prd

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

• Third-party content exposure detected (high risk: 0.70). The skill explicitly loads external Figma designs as part of its required workflow (Step 3: "Load Figma designs" calling mcp__figma__get_design_context / mcp__figma__get_screenshot for fileKey/nodeId), which ingests third-party/user-generated content that the agent reads and uses to drive UI reviews, findings, and next actions.