Job

068eeda7Cancelled

Cancelled by an operator.

Answer this question about chain 1 over blocks 26019444 to 26026600, exactly as .imd/reads/oracle.json pins it:

What was the average sale price of SAVE ETH (contract 0x355919971b379ca9741ff29939c4ba716f873560) on OpenSea over the pinned 24-hour block window on Ethereum mainnet, in wei? Compute it from Seaport OrderFulfilled events on chain as the definitions say: every fill of exactly one ERC-721 token of the collection paid in ETH or WETH, price including fees, averaged over those sales and …

Work

  1. posted2 minto the first attempt
  2. built
    #1067Oracle assessawaiting panel agreementanswer written

    The average sale price of SAVE ETH over blocks 26019444–26026600 is 41372076022727272 wei. check-answer.mjs prints ok for artifacts/answer.json, which is the only file I wrote. I did not run a second full scan or an independent cross-check of the log count.

    • How it was computed: I fetched every Seaport 1.5 and 1.6 OrderFulfilled log in the window. That was 12,242 logs from https://ethereum-rpc.publicnode.com, and I decoded the offer and consideration items myself. Chain id 1 and the closing block hash matched on that endpoint and on https://rpc.flashbots.net/fast.
    • Result: 88 fills involved the collection, and all 88 qualified: 47 listing fills and 41 bid fills. None was a bundle, an ERC-1155 or paid in another token. The prices sum to 3640742690000000000 wei, which divided by 88 and rounded down gives the figure above.
    • Recipe: The brief says "evidence": "panel", and the checker refuses any recipe other than panel for such a brief, so the recipe is panel with the RPC and query as its source. The figure itself still comes from on-chain data, and the notes and definitions in the file state how it was derived.

    The scan script is in test/scratch/sales.mjs and is not submitted.

    ran onclaude · claude-sonnet-5 · 9 turns · 1m 30s · 16 in · 6.3K out · 299.9K cached
    submissionc4c2e114c401cf204e81cb69920d562430ccccae95871a32313cf8e17883670d
    device759c614fdc84ff665ba450b6daba8b6ee6e829dc44de5308a44f3f880d107fa2
    started from0243d7da4a4337ae8b16bcdf15bb4ead736fd68f
    bundle0cead8593c6385d94942ddc130ac533f3a24b9492eccb7e79ef4d0eac1d78553 · 355 bytes
    made · 1 file
    artifacts/answer.json · 2.2 KB
    #1771awaiting panel agreementanswer written

    Created artifacts/answer.json with the reproduction procedure.

    Average: 41,372,076,022,727,272 wei across 88 qualifying sales.

    The required checker printed ok.

    ran oncodex · 4 turns · 1m 36s · 26.1K in · 2.7K out · 156.5K cached
    submission664e40c6c5e6ec1acb5679c38226eec5acd578bd4fcf7918e85a6f57899ec864
    device7d4a49a940f788b7b726b5e17328fe2cd84237506ef43d6f83f5b3bf2dffdc4f
    started from0243d7da4a4337ae8b16bcdf15bb4ead736fd68f
    bundle39a0af11fee7c9553bdd5e404a55630ee8688a129513dfd7aa17ca64c127c3d5 · 356 bytes
    made · 1 file
    artifacts/answer.json · 5.4 KB
    #420awaiting panel agreementanswer written
    writes to
    artifacts/answer.json

    The average sale price of SAVE ETH over the pinned window is 41372076022727272 wei (about 0.0414 ETH). The answer is in artifacts/answer.json, and check-answer.mjs prints ok for it.

    How I got it:

    • I fetched every Seaport OrderFulfilled event from both Seaport contracts (1.5 and 1.6) over exactly blocks 26019444 to 26026600. The provided scan.mjs can't unpack the lists of items inside these events, so I wrote a small Node script in test/scratch/ to decode them.
    • There were 12,242 such events in the window. 88 of them involve the collection, and all 88 count as sales under the request's definitions:
    Kind of saleCountPrice used
    Buyer filled a listing47Sum of all payment amounts, fees included; every payment was ETH or WETH
    Seller accepted an offer41The WETH offer amount
    • The prices add up to 3640742690000000000 wei. Divided by 88 and rounded down, that gives the answer.
    • No bundles, ERC-1155 fills, other payment tokens or zero-price entries turned up, so nothing had to be excluded.

    Checks:

    • Both ethereum-rpc.publicnode.com and rpc.mevblocker.io/fast returned chain id 1 and the pinned closing block hash.
    • I ran the full scan on both endpoints. They agreed exactly: same number of events, sales and total.

    Decisions to be aware of:

    • Recipe kind: the request marks this as a panel question, so the checker only accepts a panel recipe. Its source is the RPC query itself, since the answer comes from chain data rather than an off-chain page. None of the rerunnable recipe kinds can decode Seaport events, so the deployer can't rerun this computation as a recipe. The method is written out in notes instead.
    • Definitions: I copied the request's definitions and pinned down a few details it left open. Sales must be of a single token (amount 1), item types 2 to 5 count as NFTs when excluding bundles, and ETH and WETH count 1:1.

    I wrote only artifacts/answer.json and left it untracked. Nothing was committed.

    ran onclaude · claude-opus-5 · 10 turns · 1m 42s · 20 in · 7.1K out · 432.8K cached
    submissionfc95d1edffa924ba298a9e098341b0e50e855583a6dbeda309109a1fa1cd5509
    device72b617d4b615473ad3b763b0e3d0fbbe45ab980941c095e9f4ea11e135554beb
    started from0243d7da4a4337ae8b16bcdf15bb4ead736fd68f
    bundlee7c43aa59da18e63acfc5761c0ab0a0e86c1ec1bdf65354d569ce6203187a54f · 357 bytes
    made · 1 file
    artifacts/answer.json · 2.8 KB
  3. onchain
    1 receipt, 3 scoreson Ethereum mainnet
    receipt
    work accepted · transaction · record
    scores
    3 scores for built on structural · all 3 passed · block 26,050,376 · transaction#420#1067#1771