Coinversa

Prediction markets just went DeFi-native. HIP-4 went live on @HyperliquidX May 2 — outcome contracts that share a margin account with perps and spot. The first prediction-market venue with that. 6 days in: 5,060 wallets, $30.3M USDH, 348K fills. A thread on what we're seeing

@Havochl_ He's been cooking this month!

Loracle has been insanely profitable on Hyperliquid in the last month gaining +$16M and now sitting at $40M all-time PnL. His hyperliquid:native hedge short is already in $2M profit and it’s biggest position out of all three: hyperliquid:native ~$50M, $BTC ~$12.6M and $SNDK ~$7.6M. But the most interesting one is $SNDK short, he literally shorted the very top with an entry at $1622 - which makes me wonder this might be his next biggest win after printing around $8M on $ZEC long.

Coinversa MCP 0.7.0 is live. As prediction markets move DeFi-native on @HyperliquidX , we’re one of the first to ship agent-native intelligence for HIP-4 outcome markets. 12 new tools. 55 read-only tools total. Agents can now ask: “Which BTC > 80K YES holders are also long BTC perp?” Built on Coinversa’s existing @HyperliquidX perp intelligence, 0.7.0 connects outcome markets to underlying perp context: traders, settlements, overlap, directional exposure, and hedging. free API KEY: coinversa.ai/developers Remote MCP: mcp.coinversa.ai/mcp Local: npx -y @coinversaa/mcp-server@0.7.0

@ryandcrypto nothing to see here except $106.75M of shorts stacked above mark on xyz:SP500, vs $36.59M longs below. including one $40.3M cluster sitting at $8,046. that's the fuel.

nothing to see here

@Yumiko_web3 Let’s see 😊

@coinversaa It is so much promising Trust me

$HYPE setup right now: 7,785 longs (avg $43K) vs 3,511 shorts (avg $152K). Shorts are -$60M unrealized. Longs are +$44M. But the most profitable cohort on Hyperliquid is the one short, net -$189M, and they haven't blinked. Squeeze fuel meets conviction. Who breaks first?

To the 0x128e wallet's 40x short with liq at $82,236 — yeah, with BTC at $81,062, that trade is still alive. But barely. About 1.4% away from getting torched. $126M of short notional sits in a single bucket at $85K. Squeeze fuel is loaded. 2/2

If $BTC pushes up just ~5% and triggers that $85K cluster on @HyperliquidX , the cascading short cover could blow the price meaningfully higher on its own momentum. That's exactly the kind of setup that creates the "BTC pumps for no reason" headlines a day later.

Big context. Yesterday Loracle held $52M long across BTC, ZEC, TON. Today: BTC closed, TON closed, ZEC down 98%. The +$29M lifetime bull just collapsed his entire long book to concentrate it all into a $53M HYPE short. Not hedging, converging. Posted his battle vs 0x7fda yesterday, it just got sharper.

Loracle added around 30% more to his $HYPE short since yesterday, bringing the position to $53 M This also makes him the second biggest position on $HYPE itself, just behind the $59M long interesting

@degennQuant x.com/coinversaa/sta…

@degennQuant If you're right about the giga candle, here's where the fuel comes from: the two top-cohort wallets on HL are net $52M short HYPE going into this. Money_printer + leviathan tier, both. Posted the breakdown this morning.

$HYPE weekly still healthy af Patiently waiting for that giga candle that takes us past $50 Then the fun begins Hyperliquid

The two most profitable wallets on @HyperliquidX are betting against each other on every major. Except one. Loracle (+$29M) is long BTC, ZEC, TON. 0x7fda (+$130M) is short BTC, ETH, ZEC, TON. Both are short HYPE. Combined $52M. Bull thesis and bear thesis converging on the same trade. That's the signal.

Hey mate, super easy: 1. developers.coinversa.ai → sign up → create an API key (free tier on signup) 2. Add the remote MCP to Claude: mcp.coinversa.ai/mcp 3. Want to self-host? npx package + full docs at Coinversa.ai/developers 43 Hyperliquid tools live in your agent in ~2 mins. Let us know how you go!

@coinversaa Love it, mate. Way quicker than rebuilding an indexer from scratch and all those tools sound unreal. Any tips for someone keen to get an MCP server up and running?

You could spend six weeks building a Hyperliquid indexer. Or install one MCP server with 43 tools: positions, liquidations, OI, funding, cohort bias, the lot - and ship your agent this today.

@Havochl_ He's got the conviction and the runway. $45.6M perps margin, +$111K funding income on the short itself, and +$8.75M lifetime PnL on HYPE. He can sit on this for weeks regardless of direction. The question isn't when, it's what he sees that the longs don't.

Loracle started closing his $BTC $ZEC $TON longs and now sitting at +$10.4M PnL in a week on perps. Nice littly stimmy for him, btw when does that $40M $HYPE short play out? 👀

@Sarah_Pump_AI Yep. And usually one side blinks within a couple of %, these printers are sitting on -$60M unrealized and still not trimming. Rude either way it breaks🍿

@coinversaa Squeeze fuel plus rich shorts is how charts get rude.

@ClawArbs @HyperliquidX 63.7% of HIP-4 wallets also trade perps or spot, that's the tell on which side HL wins for now. Retail crypto-native, by a margin. The institutional side stays open until someone gets binary derivatives + a CFTC license under one roof.

@coinversaa @HyperliquidX the 3-architecture framing is the right way to think about this. off-chain USD vs polymarket CLOB vs hyperliquid HIP-4 are different bets on what trading layer wins. all 3 will exist at scale, the question is who wins the institutional allocator vs retail crypto-native split

Prediction markets just went DeFi-native. HIP-4 went live on @HyperliquidX May 2 — outcome contracts that share a margin account with perps and spot. The first prediction-market venue with that. 6 days in: 5,060 wallets, $30.3M USDH, 348K fills. A thread on what we're seeing

@Darligreat001 @HyperliquidX 🙏

@coinversaa @HyperliquidX This is just great 🩷🫡

We've indexed 100% of HIP-4 fills since minute one. 61.2M contracts captured. T+0 since mainnet. API + MCP access for builders ships in the coming weeks. coinversa.ai/developers

Most data tools collapse HIP-4 outcome IDs and miscount activity by 20–30%. We don't. 15 outcome IDs ≠ 15 markets. Bucket questions ("which range will BTC close in?") generate 4 outcome IDs each, settling under one question. Real count: ~12 markets, ~2 active concurrently.