lp-hedge

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). This skill explicitly ingests public third‑party data (DeFiLlama MCP for pool and incentive metadata, OKX REST for perp/funding/orderbook data, and GoPlus MCP for token security) as mandatory parts of its pre‑flight and data‑collection workflow, and those external fields are parsed and used to drive hedge, rebalance, and recommendation decisions—creating a clear channel where attacker‑controlled or untrusted content could influence agent behavior.