ai-short-drama-production

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

• Third-party content exposure detected (medium risk: 0.30). Skill runtime ingests user free text via “从用户输入提取” and then builds controllability contracts before passing to downstream skills (ai-storyboard-director / prompt-engineer / qc-supervisor), so outsider-authored text can enter the workflow without first selecting a specific item.