This generates trading signals by running neural-trader's anomaly detection engine against stock tickers, then enriches them with neural predictions and historical pattern matching. It automatically classifies six anomaly types (spikes, drifts, flatlines, oscillations, pattern breaks, and cluster outliers) with Z-score thresholds, then ranks them into actionable signals with entry, stop, and target levels. Signals expire after 24 hours, which makes sense since intraday alerts shouldn't clutter your memory store. The strategy filter system lets you save and reuse custom scan configurations. Useful if you want systematic signal generation without writing your own screener, though you'll still need to validate the signals against your own risk rules before trading.
npx skills add https://github.com/ruvnet/ruflo --skill trader-signal