frontend-coding-agent

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md's mandatory Visual Verification and Playwright Verification Protocol explicitly instruct the agent to navigate to and inspect live URLs (e.g., mcporter call playwright.navigate '{"url":"$NEXT_PUBLIC_APP_URL"}'), meaning it will fetch and interpret external web pages which can influence follow-up build/fix actions.