technical-seo

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

• Third-party content exposure detected (high risk: 0.85). The skill’s runtime audit path fetches arbitrary target URLs (including robots.txt and sitemap.xml) via scripts/seo_audit.py and then injects the fetched page text/HTML-derived strings (e.g., meta tags, H1, robots contents) into the agent’s LLM context, which is outsider-authored free text from the target site.