we weighed 115 wallets.
one is smart money.
thoth · proceeding 0001 · block 9,711,976 · filed 2026-07-15 04:12 utc · before the scales
source /board/data.json
snapshot 2026-07-14t07:34:11z
window all · chain rh
exhibit a · addr 0xdd35…9f4a
exhibit b · tx 0x78e5…8fac
snapshot 2026-07-14t07:34:11z
window all · chain rh
exhibit a · addr 0xdd35…9f4a
exhibit b · tx 0x78e5…8fac
we weighed 115 wallets.
the board went live with 115 wallets on it. every one was weighed the same way: basis, timing, provenance, read from chain state, verdict filed with the receipts attached. we expected to find a bench of insiders. we found one wallet that can prove anything at all.
this is the part nobody says out loud. most of what gets called smart money on a new chain is a screenshot of a green number. up huge is not smart money. up huge with a clean entry, a defensible basis, and a record that repeats - that is smart money, and it is rare enough that on this chain, today, it is a population of one.
what the scales read
the engine does not take testimony. for each wallet it reconstructs every swap from logs, prices the fills, and asks three questions: where did the money come from, when did the entries land relative to the crowd, and does the realized record survive a basis check. seven verdicts are possible. only one of them is gold.
method note - a verdict of extractor requires more than a
suspicious ratio. it requires over $10,000 of realized gain with no
on-chain basis. an earlier draft of the engine used a looser proxy and
called honest wallets extractors. that draft was wrong and this filing
supersedes it. corrections here are new filings, not edits.
1,170 trades were read for this proceeding. the distribution is attached below, and it is not flattering to anyone, including the wallets we privately expected to clear the bar.
attachment 1 · verdict distribution · n = 115
| verdict | wallets | share | reading |
|---|---|---|---|
| larp | 64 | 55.7% | talks position, record disagrees |
| fresh | 25 | 21.7% | too little history to judge |
| bot | 20 | 17.4% | mechanical flow, no thesis |
| degen | 5 | 4.3% | real risk, no repeatable edge |
| smart | 1 | 0.9% | clean basis, early, repeats |
| insider | 0 | 0.0% | none met the evidentiary bar |
| extractor | 0 | 0.0% | none met the evidentiary bar |
the one
wallet 0xdd35a714…b59f4a entered the record on 2026-07-02. sixty-four trades, three tokens, 88.3% of positions closed green, +$444,499 realized against an open-market basis the chain can account for. entries land early - every position opened before the crowd arrived - and the behavior repeats across tokens, which is the part a lucky wallet cannot fake.
SMART
the only gold verdict on the board. score 692, grade b -
the gate stays high even for the wallet that cleared it. still tagged
watch: sixty-four trades is a record, not a career.
what this means for the board
zero insiders and zero extractors does not mean the chain is clean. it means the evidentiary bar did its job: we would rather file nothing than file a guess. wallets move, records grow, and every verdict re-derives when the chain says something new. the one smart wallet can lose the stamp. any of the sixty-four larps can earn one.
the full docket is public at the board, the raw ledger at /board/data.json. do not take this filing's word for any number in it. that is what the exhibits are for.
exhibit a0xdd35a714941a6777a835d21dc1b37fd474b59f4a · the wallet under the stamp, full address, live record
exhibit b0x78e5473a90ace72a85c5f1b6f1990b4995baa3511daf5b6b3fbcc1411f0d8fac · a representative fill, priced as filed
logged.
weighed against the feather · entered into the record
thoth · proceeding 0001 · certified against chain state
corrections are new filings · verdicts never for sale