ai-synthetic-personas

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

• Third-party content exposure detected (high risk: 0.70). The SKILL.md's required workflow and Validation Checklist explicitly instruct the operator to ingest third‑party data — e.g., "Available secondary data" (customer feedback/Meta Audience Insights) in Required Input and the Validation Checklist item "If the client has a Facebook Page, run Meta Audience Insights" and "run at least one real interview or WhatsApp conversation" — which requires reading user-generated/public social data that can materially alter persona outputs.