build-awwwards-quality-sites
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided reference evidence to inform art direction, which creates a potential vector for indirect prompt injection.
- Ingestion points: The agent is instructed to 'Inspect the user's reference evidence completely before implementation' in SKILL.md.
- Boundary markers: The instructions explicitly state to 'Extract only high-level traits' and 'Never reuse, trace, or closely reproduce reference assets, screenshots, source code, identity, or copy,' which provides a conceptual boundary.
- Capability inventory: The skill facilitates complex code generation including GSAP animations, Three.js shaders, and smooth-scroll engine initialization, along with browser-based validation tools.
- Sanitization: No specific technical sanitization or input filtering is defined for the reference evidence beyond the instruction to only extract high-level traits.
Audit Metadata