READ-ONLY AUDIT SNAPSHOT · frozen 2026-10-16 18:17:04Z · no live data, no controls, no server · paper trading only, no real-money execution

JJ Vault

connecting…
SIGNER · —waiting…
Trades —
Latest call
—

System

needs the PIN · stops new buys in ≤ 5 s · exits keep running

Daemons

pick a position
    fees—realized—

    Calls · last 24 h

    Paper tests · simulated money

    Result sound · one book

    …
    One sound per resolved paper position in the selected book, on net realised P&L after the allocated purchase cost and fees. Break-even is silent. The retained runner is separate exposure and is not in the figure. Nothing plays for an entry, a ladder rung that leaves inventory, an unrealised move, a synthetic fixture, a voided position, historical replay, or a page refresh. Preferences and the already-heard list are kept in this browser only.

    Books

    Open paper positions

    Exploratory discovery · separate experiment

    loading…

    Why call volume is what it is

    loading…

    Exports

    node src/paper.mjs --report for the table · node src/paper.mjs --export writes 06-decisions/PAPER-EXPORT.json and PAPER-FILLS.csv (both scrubbed). The append-only ledger itself is data/paper/{events,decisions,fills}.jsonl.

    Spike capture · automatic execution

    loading…

    Mode

    Stage dashboard

    loading…

    Profit cards

    loading…

    Hatchan · real prices

    loading…

    Presets

    loading…

    Discovery feed · shadow comparison

    loading…

    Age at discovery, per source

    loading…

    Outcome by token age

    loading…

    Widened gate · shadow verdict

    loading…

    Coverage · coins only one feed found

    loading…

    Paper results · simulated money

    loading…

    20-trade proof · live only

    Which gate costs us · rejects audit

    what these columns mean
    gate / bucket
    the reason a call was skipped, or the band it fell in. One row per reason, so the question "which gate is costing us money" has a per-reason answer rather than one total.
    n (N seen)
    n is how many of those skipped coins have been followed long enough to grade. seen in brackets is how many were skipped in total. When seen is much larger than n, most of that bucket is still unresolved and the row is an early read, not a verdict.
    hit +25%
    the share of graded coins in the bucket that reached +25% above the price at the moment we skipped them. It is a touch, not a fill: it says the price went there, not that we could have sold there. Treat it as descriptive.
    ladder Σ
    the total simulated dollars the bucket would have returned under today's swing exit ladder, after costs. Negative means skipping that bucket SAVED money.
    scalp Σ
    the same coins under the alternative rule — sell 100% at +25%, otherwise exit at 10 minutes. It is there to compare two exit styles on identical entries.
    rugged
    how many of the graded coins collapsed outright. A bucket with a good "hit +25%" and a high rug count is a bucket you would have had to sell into, not hold.

    None of this is money that moved. It is what the skipped calls did afterwards, priced through the same cost model as a real trade.

    Latency · hit → fill