canvas-design

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: No malicious patterns or security vulnerabilities were detected. The skill instructions are focused on artistic generation and use standard agentic prompting techniques for quality control.
  • [NO_CODE]: The skill does not contain any executable scripts or source code. It relies entirely on natural language instructions for the agent's behavior, which significantly reduces the attack surface for code-based vulnerabilities.
  • [PROMPT_INJECTION]: The skill incorporates a refinement step that uses a simulated user feedback loop ('The user ALREADY said...') to drive the model toward higher-quality output. This is a functional use of pre-filled context for quality control rather than a malicious attempt to override safety protocols.
  • [EXTERNAL_DOWNLOADS]: While the skill mentions the possibility of downloading fonts, it provides a local directory for assets and does not point to untrusted remote sources. This is a standard creative instruction within the intended use case of the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 07:43 PM