frontend-design
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [Prompt Injection] (SAFE): The use of 'CRITICAL' and 'IMPORTANT' markers in SKILL.md is strictly for stylistic and design emphasis. There are no attempts to override system safety protocols or extract system prompts.
- [External Downloads] (SAFE): The README.md links to a GitHub repository under the 'anthropics' organization. This is a trusted source according to the security policy and is used for documentation purposes.
- [No Code] (SAFE): This skill contains no executable code, scripts, or package manifests. It is entirely composed of markdown instructions for the AI model.
- [Indirect Prompt Injection] (LOW): While the skill ingests user requirements to generate code, it does not possess capabilities (like file writing or network access) that could be exploited via malicious user input. Boundary markers are inherently managed by the AI's standard processing of user prompts.
Audit Metadata