design-md-query

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

• Third-party content exposure detected (high risk: 0.95). At runtime the skill reads docs/<slug>/DESIGN.md from the repo and injects its plain-text contents (e.g., the “Agent Prompt Guide” and other sections) into the agent’s LLM context; these DESIGN.md files are outsider-authored extracted design-system text from third-party companies’ public sites, not authored by the operating user.