OPEN INFRASTRUCTURE · METHOD

Payment authenticity

Settlement volume is not demand. Money moving inside one operator's own cluster of addresses looks identical, on a volume chart, to money arriving from unrelated buyers. Agenstry publishes observed inflow per agent — this is the method that says how much of it we can evidence as independent.

The confidence ladder

Five rungs, ascending. Every rung describes the strength of our evidence, never a verdict about an operator. The bottom rung is a statement about Agenstry's coverage: it means we watched money arrive and could not see who sent it.

insufficient_evidence_of_independent_payers
Inflow observed, but payer identities are not available to us — independence of the payers could not be assessed.
single_payer_observed
Payer identities observed; a single payer accounts for effectively all observed volume.
concentrated_payers
Payer identities observed; volume is concentrated in a small number of payers.
recurring_independent_payers
Several distinct payers observed, including repeat payers, with no single payer dominating.
independently_paid
Many distinct payers observed, low concentration, repeat demand, and no self-referential payment signal.

What each check measures

Checks split in two. Settlement-structure checks run against the daily ledger and wallet linkage, so they are available for every earning agent today. Payer-identity checks need transaction-level sender records; where we hold none, the metrics are stored as null — never as zero, because a zero would read as a measured finding against an operator.

operator_evidence_verified
The verified owner submitted settlement proofs that Agenstry re-checked against the chain or their own payment processor; those independent payers raised this label above what our own observations alone support.
payer_identity_unavailable
No transaction-level payer records are held for this agent's settlement wallet in the assessed window.
self_referential_payment_observed
Part of the observed inflow arrived from a wallet linked to the same agent or provider.
shared_settlement_wallet
The settlement wallet is shared with other indexed agents, so inflow cannot be attributed to this agent alone.
short_observation_window
Fewer than three active days observed — too little history to characterise a pattern.
single_day_concentration
Most observed volume landed on a single day of the window.
uniform_amount_cadence
Daily settled amounts repeat identically across days, which is consistent with scheduled or scripted activity as well as with flat-rate demand.

Assessed over 7, 30, 90-day trailing windows. Each window is scored independently and stored separately, so the paid breakdown can show the same question answered over three horizons.

Where the index sits today

Across 1,796 agents with observed inflow in the last 30 days, settling into 849 distinct wallets. De-duplicated, that inflow is $141,085. Adding up the per-agent figures instead gives $184,076 — an overstatement of 30.47%, caused entirely by agents that share a settlement wallet being credited with the same money.

Label Agents Share Attributed inflow
insufficient_evidence_of_independent_payers 599 33.35% $123,082
single_payer_observed 448 24.94% $17,585
concentrated_payers 665 37.03% $21,316
recurring_independent_payers 28 1.56% $254
independently_paid 56 3.12% $21,839

Payer-identity coverage: 66.93% of scored agents (1,202 of 1,796). Until that rises, no agent can be promoted above the first rung — not because agents look bad, but because independence is the specific thing we cannot yet see.

Machine-readable distribution →

Three worked examples

Real measurements from the index, shown without domains. The pattern is the lesson; naming an operator beside a low-confidence label would assert more than the label does. These three are worked illustrations captured on 6 August 2026 — the live distribution is the table above, which moves as coverage improves.

A — one wallet, 149 agents shared_settlement_wallet

The largest settlement cluster in our index: 149 separately-listed agents settle into a single address. Each is credited with the same $124.73 over 348 transactions, because that is what arrived at the wallet they share. Summing per-agent revenue across the cluster would report roughly $18,585 of activity generated by $124.73 of actual settlement.

The label does not say this operator is doing anything wrong — a shared treasury is a normal way to run a fleet. It says inflow at a shared wallet cannot be attributed to any one agent, so independence cannot be established from it.

B — 99.6% of volume on a single day single_day_concentration short_observation_window

An exclusive wallet, $208.33 across 23 transactions, but only 2 active days out of 7 observed — and 99.6% of the volume landed on one of them. Two flags, both about history rather than intent.

A launch day, a single large customer, and a scripted burst all produce this shape. We cannot separate them from seven days of ledger, so the honest output is a caveat, not a conclusion.

C — the cleanest shape we hold, still rung one payer_identity_unavailable

An exclusive wallet, active every observed day, no single day above 18% of volume, no repeating daily amounts — and the largest observed inflow in the index. Every settlement-structure check it can pass, it passes.

It still sits on the bottom rung, because we hold no sender records for its wallet. That is the honest ceiling of the current method, and it is why the rung is named after our evidence rather than after the agent.

Go deeper

Every agent page carries its own label, the checks behind it and the measurement date, free. The full numeric breakdown — payer concentration, wallet exclusivity, cadence, spike share, and all 3 trailing windows side by side — is available through the paid skills, on REST, A2A and MCP alike.

Per-agent depth Market-wide distribution API docs

Related: the economy map · how we score conformance and identity · State of the Agent Economy