google-ads-landing-review

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

• Third-party content exposure detected (high risk: 0.90). The skill's required workflow (Data Acquisition — Landing Page Review) explicitly instructs the agent to fetch and browse each landing URL using "web_fetch" and "browser" for content analysis, meaning it ingests arbitrary public landing pages (third‑party URLs) and uses their content to drive diagnostic decisions and recommendations.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly fetches user-provided landing page URLs at runtime via web_fetch/browser (e.g., https://example.com/landing), and uses the fetched page content to drive the agent's analysis and draft generation, so external pages can directly control agent prompts/outputs.