Loading
The rating is free and always will be. Credibility is the asset, and a rating nobody can read is worth nothing. What is priced is the layer underneath it: who else is in the trade, how a strategy is actually being run, and what it can absorb before its own edge is gone.
Data access
A paywall here never hides whether the underlying data exists, only the part of it that would settle a decision about one named vault.
No account required
The evidence should be readable before anybody buys the data.
or $2,490/yr · 17% less
Founding offer
$149/mo$249
For the first customers, on monthly billing, held for 12 months. Then the discount ends and the plan renews at $249. We write to you before it does.
Everything on the site: cluster membership, behavioural signals, capacity, positions and exports. One seat, no API.
or $6,490/yr · 17% less · 500,000 calls a month
The same surface at ten times the volume. For a product that reads this continuously.
| Capability | Free Free No account | Analyst $249/mo Everything paid |
|---|---|---|
| Ratings and coverage | ||
Agent-managed vault ratings Every vault we can see, with the evidence behind the autonomy verdict rather than a claim. | ||
All vaults, operators, assets, platforms The full universe, not only the agent-managed part, because a crowd is measured against everyone in it. | ||
How long each series has been running
Said here rather than discovered later. These are recordings, not reconstructions: they exist for as long as somebody has been writing them down, and no amount of reading a network recovers the part before that.
USDC or USDT on Ethereum, Base, Arbitrum or Optimism, to an address issued to you alone, the same string on every one of those chains, so use whichever is cheapest. No card, no automatic renewal, no invoice chasing. Credit appears once the transfer is deep enough to be irreversible. Sign in to see yours.
| Dataset | JSON over the API, with a key | Bulk CSV |
|---|---|---|
| Vaults | /v1/vaults | /v1/vaults?format=csv, or /export/vaults on the site |
| Crowding | /v1/crowding | /export/crowding, session only |
| Operators | /v1/operators | /export/operators, session only |
Crowding and operators answer a machine today; what they do not yet do is hand it a CSV, because the export routes authenticate with a session rather than a key. That is a missing format, not a missing dataset. /export/vaults?closed=1 includes the vaults that died: a survivor-only file is the bias this site exists to argue against, and handing it to you unmarked would be an odd way to make that argument. Every return column ships beside the basis it was computed on, so a figure cannot travel into a spreadsheet without the caveat that belongs to it.
What the API commits to, availability, latency, notice before a breaking change, and what a missed month is worth, is stated on the status page beside the measurement of what it actually did, including the months that miss. One person answers support, by email, and what to expect back is written down rather than implied.
Read the methodology before relying on any of it.
30-day consistency Whether a strategy kept operating, not merely whether its return line stayed flat. |
Methodology in full How every number is derived, including what it cannot see. A rating nobody can check is one nobody should trust. |
What we can see being used How much of the capital we track we can watch being DEPLOYED, by kind of vault and per firm: positions for a perp vault, utilisation for a lending vault, nothing at all for a strategy vault. A statement about the size of our blind spot, which is free because a platform charging for its own limitations is selling the wrong half. |
What the money is denominated in Every denomination with the peg it claims and whether we can check it: a dollar claim priced against a dollar, an ether or bitcoin claim we hold no reference for, and the ticker collisions where a dollar-named token prices far from a dollar and the price is refused rather than published as a depeg. |
| Deciding whether to allocate | ||
Which vault holds which real-world instrument How much sits in named instruments, and how much is stablecoin whose issuer invests reserves, are free. The gap between them is the finding. Which vault holds what, from which issuer, and whether we recognise the contract or only its symbol, is the part somebody acts on. | ||
The thinnest positions on a perpetual venue How levered the book is, and how much of it sits above 25x, is free. That is an argument about the market. Which position is a per-cent of its own size away from having nothing left, in which market and at what size, is the part a desk acts on. | ||
The keys behind each vault That a billion dollars stands behind one owner key is free, because it is an argument about the market. Which key, over which books, and where one address holds more than one role is what a risk desk needs by name. | ||
Which vaults are liquidated, and at what price The aggregate ladder is free: how many vaults and how much equity go at each shock. Named vaults, their exact liquidation prices and the direction behind each one are the part somebody acts on. | ||
Crowding clusters and coverage Which strategies are quietly running the same book, and what share of an instrument they are between them. | ||
Who is in a crowded book The named vaults, which side, at what size and leverage. This is what turns "there is a risk" into a decision. | ||
Behavioural evidence The execution-timing signature behind an autonomy verdict, so you can disagree with it. | ||
Capacity How much capital a strategy can absorb before its own edge is eaten, with the working shown. | ||
Current exposure What each vault holds right now, by instrument and side. | ||
Gross-to-net waterfall Where a headline return goes before it reaches a depositor. | ||
What depositors actually kept, against what the vault advertised Not the vault’s return, which is what a unit held throughout would have earned, but what the people in it kept. It is read from the depositor register against the venue’s own quote scaled to how long they have held. Measured today: of 98 vaults with a comparison, 80 gave their depositors less, at a median of 30 points short. The counts and the method are free, with the widest gaps drawn; this is the rest of the list. | ||
| Monitoring what you hold | ||
The complete change feed The last few events and the daily count stay free. The full near-real-time feed, every entity and value, the filters and the archive are the plan. | ||
Execution quality Taker share, fee rate and slippage from individual fills, the part of a result that is competence rather than market view, and the first thing to degrade when a strategy is scaled too far. | ||
Deposit and withdrawal flows Money leaves before a return line turns. | ||
Depositor concentration and timing How much of a vault is one wallet that could withdraw and reprice everyone else, and whether the best-timed wallets are already going. | ||
The price each vault is liquidated at Not how much would go at a −30% shock, that is public, but which vault, and at what named price, and how far that is from where the market is now. A number to set an alarm on rather than a scenario to read. | ||
| The agent economy | ||
Every rail we read, and what it settled Nine rails counted from their own tables (ACP, the Olas mech marketplace, x402, Tempo, Masumi, and the four identity registries) with the units named per rail, because a channel is not a payment and a registration is not either. Free, because it is the denominator behind every agent figure here. | ||
Who hires whom, and who was paid The named ledger behind the counts: which client hired which provider, who was hired twice, and which providers took jobs and never delivered them. The shares and the totals stay free; this is the row somebody acts on. | ||
What a shortlist of agents shares underneath Four agents can each have a spotless record, forty payers apiece and nothing in common on any page we publish, and still stop together on the morning the one party they all pay goes down. Paste the set you are deciding between and see which of them pay the same party. Each one’s own suppliers stay free; this is the answer that only exists once you name your own set. | ||
Whether a seller’s independent payers are independent of each other The number this site leads with counts payers that never paid an address back, and has never been proof they are different parties. This finds the ones that share an obscure third party with another of its payers: a party paid by two to ten addresses in all, where 96% of addresses ever paid have exactly one payer. The counts are free beside the number itself; which payers are linked, and through what, is here. Evidence of a link, never proof of one operator. | ||
What moved in the agent record, across every agent The tape beside this is the vault feed and always has been: it carries a few agent events while the record itself moves on roughly one agent in five every day. One row per field that moved, since whatever date you name. Every agent’s current record stays free on its check and its own history is a cent a call; this is not having to ask three thousand times to find the five hundred that moved. | ||
Which address holds a fleet of identities A registry counts registrations and a reader wants parties. This names the addresses behind the busiest endpoints and the documents a hundred thousand registrations point at, which is the difference between a crowd and one operator minting in bulk. | ||
| Taking it away with you | ||
CSV export, vaults, operators, crowding Every column as a file, with ?closed=1 for the vaults that died, a survivor-only export would hand you the exact bias this site publishes a page about. Each return carries the basis it was computed on. Cells a spreadsheet would execute are defused, since the names come from strangers on a public venue. | ||
Alerts Crowding past a share you choose, consistency falling, drawdown past a limit, or a strategy version changing. Each fires on the crossing and then stays quiet, an alert that repeats is one you would mute. | ||
This table is built from the same list the site reads when deciding what to show you, so it cannot promise something you would then be refused. It grows as the platform does. Everything listed is built, nothing here is a promise.