khe-color-and-surface
Pass
Audited by Gen Agent Trust Hub on Nov 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [SAFE]: The skill is primarily instructional and provides design guidelines for colors and surfaces. No malicious patterns, hardcoded credentials, or unauthorized command execution sequences were detected. The skill's behavior matches its stated purpose.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies vulnerability surfaces where it ingests data from external project files.
- Ingestion points: Reads project-specific files such as
.design,DESIGN.md, and existing token files as defined in Step 1 ofSKILL.md. - Boundary markers: None explicitly defined in the instructions to separate untrusted data from instructions.
- Capability inventory: The skill is designed to write design tokens and CSS code to the project filesystem based on its analysis.
- Sanitization: No specific sanitization or filtering of external content is mandated before interpolation into generated code.
- [EXTERNAL_DOWNLOADS]: The skill references documentation and source code from several established organizations and developers for guidance, including Meta's Astryx library (
github.com/meta/astryx) and design resources from Benji Taylor, Jakub Krehel, and Emil Kowalski. These are documented as educational sources for design engineering principles and do not involve automatic execution of remote code.
Audit Metadata