seo

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly fetches and analyzes arbitrary public websites (e.g., SKILL.md commands like "$seo page " and "$seo audit ") and includes concrete runtime code that downloads and parses third‑party pages (scripts/fetch_page.py, scripts/parse_html.py) and renders them with Playwright (scripts/analyze_visual.py, capture_screenshot.py), so untrusted user-generated web content is read and used to drive analyses and recommendations that can change agent behavior.