mono-color
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The mono-color skill is a legitimate creative tool designed for editorial poster generation. A comprehensive review of the SKILL.md instructions, the JSON catalogs in the design-system/ directory (covering colors, compositions, typography, etc.), and the markdown guidelines in the references/ folder shows that all components are aligned with the skill's stated purpose and do not contain malicious patterns or suspicious logic. The skill is authored by Nous Research, which is a known vendor entity.\n- [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface for user-controlled content.\n
- Ingestion points: The agent extracts the subject, intent, and verbatim words from user input in the SKILL.md procedure to construct an image generation prompt.\n
- Boundary markers: The instructions specify to preserve exact supplied text verbatim and never translate or rewrite it, which acts as a logical delimiter for the primary display text.\n
- Capability inventory: The skill utilizes the image_generate tool to produce visual artifacts and writes the final recipe to the local filesystem under the mono-color-output/ directory.\n
- Sanitization: No explicit content sanitization or escaping is mentioned, as the data is intended for image generation rather than shell execution or code interpretation. The risk is considered low and managed through structured manifests.
Audit Metadata