better
Pass
Audited by Gen Agent Trust Hub on Nov 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is a comprehensive collection of interface engineering principles and documentation. It contains no external downloads or remote dependencies.
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a local Node.js script (
scripts/contrast.mjs) for WCAG contrast verification. Technical analysis of this script confirms it is a purely mathematical utility with no network access, disk write operations, or sensitive data retrieval. - [INDIRECT_PROMPT_INJECTION]: The skill processes project-specific data like
CONTRIBUTING.mdand component code. While these files are external to the skill and could contain malicious content, the skill provides specific and bounded design instructions that minimize the risk of the agent following unexpected directions from processed data. - [PROMPT_INJECTION]: The skill uses directive language to ensure UI improvements are performed automatically. These instructions are aligned with the skill's primary purpose and do not attempt to override the underlying model's safety constraints or system prompts.
Audit Metadata