aeon-monitor-polymarket

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

• Third-party content exposure detected (high risk: 0.85). Outsider free text is ingested from Polymarket comment threads (e.g., “fresh comments from watched commenters” / “Most-upvoted comments since last scan”), which the skill reads at runtime via Polymarket API calls and then includes/quotes in the agent’s LLM context as untrusted comment text.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 1.00). This skill is specifically built around Polymarket trading: it calls Polymarket market and orderbook APIs, references Bankr-compatible RPC for on-chain verification, and — critically — produces "copy-paste Submit payloads for AgenticBets or direct Polymarket interaction." Those submit payloads are explicit artifacts for placing bets/orders on a markets platform (crypto prediction market). Because it explicitly prepares actionable transaction/order payloads for market execution, it grants direct financial execution capability.