okx-ai

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

• Third-party content exposure detected (medium risk: 0.30). The skill ingests peer chat messages (a2a-agent-chat), outsider-authored task/subscription deliverables, and evaluator evidence feeds via its A2A messaging and evaluation workflows.

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

• Direct money access detected (high risk: 0.70). The skill operates OKX AI agents and marketplace workflows, including buying and managing agent services and subscriptions (SKILL.md:3), but lacks documented independent authentication or authorization controls enforced by the service or execution layer for these direct purchase operations.