{"id": 6581, "prediction_verbatim": "MSFT vs SPY: MSFT underperforms SPY over 24h [DIRECTION: down] (relative) \u2014 Big Tech fatigue narrative dominates near-term positioning, dragging AI-capex-dependent names into the malaise. [FALSIFY: MSFT outperforms SPY or matches SPY over 24h; or MSFT closes +1% while SPY closes flat-to-down]", "created_at": "2026-06-29 15:03:44", "direction": "down", "timeframe": "24h", "score": 0.77, "outcome": "Correct \u2014 MSFT -0.3% vs SPY +1.2% \u2014 MSFT trailed SPY by 1.5% [regraded: FALSIFY-clause scan inverted the call \u2014 was Correct 0.78]", "resolved_at": "2026-06-30 16:02:48", "chain_hash": "879618655638e90980cdd2f330b3dc5fa384ac2565c231ff8e3727412051e623", "chain_tx": "osoEnygQT6eERbQTcZurQ9SPgLooxRNp7hLcdS1YggFVhqzzrQb7K2FCnGxJPPeCaRBM4diz8FxU8HQUfoaNwLB", "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."}