design-critique
Pass
Audited by Gen Agent Trust Hub on Aug 3, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a structured process for reviewing designs based on brand and structural rubrics. All instructions are focused on design analysis and implementation workflows.- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection because it ingests data from untrusted external sources, such as live web pages and local documentation files. This is inherent to its primary purpose of design review.
- Ingestion points: The skill gathers subjects from live pages, URLs, and documentation files including
DIRECTION.mdandBRAND.md(SKILL.md). - Boundary markers: There are no explicit delimiters or instructions to ignore embedded prompts within the ingested content.
- Capability inventory: The skill triggers other automated tools for UI prototyping and implementation based on its analysis (SKILL.md).
- Sanitization: No content filtering or validation is described for the subject material.
Audit Metadata