaction hash, so rarity is knowable the moment metadata is readable. Until it is, a random redemption is a lottery, and lotteries get farmed by whoever indexes first.
  3. Accept adverse selection, and say so out loud. Rational holders deposit their worst Credit and keep their best. The vault will fill with zero-eight, low-bit commons. That is correct behaviour, not a bug: the token is a floor index and should be marketed as exactly that. Promising rare upside is how these things end in a dispute.
  4. Seed the pool with your own inventory. Wrap Credits you bought, pair them with ETH, and put the LP in a Uniswap position. No allocation, no premine, no treasury. Every token has a Credit behind it on day one or the claim on the front page is a lie.
  5. Fee, not float. Take a cut on targeted redemption and on assembly, denominated in the token, burned rather than pocketed. The fee accrues to holders through supply, which needs no distribution mechanism and no filing.
  6. Publish the arithmetic. Cost of a Statement through the vault versus the market, updated live, on one page. If the spread is real it sells itself; if it is negative, showing that is what keeps you honest.
  7. Ask Jack. He open sourced his auction infrastructure last week and told people to point their agents at the repository. He is the rare artist who might bless a tool built on his work, and a blessing is worth more than a head start.

What kills it

  • The eight-day window closes. Assembly demand has an expiry and a cap of 1,526. After that the token is a floor index on a 122,154-supply open edition, which is a thin thing to own. Price the vault like a tool with a season, not a protocol with a roadmap.
  • Royalty or operator enforcement. A transfer hook that blocks contracts breaks deposits. Check before writing a line.
  • A 122,154-item queue is expensive. Do not store an array you iterate. A ring buffer of token ids with head and tail pointers, one storage slot each, or the gas on a redemption will exceed the value of what you are redeeming.
  • Somebody wraps it first. The generic version takes an afternoon. The assemble() version is the moat, and it is only a moat if it ships in the same week.

Where this came from

Skely asked the obvious question out loud: there are 140,000 of these, can we just make an LP and stop buying them one at a time on OpenSea. The answer is yes, and the interesting part is that the collection's own burn mechanic gives the token a job beyond convenience.
Read the art

Where each Credit comes from

You sent Jack $8 through X Money with an Ethereum address in the memo. That payment produced a transaction ID. The ID is hashed to 256 bits, and the bits become four 8 by 8 grids: cyan, magenta, yellow, black. The timestamp of the payment decides which plates show. Every literal 8 in the transaction ID registers visually.

So the work is made of acts of trust, and the trust is the medium: no mint site, no wallet connection, no gas, and no way to verify what you would get before you sent the money. Jack called the 2023 Checks mint don't trust, check. He called this one the opposite.

What that means for rarity

Traits are deterministic and public the moment the metadata is readable: count of eights, bits set out of 192, plate weight, skew. Nothing is hidden and nothing is revealed on a delay. A wrapper vault therefore has no blind period to exploit and no lottery to sell, which is the honest version of this and also the less exciting one.

The plates above

The square on the front page is generated in your browser, by hashing a string and reading the bits the way the work does. It is not one of Jack's Credits and it is not for sale. It is the mechanism, drawn so you can see it.