landing-page
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill fetches and processes content from external websites and public sources, which introduces a potential vector for indirect prompt injection. A malicious actor could embed instructions within a website that the agent is instructed to research.
- Ingestion points: The skill fetches company websites and public sources (e.g., G2, Reddit, CRO blogs) to gather positioning and competitor data (SKILL.md).
- Boundary markers: The skill includes a dedicated security reference (references/agent-security.md) that explicitly instructs the agent to treat fetched content as data rather than instructions and to report injection attempts.
- Capability inventory: The skill generates prototype HTML (SKILL.md, Step 9).
- Sanitization: The security reference requires HTML-escaping all interpolated values and validating URLs to prevent XSS (references/agent-security.md, Rule 3).
- [EXTERNAL_DOWNLOADS]: The skill uses the Tailwind CSS CDN for generating prototype previews.
- Evidence: Step 9 in SKILL.md specifies the use of
<script src="https://cdn.tailwindcss.com"></script>. - Context: This is a well-known service used for UI styling in prototypes, and the skill provides a clear warning that it is not production-ready and requires a build step for deployment.
Audit Metadata