unified-research-synthesizer

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

• Third-party content exposure detected (medium risk: 0.65). The required workflow ingests user-provided free-form text inputs like customer_feedback (reviews/testimonials/support tickets/transcripts) and potentially competitor_data (landing-page/ad text) into the agent’s LLM context, which can include outsider-authored content.