What the assay measures.

On a curve where the key taking supply from k to k+1 costs k²/16,000 SOL, the reserve after n keys is fixed. 1e9/16,000 is 62,500 exactly, so nothing leaves the integers and there is no rounding to argue about.

reserve(n) = 62,500 · (n−1)·n·(2n−1) / 6      lamports

Held against every market creatorcapital.trade will show: exact, to the lamport, with no exceptions. A market that fails this is not mispriced — it is impossible.

The reference documentation says n is “the supply after the sale of the nth key”. The chain behaves as though it is the supply before — the first key costs nothing, and the reserve holds the sum up to n−1. Quoting from the published sentence over-prices every trade by one full step of the staircase, and nothing about the result looks wrong: right shape, right units, wrong answer.

Arithmetic — does the reserve equal what this supply requires, to the lamport. Depth — if the median holder left, how much of the reserve goes with them; above a fifth, the price is a number nobody can trade at. Breadth — two holders or more, because one wallet buying its own keys produces a price that means nothing. Liveness — traded inside a week.

Consent is examined and shown beside the others but does not decide whether a market is open, because an unclaimed market genuinely is tradeable. Marking it closed would be dishonest in the other direction. It is reported because the escrow accruing to an absent person is real money owed to someone who never asked to be listed.

Docimasy has no protocol program on Solana. Nothing here custodies funds, signs, or asks for a key. /api/state returns null for the program, and will keep returning null until there is one.

The survey reads live markets from creatorcapital.trade, and that source will only return 200 of them — every documented pagination parameter is silently ignored and limit is silently capped. The survey page prints its own coverage rather than presenting a slice as the whole, which is the same standard it holds every market to.