This UTC minute contained 2,032 eligible transactions, of which 2,003 carried an asserted meaning used for behavior counts. The minute was unusually orderly by failure rate: 61 transactions failed, or 3.0% of eligible activity, versus 8.36% across the frozen atlas. It was also extremely compressed in time—2,032 transactions arrived over 59.583 seconds, with a median gap of 0 ms, a 211 ms p90 gap, and only 441 ms of idle time.
The remarkable figure was value movement, not throughput. Sender-attributed SUI balance movements totaled 1,315,186.189513108 SUI in and 339,540.444520448 SUI out across 843 transactions, for a net of +975,645.744992660 SUI. One transaction, 843NY52W, accounted for +646,019.181478112 SUI, alongside +112,183.358588 USDC and +9.685032007 CETUS. That SUI movement was roughly 158 times the frozen corpus’s per-asset 99th-percentile absolute movement.
The concrete sequence explains why this is not evidence of a trade or profit. 843NY52W was labelled a close-liquidity-position workflow. Its five calls alternated framework empty-coin construction with two collect_reward calls, then ended with pool_script::close_position. The transaction detail says the collection calls do not reveal their individual reward amounts. The evidence supports a liquidity-position closure with associated reward collection; it does not establish profit, price, or an exchange outcome.
Most of the minute was much more repetitive. Among the 2,003 semantically described transactions, 510 were limit-order placements and 508 were cancellations—1,018 transactions, or 50.8%. The largest single sender, 0xcde6dbe01902be1f200ff03dbbd149e586847be8cee15235f82750d9b06c0e04, submitted 548 of the minute’s 2,032 transactions. Its activity was narrowly split between 219 placements and 329 cancellations. The address is an on-chain identity only; the evidence does not establish who controlled it or whether it represents an automated system.
A representative placement, BcU8pcoa, prepared an owner proof and passed it to a DeepBook placement call. It submitted 812 NS at a stated resting price of 0.01681 SUI per NS. The transaction’s own explanation is careful: the call identifies a placement attempt, but does not establish that the order entered or remained on the book. Likewise, order prices in this minute are resting prices, not execution prices.
The 61 failures were scattered rather than driven by one dominant outage. The largest recorded stop sites were an unrecorded site with 11 failures, flash_swap_internal with 9, check_min_amount with 8, and swap_in_pool with 8. In 3MH4FFgm, the transaction opened short-lived flash credit, built a swap context, reached several exchange-related calls, and then stopped because the source coin balance was too small. The remaining two submitted steps did not run, and application changes were rolled back. Its reported −0.000764404 SUI boundary change was gas; the attempted route does not establish that an exchange occurred.
Finally, 518 of the 2,032 transactions were sponsored, meaning a separate gas owner paid them. That sponsorship, the millisecond-level cadence, and the concentration in repeated order maintenance make the minute resemble coordinated or automated market activity—but they do not prove a bot, shared operator, or trading result.
For a closer look, start with 843NY52W and the sui/close-liquidity-position/roster-f0137/v1 pattern, then compare it with BcU8pcoa and the sender 0xcde6dbe01902be1f200ff03dbbd149e586847be8cee15235f82750d9b06c0e04.