Skip to main content
Each agent draws from 20 specialized data layers served through background-polled caches. UI interactions read from cache rather than triggering fresh API calls — on-demand fetches happen only when agent reasoning or a trader query requires data beyond cache coverage.

Crypto Layers (L1-L13)

Powering every agent’s comprehensive analysis across 200+ perpetual and spot markets on Hyperliquid.

TradFi Layers (L14-L20)

Covering Hyperliquid’s expanding TradFi perpetuals — gold, silver, crude oil, wheat, natgas, S&P 500 (via the S&P Dow Jones Indices partnership), and more. L1-L13 are integrated in v1. L14-L20 are v2.

Caching Strategy

All data falls into one of three tiers: Immutable – Cached without expiration. Historical candles, closed trade records, completed funding rounds, token unlocks, etc. Volatile – Cached at polling intervals from 1 second to several hours depending on the data type. Live prices refresh in seconds. Cohort positioning and funding rates refresh in minutes. Macro context and sentiment refresh over hours. On-demand – Fetched fresh only when agent reasoning requires it. Per-asset deep dives, specific cohort breakdowns, detailed analysis for trade candidates.