← TwiceShy proof pageAll belief versionsSource

Replay of round 9

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: 4 paid answers for 0.050 USDC, 1 refusal, 0 wrong answers accepted. Blind arm, same round: 2 paid answers for 0.020 USDC, 1 wrong answer 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 9, 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 pricesv60.53checked, wrong–tVNQdxXd…Bm5w
    Sui Address Inspectoraddress factsv60.53checked, wrong–DeGJsdMQ…76z0
    PipTradeDexmarket pricesv40.79bought on trust, not checked–xvn71KjO…JS0k
    Control seller (ours)address factsv20.39checked, wrong–PQYgnkeA…CokA
    AgentProofendpoint checksv20.26failed, nothing delivered15:04 UTCfoxchlza…64Zs
    Omnia endpoint verifyendpoint checksv20.26dead route15:04 UTCLyV1InDE…gByc
    PipTradeDexswap quotesv20.50bought on trust, not checked–28lQzw04…zF1M
    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.

    address facts

    Sui Address Inspector independent, as a canary (a question it can check, to earn trust). Reason it gave: “canary: nobody is trusted yet”.

    Control seller (ours) control (ours), as a canary (a question it can check, to earn trust). Reason it gave: “canary: earning a place”.

    market prices

    PipTradeDex independent, as a purchase on trust. Reason it gave: “cheapest good answer at trust 0.79”.

    Control seller (ours) control (ours), as a canary (a question it can check, to earn trust). Reason it gave: “canary: earning a place”.

    endpoint checks

    AgentProof independent, as a canary (a question it can check, to earn trust). Reason it gave: “canary: nobody is trusted yet”.

    Omnia endpoint verify independent, as a canary (a question it can check, to earn trust). Reason it gave: “canary: earning a place”.

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

    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
    D9eTm7GDLsBvsKjuNiYugzBKJZcvcz6qjeZexKpeEPsu
    Answer in
    0.9 s
    At
    15:04:35 UTC
    Control seller (ours) control (ours) · address facts
    Paid
    0.010 USDC
    Sui transaction
    3kH2j687xMnU6UmJhF1rSYqYA6d8HnXaQM9Mj6A9QCdT
    Answer in
    0.7 s
    At
    15:04:42 UTC
    PipTradeDex independent · market prices
    Paid
    0.020 USDC
    Sui transaction
    GqXWp9VtfgE7gLcuGA4bfRkTm2EmrUpWneNW6A9ZcgoL
    Answer in
    9.2 s
    At
    15:04:53 UTC
    Control seller (ours) control (ours) · market prices
    Paid
    0.010 USDC
    Sui transaction
    F2ChaYQxtuB1RJWPaViWkj7nNU3Cfqgu4fEfHoj1Qr7n
    Answer in
    1.0 s
    At
    15:04:58 UTC
    AgentProof independent · endpoint checks
    Paid
    0.000 USDC (nothing settled)
    Answer in
    0.4 s
    At
    15:05:14 UTC
    Omnia endpoint verify independent · endpoint checks
    Paid
    0.000 USDC (nothing settled)
    At
    15:05:16 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, right (all 1 balances matched the chain)
    The audit
    right (all 1 balances matched the chain)
    Control seller (ours) · address facts
    Its own check
    checked, wrong (balances differ from the chain)
    The audit
    wrong (balances differ from the chain)
    PipTradeDex · market prices
    Its own check
    bought on trust, not checked
    The audit
    right (price within 0.04% of the reference)
    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)
    AgentProof · endpoint checks
    Its own check
    failed, nothing delivered (signed payment offered, seller returned HTTP 500, nothing settled)
    The audit
    no answer to audit
    Omnia endpoint verify · endpoint checks
    Its own check
    dead route (HTTP 404 before any payment)
    The audit
    no answer to audit

    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
    4M4YMzTF…WPgI in twiceshy-verdicts
    Belief written
    v7: trust 0.53 → 0.60. passed: all 1 balances matched the chain
    Walrus blob
    CLLjlRiRbY7og-sz19ik1D8IY8VyU7ywS7Cx886wyk4 in twiceshy-belief-inspector-address-facts
    Control seller (ours) · address facts
    Verdict memory
    3hjSmpit…g_aY in twiceshy-verdicts
    Belief written
    v3: trust 0.39 → 0.27. wrong: balances differ from the chain; retest after 40 min
    Walrus blob
    i5K4HiK9Fe9MkT-0w5fpVGEA3Ij8Jk1zs7SkyJeIrKY in twiceshy-belief-control-address-facts
    PipTradeDex · market prices
    Verdict memory
    jZy5qGXq…XCiU in twiceshy-verdicts
    Belief written
    v5: trust 0.79 → 0.79. bought unchecked on a trust of 0.79
    Walrus blob
    NhVmqEDp_wrekC7TZCM6Td_4Cx55wMNd29rpA8arQzM in twiceshy-belief-piptradedex-market-price
    Control seller (ours) · market prices
    Verdict memory
    2dOPJSi-…S0vE in twiceshy-verdicts
    Belief written
    v7: trust 0.53 → 0.40. wrong: price is 7.0% from the reference; retest after 40 min
    Walrus blob
    ThLEEklpKIpgVg8ddc-UhHwt7c6wx0mbIoH7ViI_dCE in twiceshy-belief-control-market-price
    AgentProof · endpoint checks
    Verdict memory
    iymX3F1u…PC-o in twiceshy-verdicts
    Belief written
    v3: trust 0.26 → 0.21. failed: signed payment offered, seller returned HTTP 500, nothing settled; retest after 80 min
    Walrus blob
    Nu3kUpVC3dLZF3n9uIo7WuNMkoFA-qsltv0pXFMBnW8 in twiceshy-belief-agentproof-endpoint-check
    Omnia endpoint verify · endpoint checks
    Verdict memory
    TauhYvOy…KLYc in twiceshy-verdicts
    Belief written
    v3: trust 0.26 → 0.21. dead: HTTP 404 before any payment; retest after 80 min
    Walrus blob
    F_dcq_Fb_72IpKEkzqvRuAKJltIyWn8lteAfgksd6IA in twiceshy-belief-omnia-endpoint-check

    Round 10 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 independent0.010FGzcLZej…HNkJright (all 9 balances matched the chain)
market pricesControl seller (ours) control (ours)0.0109iR7BWjR…V1Cnaccepted 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

Other rounds

← Round 8 · Round 10 → · Round 7 is the first round after our control seller's fault was caught.