Across its complete frozen-window history, this address repeatedly assembled transactions that opened flash-credit positions and passed coins through exchange-related calls. The activity resembles automated route execution, but the evidence does not identify a person, company, operator, bot or strategy—and it does not establish profitability.
The complete-history figures below cover all 1,206 transactions from 13:02:01 to 23:37:29 UTC on 19 July 2026. Recent examples and recent counts use the separate latest-500 transaction stream.
The rhythm was fast, then repeatedly broke
In the complete history, 738 of 1,206 transactions failed—61.2%—and the longest consecutive failure run was 17 transactions. The median gap was 1.933 seconds, although the 90th-percentile gap stretched to 88.403 seconds. The longest idle period was 997.803 seconds, and the busiest minute was 17:01 UTC with 42 transactions.
The failure record is unusually opaque. Of the 738 failures, 735 have no more specific recorded abort location. Two were attributed to price_oracle::get_valid_price, and one to pool::simulate_buy_swap. A reached call proves an attempted step, not a surviving exchange: when a later command aborts, application changes are rolled back.
Only 17 of the 1,206 transactions received a whole-transaction route explanation. The other 1,189 were described through individual connected call steps or incomplete compositions. The repertoire stayed broadly similar throughout the window: every transaction touched the same main package revision, while the route-labeled cases split between two variants, used 13 times and 4 times respectively.
What the recent 500 reveal
The latest-500 stream begins at 18:54:01 UTC and ends at 23:37:29 UTC. It contains 288 failures and 212 transactions without an execution error. Only 10 of these 500 received an asserted route label; the remaining 490 did not receive a whole-transaction purpose. Their 185 distinct ordered call sequences show a family of related attempts rather than one perfectly repeated script.
The recent stream also shows the same cadence: a 1.686-second median gap. The largest recent gas charge was 0.5 SUI in 7fn6iMf1; the ledger records an equal-sized SUI sender movement alongside it. Those evidence categories overlap and must not be added together.
Three transactions show the boundary clearly:
6bMSEA3DziiAXR3RHdcLCkwpw4bVFmnwikDmGJRPfwqWreceived the clearest route explanation. It opened flash funding, created a zero-valued coin for a DeepBook step, passed assets through a router, returned the flash obligation and reorganized the remaining asset. The transaction detail still says the exchanged quantity is transaction-specific and does not establish an exchange outcome.2kNJ6ZtmP8zZzZETwcAg5mrNrbLU2ESDNCBNeKjQ1Rn7opened flash funding, reached DeepBook and a router-related call, then stopped at command 4 because a source coin balance was too small. The application changes were rolled back. Its boundary record shows −0.327955844 SUI in sender movements and the same amount in gas accounting; those figures are not additive.7DUx1rZHxsPSqhQDVpUvwzW8L2ePCMwsDyoQaWxPqLAJdid not report an execution error, but its connected steps still did not establish the complete transaction purpose. The boundary record shows −0.007370684 SUI and +0.005203 USDC for the sender. That is a bounded movement observation, not proof of a completed trade or gain.
The money is real, but the economics remain unresolved
Over the complete history, gas accounting records −11.700682048 SUI. Separate sender-boundary SUI movements record +13.0632966 SUI in and −9.730423169 SUI out, for a net +3.332873431 SUI. Other sender-boundary inflows total +5.575336 USDC, +9.116732 DEEP, +1.2628944 WAL, +0.204418 USDSUI and 870 raw units of an XBTC asset whose display denomination is unavailable.
In the latest 500, gas accounting totals −8.417394308 SUI. Sender-boundary movements separately record −4.646597549 SUI net, +3.277569 USDC, +0.6665862 WAL, +1.648447 DEEP, +0.113392 USDSUI and 431 raw XBTC units. These assets have no prices in the corpus, and the figures cannot establish cross-asset profit.
No transaction in the complete history was sponsored: the gas owner matched the sender on all 1,206 records.
The durable character is therefore not a confirmed trader or a named operator. It is an on-chain address producing frequent, flash-funded, multi-venue route-shaped attempts, with a high failure rate and only a small subset receiving a complete semantic explanation. The evidence supports repeated execution consistent with automation; it does not resolve who or what operated the address, whether any route completed an exchange, or whether the activity was economically successful.
Useful exits are 6bMSEA3DziiAXR3RHdcLCkwpw4bVFmnwikDmGJRPfwqW for the clearest route, 2kNJ6ZtmP8zZzZETwcAg5mrNrbLU2ESDNCBNeKjQ1Rn7 for the balance failure, and 7DUx1rZHxsPSqhQDVpUvwzW8L2ePCMwsDyoQaWxPqLAJ for a non-failed but unresolved composition.