best-practices
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to audit user-provided code, which acts as an ingestion point for untrusted data. Evidence: 1. Ingestion points: User-supplied HTML, JavaScript, and CSS snippets in SKILL.md. 2. Boundary markers: Absent in the prompt templates. 3. Capability inventory: The skill provides code recommendations and snippets but does not autonomously execute code or perform network operations. 4. Sanitization: Explicitly recommends the use of DOMPurify and textContent to prevent XSS.
- [EXTERNAL_DOWNLOADS]: The skill provides examples that reference the polyfill.io script service and links to several well-known security auditing tools including Mozilla Observatory, SecurityHeaders.com, and the W3C Validator.
- [DYNAMIC_EXECUTION]: Contains documentation on conditional script loading using document.write, which is properly categorized in the text as a deprecated API to be avoided in favor of modern dynamic loading techniques.
Audit Metadata