One route, two outcomes

For roughly ten hours, 0x755093…54fe2dc did one thing: it executed a route-shaped transaction 1,459 times. The route succeeded structurally 1,174 times and failed 285 times. Every failure stopped at 0xd07533…486e9c::pool::swap. The regular 2.433-second median gap is consistent with automated execution, but the evidence does not establish who operated the address.

These figures cover the address’s complete history in the frozen window, from 13:02:00 to 23:37:22 UTC. The detailed examples below come from the separate latest-500 stream, which covers only 21:40:27–23:37:22 UTC.

The rhythm

The complete-history stream had a 2.433-second median gap, a 26.147-second mean, and a 61.280-second 90th-percentile gap. Its longest idle period was about 31 minutes. Activity peaked at 23 transactions during 22:34 UTC. The longest uninterrupted successful run was 314 transactions; the longest failure run was 97.

The latest-500 stream looks harsher: 269 of 500 transactions failed, or 53.8%. Those failures were not scattered across unrelated calls. They all stopped at the same pool::swap location. The largest latest-stream failure run lasted 97 transactions, from EFwwx2hd through AaaknRkQ.

The repertoire

There was only one asserted behavior across the complete history: execute-swap-route. Five exact route compositions supplied that label:

• roster-f0020/v1: 1,150 of 1,459 transactions
• coverage93-r054-whole/v1: 152
• coverage93-r028-gateway-reached/v1: 134
• packed-e118-d003c8577562-718b895db74e/v1: 14
• packed-e118-e54c9a3f7c79-30537ddc4cf4/v1: 9

The latest 500 contain all five compositions, but their mix shifts toward the two coverage variants. This is variation in route construction, not evidence of a change in the address’s broader activity: every hourly bucket still contains only execute-swap-route.

A successful example, 4Ck2Aaej, used the framework coin helper at revision 0x000000…000002@56 and then called gateway::swap_assets in package revision 0xd07533…486e9c. Its boundary movements were −200 WAL, +5.779774 USDC, and −0.001553152 SUI. The stored explanation explicitly says those movements do not establish an exchange.

A failed example, 6z9MBhzv, used the same two-step shape but stopped at pool::swap with abort code 1010. Application changes rolled back; the reached call proves an attempted route, not a completed economic effect. The latest transaction, CGELWgt1, repeated the successful shape with −200 WAL, +5.749351 USDC, and −0.001552152 SUI, again without establishing an exchange.

The money

Across the complete history, sender-boundary movements show:

• SUI gas accounting: 1.135038344 SUI in and 6.820210704 SUI out, net −5.685172360 SUI.
• WAL holding movements: 118,600 WAL in and 116,200 WAL out, net +2,400 WAL.
• USDC holding movements: 3,391.077235 USDC in and 3,456.913372 USDC out, net −65.836137 USDC.

The SUI charge also appears in a separate holding record. Those accounting categories are not additive. These boundary movements describe what was observed for the sender; they do not establish profit, an exchange, or the economic outcome of each route.

All 1,459 transactions list this same address as gas owner, so none were sponsored. The address therefore presents a narrow, highly repetitive picture: one route family, machine-like timing, a persistent failure point, and substantial SUI expenditure concentrated in failed attempts. Whether that reflects a strategy, a misconfigured caller, or repeated testing remains unresolved.

Exits

Start with successful route 4Ck2Aaej, failed route 6z9MBhzv, or the latest successful route CGELWgt1. The reusable route cohort is sui/execute-swap-route/roster-f0020/v1, and the recurring failure site is 0xd07533…486e9c::pool::swap.