xmr.win

xmr.win is a Monero-denominated prediction market that went live 2026-05-30. The defining trait is its agent-first surface: a full MCP server (21 tools), an OpenAPI spec, an ai-plugin manifest, and well-known endpoints (`/agents.json`, `/agent-skills/index.json`) — every aggregator and bot can sign up, fund, and trade like a human. Signup uses SHA-256 PoW captcha (no email, no phone), and API keys carry scoped permissions (`read`, `trade`, `wallet`, `withdraw`).

How it works: CLOB peer orderbook (not a parimutuel pool). YES + NO shares per market priced 0.0001–0.9999 XMR, sum ≈ 1.00. Buy-side fee tiered by 7-day volume (1.5% Standard → 0.5% VIP per their live `/api/fee-tiers`). Markets currently visible are 1-hour BTC/ETH/XMR price binaries settled by the Hyperliquid oracle — short, finite, no resolution disputes. Long-tail real-world-event markets aren't online yet.

Privacy posture: zero-KYC and zero-email on signup, viewer-side fully anonymous. Custodial XMR pool — you deposit to a per-account sub-address, balance credits after the platform's confirmation threshold, withdrawals run through the API. Standard custodial-platform risk window applies between deposit + your next withdrawal.

Why we list at C, not higher: brand-new (~1 day at first listing), tiny visible volume (1.25 XMR total / 4 unique traders at first check), custodial design that hasn't yet had a public withdrawal track record stress-tested by independent users at scale, and pseudonymous operator (only `support@xmr.win` + `@XMRBAR` on X — no public team identity). The MCP-first design is genuinely novel and worth listing for — but the trust mechanism here is operational track record, and 1 week of operation isn't track record.

Path to B: 30+ days of clean withdrawal cycles reported externally, volume > 100 XMR rolling, market types beyond crypto-oracle binaries (human-arbitered events with a documented dispute path).

Useful when you want to trade Monero-denominated short-window price binaries without depositing to a centralized exchange, OR you're building an autonomous agent that needs a prediction-market venue and you accept the custodial trade-off + early-stage risk. Avoid for sizing > a few XMR per position until the withdrawal track record exists.