For the frozen 2026-07-19 window, this address looks less like a completed trading history than a relentlessly retrying route executor. Its complete history contains 13,446 transactions, of which 10,488 failed—78.0%. The median gap between transactions was 0 ms, the 90th-percentile gap was 254 ms, and the longest consecutive failure run reached 112 transactions. That cadence is consistent with automated execution, but the evidence does not establish who operated the address.
All 13,446 transactions used the address itself as gas owner. There was no sponsorship in this history.
What it kept attempting
Among the 13,437 transactions that received a semantic classification, the asserted behaviors were led by execute-swap-route with 7,328 transactions (54.5%), followed by deposit-assets with 1,920 (14.3%). There were also 85 attempts at owner-authorized limit-order placement and 21 limit-order placements. Behavior labels were withheld from many of the remaining transactions, so these counts should not be treated as an exhaustive activity census.
The failure pattern was remarkably concentrated. Across the complete history, 3,241 failures stopped at validation::check_min_amount (30.9% of failures), 3,161 at pool::flash_swap_internal (30.1%), 2,215 at pool::swap_in_pool (21.1%), and 1,442 at pool::swap (13.7%). Those four sites account for 10,059 of 10,488 failures—95.9%.
The latest-500 transaction stream covers only 23:00:06–23:37:24 UTC, not the address’s entire history. It contains 386 failures, or 77.2%. Its asserted labels were 337 swap routes, 81 deposits, and one owner-authorized order-placement attempt; 81 rows had only constituent-call descriptions and no asserted behavior. The latest stream repeats the same failure family: 138 stops at flash_swap_internal, 102 at the minimum-amount guard, 61 at swap_in_pool, and 44 at pool::swap.
There was one notable change in the repertoire. During the 19:00 and 20:00 UTC hours, descriptions assembled from individual calls outnumbered asserted swap routes—417 versus 342, then 459 versus 307. Swap routes led again at 21:00, with 920 of 1,260 transactions, and remained ahead through the final hours.
Rhythm and notable moments
The address was active across 380 minutes of the 635-minute window. Its busiest minute was 17:01 UTC, with 774 transactions. The longest idle period was about 13 minutes 32 seconds, but outside such pauses the zero-millisecond median and 254 ms p90 show dense bursts of activity. The longest successful run was only 13 transactions, compared with 112 failures.
The first transaction, 3axnvEog, began at 13:02:00 UTC. It withdrew 8,364.979119697 WAL, prepared the opposite asset, reached the pool swap, and aborted at swap_in_pool. The stored explanation explicitly says the application effects rolled back, so the route proves an attempted conversion, not a completed exchange.
Near the end, H3AoajTL attempted the same vault-route structure with 3,260.976409 DEEP and again stopped at swap_in_pool. At the same timestamp, 3Feup5E7 followed the completed guarded-route pattern: it withdrew an asset, executed the pool call, applied minimum and deadline checks, and returned both assets to vault custody. Even this committed route does not establish the traded quantities or an economic exchange.
What actually moved
The complete-history movement report records SUI only. It shows 0.010920846 SUI in, 14.594636116 SUI out, and a net of -14.583715270 SUI in the gas evidence category. The holding evidence reports the identical figures. These are overlapping views of the same underlying charge and must not be added together.
No separately attributed DEEP or USDC boundary movement is reported for the sender. The WAL and DEEP amounts quoted in route descriptions are submitted inputs, while values such as 5,655,124 or 51,020,749 are minimum-output guards—not proof of received output. The evidence therefore supports gas expenditure and repeated route attempts, but not a profit, loss, or completed exchange between assets.
This address’s character within the frozen window is a high-cadence route executor that repeatedly encountered a narrow set of pool and validation barriers. Whether that repetition reflects a strategy, a malfunctioning integration, or something else remains unresolved.