canvas-design

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The instructions utilize behavioral reinforcement and simulated context (e.g., "The user ALREADY said 'It isn't perfect enough'") to force the agent into a specific creative mindset. While these represent context manipulation techniques, they are used here for quality control rather than bypassing safety protocols.
  • [EXTERNAL_DOWNLOADS]: The skill references a directory of font license metadata and instructs the agent to use external fonts. These references target well-known and reputable open-source font projects and organizations.
  • [DATA_EXFILTRATION]: The skill contains an indirect prompt injection surface because it processes user-provided topics to generate "design philosophies" without explicit boundary markers. However, the impact is limited as the skill's capabilities are restricted to visual document generation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 05:19 PM