compatibility-testing

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

• Third-party content exposure detected (high risk: 0.80). The skill instructs agents to navigate to and test arbitrary websites (e.g., the Playwright example uses page.goto('https://example.com') and the "Agent-Driven Compatibility Testing" Task accepts a url parameter), so the agent will fetch and interpret content from open/public third-party sites as part of its workflow.