serenity-research-model

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

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow ingests outsider-authored free text from user-provided public X exports / saved post URLs (e.g., extract --posts <file> reads text and quoted_text fields from CSV/JSON/TXT/MD), which are then embedded into the agent’s LLM context for semantic review/labeling (via auto-review/report generation).