product-feedback-roadmap-weapon

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

• Third-party content exposure detected (high risk: 0.78). Outsider-authored free text can enter the LLM context via the runtime workflow implied by “Read the user’s request” and by guides/templates that require reviewing “new submissions” / request descriptions from customer feedback platforms (e.g., user-submitted feature request/comment text in Canny/Featurebase/Userback/Jira items), which are not authored/chosen by the operating user.