figma-to-compose

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

• Third-party content exposure detected (high risk: 0.75). The required workflow ingests Figma node tree content via the Figma Desktop MCP calls (get_metadata/get_design_context) to extract icon/vector SVG/path data and then converts it, and that Figma content is outsider-authored (public/third-party design files not authored by the operating user) into the agent’s LLM context.