NPS Action Planner

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

• Third-party content exposure detected (high risk: 0.70). The skill explicitly ingests and analyzes open-ended, user-generated "verbatim_feedback" (see Required Inputs) and Step 2 ("Apply topic modeling to detractor verbatims") as part of its workflow, meaning untrusted respondent text could influence driver analysis and downstream action decisions.