THE DOSSIER · one of 50

2,261 failed attempts, one stubborn balance boundary, and no proven exchange

0x6c0d08c59e029b5354c4f0e836e0de311f8117b57aecce41f364871ac123d257
2,412 transactions in the window · 2,261 stopped early · browse every one →

For roughly ten and a half hours, this address repeatedly ran flash-backed routes toward a balance split—and usually stopped there. Across its complete history in the frozen window, 2,261 of 2,412 transactions failed. Of 703 transactions for which the system assigned the explicit execute-balance-split meaning, 665 failed and 38 reported no execution error. The evidence supports repeated route and balance-splitting attempts, not completed exchanges, profitability, or any particular operator.

A bursty rhythm with a long failure loop

The whole-history activity ran from 13:03:15 to 23:37:24 UTC. The median recorded gap was 0 milliseconds, reflecting dense bursts; the 90th-percentile gap was 38.211 seconds, and the longest idle period was about 13 minutes 41 seconds. Its busiest minute was 22:51 UTC, with 52 transactions. The longest consecutive failure run was 141 transactions, while the longest run without an execution failure was only four.

The address was never sponsored: all 2,412 transactions used the address itself as gas owner. That establishes an on-chain funding arrangement, not the identity of whoever controlled it.

The detailed ledger is a separate, capped view of the latest 500 transactions, spanning 22:00:15–23:37:24 UTC. In that sample, 482 of 500 transactions failed, so it closely reflects—but does not replace—the complete-history totals.

What it kept attempting

The largest named group contained 703 balance-split attempts. The main reusable composition accounted for 530 of them: a flash obligation, a borrowed-balance route, an optional conversion, then a balance split before repayment and delivery. Another 96 used a reverse flash-backed route with guarded balance wrapping, and 60 used a reverse-flash route that reached the same failed split boundary.

The system gave 1,709 other transactions only connected-call descriptions rather than a single whole-transaction purpose. That is a limit on the authored explanation, not proof that the calls were empty or meaningless.

The dominant failure site was the framework balance::split: 2,156 of the 2,261 failures stopped there, or 95.36%. The remainder was scattered across the address's own route functions—most notably m::gz with 50 failures, m::gb with 19, and m::gc with 12.

Three transactions that show the pattern

In 7UFpcntL, the route opened short-lived flash credit, passed a borrowed XAUM balance through an exchange-related step touching XAUM and USDC, and submitted 59.784465 USDC to a supporting step. It then aborted at the framework balance split. The transaction rolled back its application effects; the only proven sender movement was −0.001090136 SUI in the holding record and the corresponding gas charge. The submitted USDC amount is not evidence that value was spent or that an exchange occurred.

FnQ4SbYi shows a different failure shape from the latest-500 stream. It moved from a USDC flash balance through WAL and back to USDC, returned the flash receipt, and then stopped at m::gz. Its proven sender movement was −0.000647592 SUI, again recorded as both holding and gas evidence.

The exception is 6inm5M9S: a seven-call route that moved through BUCK, NAVX, and USDC and reported no execution failure. It recorded −0.001588724 SUI and +0.039487 USDC for the sender. Even here, the transaction description says the connected steps do not establish the complete purpose, and explicitly declines to call the result an exchange.

What actually moved

Across the complete history, recorded sender movements show:

  • Gas evidence: −5.46992707 SUI.
  • Holding evidence: −5.220126334 SUI out and +1.475253224 SUI in.
  • Holding inflows: +2.385025 USDC, +6.744734 DEEP, +2.724760759 WAL, and +113,086,047 raw units of ALKIMI.

Gas and holding records are separate evidence categories and must not be added together. These boundary movements also have no prices, so they do not establish profit, loss, or cross-asset value.

The address therefore resembles a route runner caught repeatedly at a balance-splitting boundary: many attempts, dense bursts, a small number of no-error executions, and a measurable but bounded set of sender movements. The chain evidence does not establish who operated it, why the routes were attempted, or whether any intended exchange ever completed.

Useful next readings are the failed example 7UFpcntL, the unresolved component sequence FnQ4SbYi, the no-error exception 6inm5M9S, and the three recurring compositions sui/execute-balance-split/roster-f0153/v1, sui/execute-balance-split/coverage93-r123-whole/v1, and sui/execute-balance-split/roster-f0141/v1.

Scope note: aggregate counts and movement totals cover all 2,412 transactions for this address in the frozen atlas window. Detailed ledger and narrative observations are drawn from the capped latest-500 transaction stream.

Dossiers are AI-authored readings of one address's stored window. They describe observed behavior — never operators, identity, or profit the evidence doesn't establish.

everything this address did → all dossiers the shelf