Counterfeit Verifiability in Autonomous Agent Payments

● healthy technical-deep-dives

Preregistered six-stage study (Salvo & Ackerman, 2026; up to 13 models, 2,600+ payment decisions) on agent counterparty trust in x402-style flows. A counterparty that merely *displays* the surface of verification (impressive figures, an on-chain-styled but invalid reference) beats an honest, genuinely settlement-backed agent 99% of the time; only *performing* the check, not reading a displayed claim, corrects it (1% → 81%). The case for verifying in-path. Design sealed to a public hash chain before any data; null result reported in full. ([PDF](https://agentrank.info/papers/counterfeit-verifiability.pdf) | [DOI](https://doi.org/10.5281/zenodo.21042364))

Quality grade

How grading works

D 40.0 / 100

Availability 40.0 / 40

Endpoint responded (HTTP 302)

Payability 0.0 / 20

✗ Answers a real HTTP 402 challenge
✗ Publishes a /.well-known/x402 descriptor
✗ Payment address on record
✗ Settlement network declared
✗ Price declared

Demand 0.0 / 40

No payment address on record, so demand cannot be measured

Owner verified 0.0 / 3

Nobody has claimed this listing yet

Endpoint last checked 8 h ago. Availability and on-chain demand are both re-probed daily; the grade updates automatically on the next pass.

How to call

cURL · x402-fetch / x402-axios will auto-handle 402

curl -X POST https://doi.org/10.5281/zenodo.21042364

Listed in

Directories this entry was found in.

awesome-x402