marketing-analytics

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

• Third-party content exposure detected (high risk: 0.90). The skill's Research Mode (SKILL.md) explicitly instructs the agent-browser to open and scrape live, arbitrary public URLs — e.g., pagespeed.web.dev, search.google.com/test/rich-results, validator.schema.org, and "https://{domain}" — and to extract/interpret their contents to drive audits and recommendations, so untrusted third-party page content can materially influence the agent's decisions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill instructs installing and using the external agent-browser skill via "npx skills add https://github.com/vercel-labs/agent-browser --skill agent-browser", which fetches and installs remote code used at runtime to execute browser commands (e.g., agent-browser eval), so this URL is a runtime dependency that can execute code.