See the whales and find the edge of liquidation — as an API.

Real-time Hyperliquid liquidation & smart-money signals your agents can query directly. The biggest leveraged positions about to blow up, net smart-money bias, cascade risk — one call, no dashboard.

Read the docs → Start in 2 minutes llms.txt

Built for AI agents·developers·automation teams Read-only

Your first call — biggest positions nearest liquidation

# the powder kegs, largest first
curl "https://goose-liq.duckdns.org/v1/positions/at-risk?sort=groesse&limit=3" \
  -H "X-API-Key: YOUR_KEY"

Response (trimmed)

{
  "data": [
    { "coin":"HYPE", "side":"long", "notional":82772198,
      "leverage":5, "dist_to_liq_pct":9.8, "upnl":29344446,
      "tags":["smart","whale"] }
  ],
  "count":3, "as_of":1784325592778
}

What your agent gets

🎯 Positions at risk

The largest leveraged positions closest to forced liquidation — spot the powder kegs before they go off.

🧠 Smart-money bias

Net long/short per coin among proven smart-money wallets. Where the profitable money leans.

💥 Cascade risk

Clusters of positions bunched near liquidation — anticipate the chain reaction.

🐋 Whale entries & exits

Fresh smart-money entries (with price) and voluntary exits (with realized PnL). The full cycle.

Start in under 2 minutes

Grab your API key and base URL (or run it locally on :8000).

Make the call above. Every human action in the dashboard is one endpoint — no menus, no clicking.

Wire the MCP server into Claude so the assistant calls the tools itself. Full setup in the docs.

MCP — Claude Desktop config

{
  "mcpServers": {
    "liq-whale-signals": {
      "command": "python",
      "args": ["…/on_demand/mcp_server/server.py"],
      "env": { "ODMD_API_URL":"https://goose-liq.duckdns.org", "ODMD_API_KEY":"YOUR_KEY" }
    }
  }
}