@1247/vampire-attack-hl

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

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow takes outsider-supplied --wallet input and then reads Hyperliquid fill-level data plus candle snapshots via skills/vampire-attack-hl/scripts/analyze_hl_wallet.py calling https://api.hyperliquid.xyz/info endpoints (userFillsByTime and candleSnapshot), where the returned fields (e.g., fill coin, side, px, sz, fee, builderFee, time) are free-text/JSON content originating from an external party and then used to generate the client-facing markdown.