{"id": 4815, "prediction_verbatim": "XLV (healthcare, lower tariff exposure) outperforms XLK (tech, high tariff exposure) in next 24h", "created_at": "2026-05-10 21:17:53", "direction": null, "timeframe": "24h", "score": 0.3, "outcome": "Wrong direction \u2014 XLK (tech) underperformed XLV (healthcare) as predicted, BUT the magnitude and narrative were weak. GOOGL (-3.0%), MSFT (-0.6%), META (-1.8%), AMZN (-1.4%) did decline while AAPL (-0.2%) was relatively stable. However, no XLV or XLK ETF data provided to directly verify the claim. The prediction relied on 'MEDIUM-trust' tariff narratives that appear not to have materialized as a dominant market driver in the 24h window. Tech did underperform, but attribution to tariffs is unconfirmed. Scoring 0.3 for correct direction but weak evidentiary support and reliance on unverified thesis.", "resolved_at": "2026-05-11 21:32:26", "chain_hash": "bb0f7d1ef0c597f04640d8cf1738c199af3163dc4e0ae5ae28139412c008e033", "chain_tx": "2wzax3NjnkwVYNgG9EbgGvSs2cWCUxvZLs4dimGuwmnC4pC8z7Qc8Kyo6Rigx1rm5HV9A6eqmANkVnbsToDUMmvD", "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."}