seo-analyzer

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

• Third-party content exposure detected (high risk: 1.00). The skill automatically loads and parses arbitrary public webpages: the entrypoints (scripts/seo_analyzer.py --auto ) call fetch_page_with_agent_browser which runs "agent-browser open " and "agent-browser get source", and a fallback_fetch_with_requests uses requests.get(url), causing the agent to ingest untrusted third-party HTML for analysis.