I've written the report to artifacts/report.md. It lists ten contested Polymarket resolutions, each turned into a yes/no question you can ask the oracle.

Each entry has a short question, a longer version that includes the market's own rule text, and what Polymarket actually decided. Notes under each are labelled as fact (linked to a source), my inference, or uncertain, and the report ends with the questions I couldn't answer. Outcomes, rule text, volumes and dispute histories come from Polymarket's own public market data, pulled today. I cross-checked every result and volume in the report against that data.

#	Ask the oracle	What Polymarket decided
1	Did Zelenskyy wear a suit between May 22 and June 30, 2025? ("suitgate", about $242M bet)	No
2	Was the US–Iran ceasefire officially extended by April 22, 2026?	No
3	Did the US and Iran agree a permanent peace deal by June 15, 2026?	Yes
4	Was TikTok banned for most Americans before May 2025?	Yes
5	Did the Trump administration declassify UFO files in 2025?	Yes
6	Did the US invade Venezuela by January 31, 2026?	No
7	Did the US and Ukraine agree a rare-earths deal by March 31, 2025?	Yes
8	Did Edmundo González win Venezuela's 2024 election?	Yes
9	Was the Titan submersible's passenger cabin found by June 23, 2023?	Yes
10	Is there a preponderance of evidence that Barron Trump helped create the $DJT token?	No (Polymarket publicly said UMA got it wrong)

Things to know before you use it:

• The ranking is my judgment. There's no official list. I picked markets with press coverage of the dispute itself, a UMA dispute or Polymarket intervention on record, and real money involved. Two near misses (Fort Knox gold, Israel–Hezbollah ceasefire) are noted at the end of the report.
• Not every case went to a UMA vote. The Venezuela invasion market was decided by Polymarket's own clarification of what counts as an invasion, with no dispute filed. I kept it because of the size of the backlash.
• Whale-voting claims are unchecked. Figures like "one holder cast 25% of the votes" come from news reports and third-party analysts; I didn't check them against blockchain records.
• Some details are thinly sourced. The reported refund to Barron "Yes" bettors comes only from a Manifold market's notes. The Iran peace deal dates differ between sources.
• There's no "correct" answer list. The report gives Polymarket's decisions only as a baseline to compare the oracle's answers against.
• No independent review. I did this research alone.

The artifacts/ folder is git-ignored, so nothing was committed. Besides the report, it holds a short README.md and a trimmed copy of the Polymarket data in evidence/; the report links to that copy. Only the report is a named output, so the other two may not be uploaded with it.