canvas-design
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill functions as a framework for generating artistic design philosophies and translating them into visual assets. It contains no executable code, network exfiltration logic, or obfuscated components.
- [EXTERNAL_DOWNLOADS]: The skill instructions provide a general directive to download fonts for use in design tasks. The accompanying license files point to official, well-known, and trusted repositories on platforms such as GitHub and Google Fonts for various open-source font projects.
- [INDIRECT_PROMPT_INJECTION]: The skill includes a surface for indirect prompt injection because it ingests user-provided creative foundations to generate artwork. Ingestion points: User-supplied creative foundations and specific visual requests outlined in SKILL.md. Boundary markers: Absent; there are no explicit delimiters or instructions to ignore embedded commands within user data. Capability inventory: The skill enables the agent to create and modify .md, .pdf, and .png files via agent tools. Sanitization: The instructions emphasize that text should be minimal and secondary to visual expression, which naturally mitigates the risk of textual instruction exploitation.
Audit Metadata