signals-scout-general

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

• Third-party content exposure detected (medium risk: 0.65). The runtime workflow uses PostHog project data plus stored scratchpad/inbox contents (outsider-authored free-form messages/issues/comments), which are then incorporated into the agent’s LLM context via steps that list/retrieve reports and explore/corroborate findings; SKILL.md’s orient/decide steps imply reading those retrieved texts before authoring/editing reports.