kai-repurpose

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md indicates the agent reads pillar input via user-provided content (from the user’s MARKETING.md, which it may create from their project files, and then extracts from that pillar content) and then generates derivatives, meaning outsider-authored free text can be ingested through user-supplied source content/MARKETING.md.