marketing-podcast-strategist

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

• Third-party content exposure detected (medium risk: 0.65). The skill’s runtime workflow ingests outsider-authored free text via “Social media distribution / community engagement” and “collect feedback, run topic polls,” where listener comments and poll responses (authored by others) are likely fed into the agent’s LLM context for analysis or response generation.