color-system
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided design briefs and brand evidence to generate color systems, which serves as an ingestion point for untrusted data.\n
- Ingestion points: Brand evidence, audience descriptions, and product requirements gathered from user input (SKILL.md, contract.json).\n
- Boundary markers: The skill does not employ specific delimiters or warnings to the agent to ignore instructions embedded within the user data.\n
- Capability inventory: The skill's functionality is limited to generating text and markdown documentation. It does not perform subprocess execution, file system modifications, or network operations.\n
- Sanitization: No input validation or sanitization of user-supplied briefs is documented.\n- [SAFE]: The skill provides instructional JavaScript and Python code snippets in 'assets/oklch-ramp.md' to demonstrate color ramp math. These are static reference templates intended for design workflows and do not execute untrusted logic or interact with the operating system.
Audit Metadata