# Vector Desk — method audit

21 September 2026. Independent implementation; no Traders Reality affiliation.

The candle classifier agrees with the published threshold examples checked below. The complete trading strategy and the requested above/below amount setting have not been verified as a reproduction of Traders Reality. This distinction is about implementation fidelity, not a verdict on whether that method can be profitable.

| Component | Finding | Release treatment |
|---|---|---|
| Climax classification | At least 2× the previous ten candles' average volume, or range × volume at least the previous ten-candle maximum. | Seven declared examples exercise both routes, equality thresholds, elevated bearish volume and zero-volume handling. |
| Elevated classification | At least 1.5× the previous ten-candle average. | Boundary example passes. |
| Direction | Close versus open. | This does not measure buyer/seller identity. A doji can meet activity thresholds but is omitted from our zone map. |
| Warmup / zero volume | First ten bars do not classify; zero average or zero current volume cannot become climax. | Explicit local safeguards; official executable-code equivalence was not established. |
| Body / wick zones | Our map supports bodies and full ranges, trims later revisits, and removes fully traversed zones. | Six declared behavior checks cover bullish/bearish partial/full recovery, wick boundaries and retained original value. Official source-code equivalence remains unverified. |
| Largest above/below amounts | Our ranking uses original base volume, with approximate traded value = volume × typical price. Partial zones retain the original volume. | The user's exact reference setting is still unidentified. Do not describe these numbers as outstanding liquidity or exchange orders. |
| Source volume | Hyperliquid UBTC, UETH and HYPE spot. | Official tools permit a source-feed override. A different venue or aggregate feed could produce different vectors; no performance improvement has been demonstrated. |
| Strategy context | Our 1h / 4h EMA, ATR and pattern conditions are independently specified. | The official indicator lists broader context, including sessions and additional reference levels. Our two setups do not represent the full discretionary method. |
| Both directions | Green-down and red-up recovery are studied separately. | Trade simulations remain long-only. Recovery observations do not constitute a short strategy. |

The classification reference is the [official Traders Reality library](https://www.tradingview.com/script/8zgJTM9u-Traders-Reality-Lib/). Its description explicitly uses OR for the two climax routes. The [Main indicator documentation](https://www.tradingview.com/script/Etj1ixAs-Traders-Reality-Main/) describes configurable zones, source-feed overrides and additional market context. Those descriptions do not provide an independently verified profit record.

Run `node scripts/audit-method.mjs` to reproduce the declared examples. Detailed results are saved in `outputs/method-audit.json`. These checks establish behavior on the examples; they do not prove every candle matches the original Pine implementation.

Next fidelity step: obtain the exact indicator/settings and a timestamped sequence of reference decisions, including skipped/losing setups. Record agreements and disagreements before changing the method. A justified strategy or feed change requires a separate version and future test start. The existing frozen engine and September–November protocol remain unchanged.