iblai-api-management

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

• Third-party content exposure detected (medium risk: 0.65). The skill’s required runtime workflow reads OUTSIDER-authored free text by fetching organization data from external HTTP APIs (e.g., user/group/team/policy “list/detail” responses) and passing those response bodies into the agent’s LLM context, where fields like names/descriptions could contain attacker-controlled content (public internet / external service input category).