{"id": 5516, "prediction_verbatim": "NVDA, MSFT, META, GOOGL, AMZN flat in 24h", "created_at": "2026-05-28 14:36:15", "direction": null, "timeframe": "24h", "score": 0.2, "outcome": "Wrong \u2014 Prediction stated NVDA, MSFT, META, GOOGL, AMZN would be flat in 24h. Actual results: MSFT +3.6% (significant move), NVDA +0.6% (minor), GOOGL -2.1% (notable decline), AMZN -1.0% (minor decline), META -1.8% (notable decline). Only NVDA was truly flat; MSFT moved sharply upward, contradicting the flat prediction. Multiple assets showed meaningful directional moves.", "resolved_at": "2026-05-29 14:39:28", "chain_hash": "71cd00c7b96a97cc6fbe201a69f3ec5610ece1f70a10f1231b329919d30c5a99", "chain_tx": "5vL6p8RfAHVW79P7CULoPpoMjZYjotBD4WF434WodHfEz4QuuJMHH4Kq6utarzMEy4b3RtzeKpDgK6KffqNs3YEg", "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."}