{
  "direction": "long",
  "leverage": 800,
  "take_profit_bp": 12,
  "take_profit_price": 81452.1254,
  "bust_distance_bp": 7.7,
  "bust_price_estimate": 81291.857035,
  "expected_net_return_on_margin_pct": 71.743709,
  "confidence": 0.55,
  "regime": "trending_up",
  "key_levels": {
    "support": 81290,
    "resistance": 81357
  },
  "entry_mid": 81354.5,
  "data_time_utc": "2026-10-17T17:59:11.866512+00:00",
  "reasoning": "Hyperliquid https://api.hyperliquid.xyz/info candleSnapshot and l2Book snapshots show a bid/ask of 81354/81355, a 13.54 bp rise over 30 minutes, rising 1m/5m EMA9 slopes, and 1m return volatility contracting to 1.10 bp over 15 minutes versus 2.18 bp over the available hour; I infer short-horizon upward continuation despite falling 15m EMAs. At 800x, the 7.7 bp bust buffer allows more retracement than 900-1000x while the 12 bp target yields about 71.74% net margin return conditional on winning, after the supplied haircut and fee. The 12 bp target requires a breakout above 81357 but is closer than 15-20 bp targets in this quiet regime; this is an inference, not an assured 30-minute outcome. Confidence is only 0.55 because price is extended above its 1m EMA9 near 81335, visible asks outweigh bids, and trade candles cannot establish historical mid-price liquidation probabilities.",
  "risks": "Rejection at 81357 or the 24h VWAP near 81420 could reverse the rebound into the broader 15m downtrend; a decline to approximately 81291.86 liquidates the full margin. The book is shallowly sampled and mutable, news risk is unquantified, and no calibrated win probability or positive unconditional expected return is established."
}