marketing-email

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

• Third-party content exposure detected (high risk: 1.00). The skill's Shared Patterns (references/shared-patterns.md — "Context C — Live Website URL Audit") and the agent-browser setup in SKILL.md explicitly direct the agent to fetch and audit user-provided public URLs (live websites) and treat those pages as the working source of truth, meaning untrusted third‑party content will be read and used to drive recommendations and next actions.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill's agent-browser setup explicitly instructs installing remote code at runtime using "npx skills add https://github.com/vercel-labs/agent-browser", which fetches and executes external repository code used by the skill for browser-based research (runtime dependency) and therefore represents a remote code execution dependency.