signals-scout-inbox-validation

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

• Third-party content exposure detected (high risk: 0.75). The required workflow ingests outsider-authored free text from the inbox-report surfaces via inbox-reports-retrieve / inbox-reports-list (resolved/suppressed reports have titles/summaries and evidence fields that can contain arbitrary human-authored wording), which is then placed into the LLM’s context during validation/report-authoring and recurrence checks.