ad-to-landing-page-auditor

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

• Third-party content exposure detected (high risk: 1.00). The skill's Phase 2 "Landing Page Audit" explicitly instructs the agent to fetch landing page content from provided URLs (via fetch_webpage: [landing_page_url] or curl) and to read/score that HTML to drive recommendations, which exposes the agent to untrusted third-party webpages.

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

• Potentially malicious external URL detected (high risk: 0.90). This skill fetches user-provided landing page URLs at runtime (via fetch_webpage or curl, e.g. https://<user‑landing‑page>) and injects the retrieved HTML into the agent's analysis prompt, so external page content directly controls model behavior.