ui-workflow

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

• Third-party content exposure detected (medium risk: 0.30). Step 2 “Mockup Generation” via the Stitch MCP workflow ingests outsider-provided PRD/free text only when the user supplies “the PRD” to generate_screen_from_text, after which those contents are used for mockup generation; this is a direct runtime path where attacker-controlled text can be fed without selecting a specific trusted item first.