reviewing-security

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md describes a “Queue or event consumer” and “LLM / tool output” as an entry point where “anyone who can influence the model’s context” can steer tool calls, implying runtime ingestion of outsider-authored free text if the workflow monitors/consumes such sources without selecting a specific item first.