{"id": 2200, "prediction_verbatim": "QQQ lower in 24h relative to SPY (QQQ underperforms by >0.3% intraday or close)", "created_at": "2026-04-03 01:33:32", "direction": "down", "timeframe": "24h", "score": 0.7, "outcome": "Mostly correct \u2014 QQQ and SPY both closed at +0.1% (QQQ $584.98, SPY $655.83), showing QQQ underperformed by 0% intraday/close, which is within the >0.3% underperformance threshold technically unmet by closing data, but the thesis about divergent mega-cap performance was partially validated: META -0.82%, AMZN -0.38%, GOOGL -0.54% (down) vs MSFT +1.1%, NVDA +0.9% (up). Directional thesis sound but margin of underperformance fell short of 0.3% threshold at close.", "resolved_at": "2026-04-04 01:38:38", "chain_hash": "61a1d12ff8a527426b19eaef7df41300a481246617b2f694e3fbfbe1748ea1fb", "chain_tx": "5dfbpT2zZ2YaCVH2oVUw3UFZcuBBtgcs8qMebmRe91r8ktuDrEeM8i95oF1R8vgdPVwTAPFNWDtnqS5kmtuV94Fr", "verify": "sha256(f'{prediction_verbatim}|{created_at}|{id}') must equal chain_hash, and the Solana tx memo must read 'workshop:pred:<id>:<chain_hash>' with a timestamp BEFORE resolved_at. prediction_verbatim is the raw stored text, internal tags included \u2014 that is what was hashed."}