marketing-pricing

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 § "Research Mode: Live Competitive Pricing Intelligence") explicitly instructs the agent to use agent-browser to open and scrape live public URLs (e.g., competitor pricing pages, G2 reviews, Reddit search, Wayback Machine) and to "get text body" and extract signals that will drive pricing recommendations, thereby exposing the agent to untrusted, user-generated third‑party content that can materially influence actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's Research Mode requires installing/using agent-browser at runtime via commands that fetch and run remote code from https://github.com/vercel-labs/agent-browser (npx skills add https://github.com/vercel-labs/agent-browser and npm install -g agent-browser), making that external repository a runtime dependency that can execute code and affect agent behavior.