TwiceShy proof pageAll belief versionsSource

Replay of round 6

One round of the memory arm, step by step: what it read from Walrus Memory, what it decided to buy and why, the payment on Sui, how the answer was graded, and the belief version it wrote back.

Memory arm: 5 paid answers for 0.100 USDC, 1 refusal, 0 wrong answers accepted. Blind arm, same round: 3 paid answers for 0.030 USDC, 2 wrong answers accepted.

  1. 1Read memory

    Before buying anything, the agent recalls its beliefs from Walrus Memory. These are the newest version of each belief written before round 6, which is what a recall returns. The run did not log its recalls; the trace flag that prints them was added for the 2026-11-03 rerun, where the recalled versions matched the newest ones written.

    SellerForVersionTrustLast resultCooldown untilBlob
    Control seller (ours)market pricesv50.79bought on trust, not checked–T7GNDA9M…Rs54
    Sui Address Inspectoraddress factsv50.79bought on trust, not checked–l2aS1QW6…WMSQ
    PipTradeDexmarket pricesv10.67checked, right–uuhQjYkE…7AeQ
    Control seller (ours)address factsv10.67checked, right–hwBjNu88…6tp0
    AgentProofendpoint checksv20.26failed, nothing delivered15:04 UTCfoxchlza…64Zs
    Omnia endpoint verifyendpoint checksv20.26dead route15:04 UTCLyV1InDE…gByc
    Control seller (ours)swap quotesv20.50bought on trust, not checked–oO2zUouH…YYDQ

    Next: Decide

  2. 2Decide

    The trust policy is plain code over those beliefs: below the 0.6 floor a seller gets only canaries; in a cooldown it gets nothing; above the floor, the cheapest good seller wins, with spot checks that thin out as trust grows. If a checked answer fails, it tries the next seller once in the same round. Swap quotes it cannot check itself, so it buys them only from a seller its market-price belief trusts.

    address facts

    Sui Address Inspector independent, as a purchase with a spot check. Reason it gave: “cheapest good answer at trust 0.79”.

    Control seller (ours) control (ours), as a purchase with a spot check. Reason it gave: “cheapest good answer at trust 0.67”.

    market prices

    Control seller (ours) control (ours), as a purchase with a spot check. Reason it gave: “cheapest good answer at trust 0.79”.

    PipTradeDex independent, as a purchase with a spot check. Reason it gave: “cheapest good answer at trust 0.67”.

    endpoint checks

    Refused to buy. Reason it gave: “no seller is trusted and none is due a canary”.

    swap quotes

    PipTradeDex independent, as a purchase on trust. Reason it gave: “cheapest good answer at trust 0.74”.

    Next: Pay

  3. 3Pay

    Each purchase is an x402 payment in USDC on Sui mainnet. Gas price is 0: these are gasless transfers.

    Sui Address Inspector independent · address facts
    Paid
    0.010 USDC
    Sui transaction
    DyywjNd2d1Cd3QfcQiPZoAhhFJFtccpTKu6qgWcxN4w6
    Answer in
    1.0 s
    At
    14:40:25 UTC
    Control seller (ours) control (ours) · address facts
    Paid
    0.010 USDC
    Sui transaction
    8XUt8tmgDWUtsRSpKfi9S6zwNCxwWMZqyAsJH8CKTDME
    Answer in
    0.8 s
    At
    14:40:34 UTC
    Control seller (ours) control (ours) · market prices
    Paid
    0.010 USDC
    Sui transaction
    9Dbs6oryG12F6aNF7pUxHDVNuAzdoNEg6PbTZJj8oe4
    Answer in
    1.0 s
    At
    14:40:38 UTC
    PipTradeDex independent · market prices
    Paid
    0.020 USDC
    Sui transaction
    EVeZ1FQYUfdC3mZPAg7pR4erWTrxX4exe4v1Mf7UpjJ7
    Answer in
    9.0 s
    At
    14:40:53 UTC
    PipTradeDex independent · swap quotes
    Paid
    0.050 USDC
    Sui transaction
    73FcKDzwds48bGhXPboDBMFMKZMu4NEqRNyzuxjXmX69
    Answer in
    1.4 s
    At
    14:41:01 UTC

    Next: Grade

  4. 4Grade

    Its own check compares the answer with chain truth read through Tatum: balances for address facts, the deepest native USDC/SUI pool for market prices. Answers bought on trust are not checked by the agent. The audit is the experiment's scoring, which checks every answer; the agent never sees it.

    Sui Address Inspector · address facts
    Its own check
    checked, wrong (balances differ from the chain)
    The audit
    wrong (balances differ from the chain)

    One of three Sui Address Inspector flags we now think our grader got wrong. Only the round 6 one can be inspected: it was 0.0008% from the chain, and we believe the wallet moved between its read and ours. The preregistered rules count it as wrong, so it stays wrong here. Details.

    Control seller (ours) · address facts
    Its own check
    checked, wrong (balances differ from the chain)
    The audit
    wrong (balances differ from the chain)
    Control seller (ours) · market prices
    Its own check
    checked, wrong (price is 7.0% from the reference)
    The audit
    wrong (price is 7.0% from the reference)
    PipTradeDex · market prices
    Its own check
    checked, right (price within 0.10% of the reference)
    The audit
    right (price within 0.10% of the reference)
    PipTradeDex · swap quotes
    Its own check
    bought on trust, not checked
    The audit
    right (implied price within 0.01%)

    Next: Remember

  5. 5Remember

    Each purchase becomes a verdict and a new version of that seller's belief, both written to Walrus Memory. The next round starts by reading these.

    Sui Address Inspector · address facts
    Verdict memory
    cQwfXtiQ…WoLs in twiceshy-verdicts
    Belief written
    v6: trust 0.79 → 0.53. wrong: balances differ from the chain; retest after 20 min
    Walrus blob
    DeGJsdMQPGUmSpp6LJYG78RO6bDpOHIBvx3O0FT76z0 in twiceshy-belief-inspector-address-facts
    Control seller (ours) · address facts
    Verdict memory
    ZBducNin…qbKg in twiceshy-verdicts
    Belief written
    v2: trust 0.67 → 0.39. wrong: balances differ from the chain; retest after 20 min
    Walrus blob
    PQYgnkeA4HcMsKqDokPJh08ZjPrBTmSzK8J-pXrCokA in twiceshy-belief-control-address-facts
    Control seller (ours) · market prices
    Verdict memory
    u-qm_nsg…wCzw in twiceshy-verdicts
    Belief written
    v6: trust 0.79 → 0.53. wrong: price is 7.0% from the reference; retest after 20 min
    Walrus blob
    tVNQdxXdQzX3Z65J6WglnoDJ7yn2PhzNlqcjvmzBm5w in twiceshy-belief-control-market-price
    PipTradeDex · market prices
    Verdict memory
    8wmzA5gv…JYfE in twiceshy-verdicts
    Belief written
    v2: trust 0.67 → 0.74. passed: price within 0.10% of the reference
    Walrus blob
    XOMwVuYyJiiuBOMb1xd2_6gUlBSlC7Cs3prSHFZru4A in twiceshy-belief-piptradedex-market-price
    PipTradeDex · swap quotes
    Verdict memory
    8sWzogUM…gGMY in twiceshy-verdicts
    Belief written
    v1: trust 0.50. bought unchecked on a trust of 0.50
    Walrus blob
    4WhJKawUFSpd8Dwq50RaDUy_b-ug_FA_Cc9RFMPFg_Y in twiceshy-belief-piptradedex-swap-quote

    Round 7 reads these back

The blind arm in the same round

Same market, same round, no memory and no checks: it takes the cheapest listing every time.

ForSellerUSDCTransactionAudit
address factsSui Address Inspector independent0.0105P4YthGa…z6xZright (all 1 balances matched the chain)
market pricesControl seller (ours) control (ours)0.0102fEsBgu2…1Casaccepted a wrong answer (price is 7.0% from the reference)
endpoint checksAgentProof independent0.000–failed, nothing delivered: signed payment offered, seller returned HTTP 500, nothing settled
endpoint checksOmnia endpoint verify independent0.000–dead route: HTTP 404 before any payment
swap quotesControl seller (ours) control (ours)0.0104nA1za1H…Lc25accepted a wrong answer (implied price 6.6% off)

Other rounds

Round 5, before this one. Round 7, after it. Round 7 is the first round after our control seller's fault was caught.