Hyperliquid Whale Tracker — scored trades, funding extremes, wallet PnL

Most "whale alert" feeds tell you a big trade happened. This one tells you how much attention it deserves — every large Hyperliquid perp trade is scored 0–100 on the patterns that correlate with informed flow, and every record carries both counterparty addresses plus the tx hash (when Hyperliquid provides one — see FAQ), so you can verify records on-chain yourself. No API keys, no wallet connection, pay per result.

What it does

🐋 Whale feed (whale-feed)

Each run grabs the most recent trades per coin (instant bootstrap), then live-streams the actual trade tape over websocket for your listenSeconds window (default 90s, max 300s). Honest coverage model: you see everything that prints during your listen window — schedule runs back-to-back for near-continuous coverage. Every qualifying trade is scored:

Signal	Why it matters
Notional size ($10k → $1M+)	Base attention weight
Single print ≥1% of the coin's open interest	One trade moving a whole market
Contrarian vs crowded funding	Buying when shorts pay / selling when longs pay — fading the herd
Thin market (<$5M daily volume)	Big fish in a small pond

Every row: coin, side, price, size, USD, score, buyer address, seller address, tx hash, trade id, plus the coin's funding rate, open interest, and 24h volume for context. Cross-run dedup means a scheduled run never returns (or bills) the same trade twice.

📊 Market snapshot (market-snapshot)

Per-coin table: mark price, 24h change, 24h volume, open interest, hourly funding, annualized funding APR, and a crowdedFunding flag (≥25% APR — somebody's paying dearly to hold that side). The fastest way to find squeeze candidates and funding-farm setups.

👛 Wallet tracker (wallet-tracker)

For any address: account value, every open position (side, size, notional, entry, liquidation price, leverage, unrealized PnL) and recent fills with closed PnL. Vet a "top trader" before copying them, or watch a known whale's book in one call.

Quick start

{ "mode": "whale-feed", "minTradeUsd": 50000, "topCoins": 25 }

Schedule it back-to-back (e.g. every 5 min with listenSeconds: 290) + a webhook to Slack/Discord/Telegram for a near-continuous whale-alert service. Cross-run dedup (insertion-ordered, oldest-evicted) means overlapping windows never bill the same trade twice.

{ "mode": "wallet-tracker", "wallets": ["0x..."] }

Honest data, honestly billed

Built by the maker of Polymarket Whale Tracker Pro, Polymarket Top Traders and the Polymarket MCP Server, with the same rules:

• A rate-limited upstream call raises — it is never passed off as "no data"
• Error records are pushed for transparency but never billed
• Cross-run dedup happens before billing — the same trade id is never charged twice
• Data comes from Hyperliquid's public info API in real time; nothing is cached or stale

FAQ

Which trades appear in the feed? Everything at or above your minTradeUsd that prints during the run: an instant bootstrap of each coin's most recent trades, plus the full live tape for your listenSeconds window. Coverage equals your listen window — back-to-back scheduled runs approximate a continuous feed, and dedup guarantees overlapping runs only ever bill a trade once.

Are the addresses real? Yes — Hyperliquid's API returns both sides of every fill. Note: Hyperliquid reports some internalized fills with a zero hash; we report those as hash: null rather than pretending — the trade id, timestamp and both addresses remain the verification keys.

Does it work for spot? v0.2 covers perps (where the whales are). Spot support is on the roadmap — open an issue if you need it.

More tools by this maker

Tool	What it does
Polymarket Whale Tracker Pro	Insider-scored large trades, top holders, wallet P&L
Polymarket Top Traders	Leaderboards with cross-window consistency + verified P&L
Polymarket MCP Server	The same live data as native AI-agent tools (MCP)
Hyperliquid Whale Tracker	Scored perp trades w/ counterparty addresses, funding/OI, wallet PnL
Web Search MCP Server	Multi-engine web/news search + instant answers for AI agents

Same engineering rules everywhere: fail-loud data integrity, dedup-before-billing, errors never billed.