landing-page-design
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by instructing the agent to incorporate external data—such as testimonials, case studies, and brand copy—into the generated landing page code.
- Ingestion points: External content is ingested via the 'Intake' (A1) and 'Proof and assets' (A2) guidelines in
SKILL.md. - Boundary markers: The instructions do not specify the use of delimiters or warnings to prevent the agent from following instructions that might be embedded in the provided marketing copy or testimonials.
- Capability inventory: The skill directs the agent to generate functional web code, including Tailwind CSS styles and JavaScript logic (e.g.,
IntersectionObserverfor animations in sections B7 and B11). - Sanitization: There are no requirements or mentions of sanitizing or escaping untrusted external strings before they are rendered into the final output.
Audit Metadata