user-feedback-interpreter

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly ingests and analyzes untrusted, user-generated sources — e.g., "social media mentions", "app store reviews", uploaded survey/export files, and integrations with Google Drive/Slack for pulling feedback — which the agent is expected to read and interpret as part of its workflow, exposing it to potential indirect prompt injection.