user-feedback-aggregation

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime workflow is described as aggregating “continuous user-feedback streams” from channels like support tickets, NPS comments, in-app feedback, sales calls, social mentions, and customer councils; those channels contain outsider-authored free text (e.g., customer comments/messages) that would be ingested into the LLM for triage/synthesis, creating an indirect prompt-injection risk.