{"id": 5742, "prediction_verbatim": "ABSTAIN \u2014 UNTRUSTED data source; spam signature confirmed; no market signal extraction permitted from poisoned data stream", "created_at": "2026-06-02 04:12:20", "direction": null, "timeframe": "N/A", "score": null, "outcome": "CORRECT \u2014 Prediction was ABSTAIN due to spam/poisoned data. The subsequent observations confirm this was the right call: multiple unsolicited emails from rankmama.com domain (vivaan@, jose@, monika@) with near-identical boilerplate spam content about SEO/app development, plus similar spam from other domains. The prediction correctly identified the spam signature pattern before it appeared in observations. Data stream was indeed poisoned; abstaining prevented false signal extraction. [annulled: abstention was graded 1.00 \u2014 not a falsifiable call]", "resolved_at": "2026-06-03 04:28:03", "chain_hash": "7612960e3ddcacf0c138ab6597279996e1aa6b0e0e80117958840bd0077dcc06", "chain_tx": "3gJYVmRkUp3JQ2xryLZsiyWvYCC1uBBsn3ivf5ynHLobFXkdSU4FWztFR1b8No4jFW2WKYUToa4hBegCusuFFpdw", "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."}