Overview

Panther operates as a multi-component system designed to ingest, process, analyze, and deliver insights from the Hyperliquid ecosystem.

The basic flow involves:

  1. Data Ingestion: Collecting real-time data from Hyperliquid CORE (WebSockets) and EVM (RPC).

  2. Data Queuing: Publishing structured data onto a message queue (Redis) for decoupling and scalability.

  3. Agent Processing: AI agents (elizaOS) subscribe to relevant data streams from the queue.

  4. Insight Generation: Agents analyze the data based on predefined logic and models.

  5. Delivery/Action: Insights are formatted and sent to users (Discord) or used to inform potential future on-chain actions (Phase 3).