canvas-design
Pass
Audited by Gen Agent Trust Hub on Apr 2, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill employs a simulated context technique by claiming 'The user ALREADY said "It isn't perfect enough..."'. This attempts to override the model's standard interaction flow by injecting a fake conversation history to force a specific quality level.- [EXTERNAL_DOWNLOADS]: Instructions explicitly tell the agent to 'Download and use whatever fonts are needed,' which introduces a dependency on external, unverified resources from the internet.- [COMMAND_EXECUTION]: The skill requires generating binary files such as .pdf and .png, which involves executing system-level rendering commands or software libraries.- [PROMPT_INJECTION]: Directives like 'Embrace ultimate design freedom' and 'Push aesthetics and design to the frontier' are intended to steer the agent away from its default operational constraints and safety-oriented formatting.
Audit Metadata