ASPERN

What the on-chain record shows about one DeFi vault: what holders actually earned compared with the advertised rate, the risk band and its components, who runs it, who can move the money (`keys`), whether control has changed hands, and how a holder gets out. `cannotSee` lists what could not be established for this vault; quote it with the figures. Rates are decimal fractions: 0.0432 is 4.32% a year. Call list_vaults first if you only have a name.

vault_check

Everything recorded about one address or endpoint URL, as edges with their source, when the fact was observed, when it was read, a confidence and the method. Edges are declared (the subject about itself), observed or derived, and are never added together. `lookedForAndMissing` lists what was searched for and not found. No identity is inferred from a shared host or funder.

evidence_for

Search the vaults covered by name or platform to get the slug the other tools take. A vault missing here is one we do not read, not one that does not exist.

list_vaults

Find vaults by chain, platform, size, risk score, or whether holders earned less (`short`), about the same (`in-line`) or more (`beating`) than advertised. The `earned` filter covers only vaults whose realised and advertised returns can be compared, a small part of all open vaults, so a short list means few comparable rather than few that performed. Call vault_check on a result for its full record.

find_vaults

How much of the market each vault reading covers, counted both by number of open vaults and by the capital in them. Quote both: they differ widely because many vaults are small, and neither alone is the coverage.

capital_coverage

For each phase of leaving a vault (a request filed, something set aside to claim, money that left, an attempt that failed), how much is visible across open vaults. Each phase is read, unread, exempt where the phase cannot exist on that vault, or not exposed by the contract. Failed attempts cannot be read, and no single percentage is given because the phases have different denominators.

exit_lifecycle_coverage

The hash-committed assessment of one vault. Each dimension is scored on its own and never blended into one number; `document` is exactly what the hash covers. A missing dimension is unread unless a refusal in `document.refusals` names it, which means a reading exists that is not stood behind.

vault_assessment

Across the vaults where both an advertised and a realised rate are held, what depositors actually kept against what was advertised. The set is narrow and often dominated by one venue (`population` says which and how much), so a share from here describes those vaults only. Depositors who withdrew everything are not in the register, so the result reads better than reality. Rates are decimal fractions and `gapPoints` is in percentage points.

vaults_earned

What has actually been withdrawn from one vault: settled withdrawals over 7, 30 and 90 days, the largest single exit recorded, and what the vault declares withdrawable now. Give `exitAmountUsd` with a position size to see whether that amount is covered now and whether a withdrawal that large has ever settled. Only settled withdrawals are visible, not failed attempts or time spent in a queue, and where no history is held the figures are null rather than zero.

vault_exits

Who runs a vault and how much of the capital is their own. The manager’s own holding can be read on Hyperliquid and Drift only; elsewhere it is absent, not zero, and a large own share is evidence of alignment rather than proof. Take the slug from vault_check’s `operator.slug`.

operator_check