scanning-experiments-with-replay-vision

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

• Third-party content exposure detected (medium risk: 0.30). The skill workflow involves reading session recordings and observations using tools like vision-scanners-observations-list and vision-scanners-scan-session, where session content (page content and user interactions) is treated as untrusted data that may be influenced by third-party users.