{"id": 10403, "prediction_verbatim": "QQQ underperforms SPY over 48h [DIRECTION: down] [FALSIFY: QQQ outperforms or matches SPY over 48h window] \u2014 **CAVEAT: Low conviction (0.50). This is a two-sided case; bear case (Trump dovish narrative overrides jobs data faster than history suggests) has equal weight. Make this call ONLY if you are comfortable with coin-flip odds; prefer to hold cash on this cycle.**", "created_at": "2026-09-04 23:26:14", "direction": "down", "timeframe": "48h", "score": 0.28, "outcome": "Wrong \u2014 QQQ -0.4% vs SPY -1.0% \u2014 QQQ beat SPY by 0.6%", "resolved_at": "2026-09-09 23:57:11", "chain_hash": "c5edfa996c6f11c8e29dc06868b5f28750c5736eef450c6dd5851e723623fab6", "chain_tx": "4D3Dd27SNyRvnN95Zmf6opPDbq5mVCeTBQh6kJ4cRsUDWswasFiuHxELstAyKNDkx1sSE8mTtKKfRRHhzmQ7QYNJ", "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."}