meta-ads-creative

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

• Third-party content exposure detected (high risk: 1.00). The skill's Phase 1 research workflow (SKILL.md "Phase 1: Research") and the referenced creative-research-methods.md explicitly require mining public/user-generated sources—e.g., Reddit/Facebook groups, Meta Ad Library, TikTok Creative Center, and other organic social posts—which the agent is expected to read and use to inform ad copy and format decisions, thereby exposing it to untrusted third-party content that could carry indirect prompt injections.