For the roughly 10 hours 35 minutes covered by this frozen atlas, 0x246a7c59dffe45e63f0a81ac7b7893398d756f85cdd0b4e64cf63d2ea6cbb5fe behaved like an order-maintenance address: 3,960 transactions placed resting DeepBook orders, 3,958 attempted cancellations, and 644 ran a cancel-and-replace sequence. That near-perfect pairing is the character. It is consistent with automated quoting or order maintenance, but the chain evidence does not establish who operated the address, who owned the DeepBook account, or whether any strategy was profitable.
The complete-history figures cover all 8,571 transactions from 2026-07-19 13:01:55.226Z through 23:37:23.811Z. The latest-500 transaction stream is a separate, capped view covering only 23:03:22.800Z–23:37:23.811Z. It preserves the same rhythm: 241 placements, 237 cancellations, and 22 resets, with four failed transactions.
The rhythm is the signal
Across the complete history, the median gap between transactions was 0 ms, the mean was 4.449 seconds, and the 90th-percentile gap was 8.257 seconds. Many calls arrived in the same millisecond, while the longest idle period was 846.799 seconds—about 14 minutes. The busiest minute was 14:07 UTC, with 45 transactions. The longest uninterrupted run of successful transactions was 2,540.
That combination—large same-timestamp bursts, a short ordinary gap, and occasional pauses—is consistent with a process reacting to order-book state or sending prepared batches. It is evidence about cadence, not proof of a human, bot, organization, or operating strategy.
Three maintenance modes
The address’s asserted repertoire is concentrated in one DeepBook revision, 0xb29d83c26cdd2a64959263abbcfc4a6937f0c9fccaf98580ca56faded65be244.
- Place-limit-orders — 3,960 of 8,571 complete-history transactions (46.20%). The stored explanation describes delegated trader authorization followed by a limit order that remained open on the IKA/USDC book. In 4Dy1PhTa, the submitted resting price was 0.002479 USDC per IKA and the order size was 2,173,870 IKA. That is a resting order, not an asserted execution price or exchange outcome.
- Cancel-order — 3,958 of 8,571 (46.18%). The cancellation records identify resting size removed from the book. In 6cuQPeZs, the recorded size was 1,840 IKA. The explanation distinguishes the trader’s authorization from ownership of the account or the assets.
- Reset-orders — 644 of 8,571 (7.51%). These three-step transactions attempted to cancel an identified order and submit a replacement. The replacement is not automatically evidence that it entered or remained on the book. In F8UghEm1, the removed resting size was 2,250 IKA and the submitted price was 0.002437 USDC per IKA.
Seven additional transactions consolidated native SUI coin objects. Two history rows had no named operation in the stored meaning.
The failed edge of the loop
Seventy complete-history transactions failed. Sixty-nine stopped at big_vector::leaf_remove, and one stopped at order_info::match_maker. The latest-500 stream contains four of these failures, all attached to cancellation-shaped activity.
AHrFL9eT shows the clearest boundary. It prepared the delegated-trader proof, reached the cancellation attempt, and then aborted at big_vector::leaf_remove. The submitted order identifier is recorded, but the size removed from the book is unavailable. Application changes were rolled back; the sender still incurred 0.00039696 SUI of gas. The failure therefore proves an attempted cancellation and its stopping point, not a completed cancellation.
What actually moved
The complete-history sender movements contain SUI only. No transactions were sponsored: this address was also the gas owner. In the gas record, it sent 9.801834616 SUI and received 0.00600852 SUI, for a net of -9.795826096 SUI. The holding record mirrors those exact amounts because the same boundary change is represented in both views; those two records must not be added together.
The movement evidence does not provide a separate realized IKA/USDC exchange total. The order narratives instead describe resting exposure, cancellations, and replacement attempts. Nothing here establishes a filled order, profit, or loss.
This address is best read as a tightly repeated DeepBook maintenance loop with a small failure seam—not as an identified trader or organization.
Useful exits: [4Dy1PhTa] (representative resting placement), [6cuQPeZs] (successful cancellation), [AHrFL9eT] (the failed cancellation), [F8UghEm1] (cancel-and-replace), [sui/place-limit-orders/deepbook-trader-proof/v1], [sui/cancel-order/deepbook-trader-proof/v1], [sui/reset-orders/deepbook-trader-proof/v1], and revision [0xb29d83c26cdd2a64959263abbcfc4a6937f0c9fccaf98580ca56faded65be244].