The Agent Skills Directory

[SAFE]: The skill instructions are centered on creative design tasks and the generation of visual documents (PDF/PNG). It does not request access to sensitive file paths, environment variables, or credentials.\n- [SAFE]: The skill manages its resources locally, directing the agent to use fonts located within the skill's own directory (./canvas-fonts). There are no hardcoded external URLs for script or executable downloads.\n- [PROMPT_INJECTION]: The skill uses strong instructional language and formatting (e.g., 'CRITICAL', 'IMPORTANT') to guide the agent toward high-quality artistic output. These instructions do not attempt to bypass safety filters or override core system behavior in a malicious way. The use of 'The user ALREADY said' is a framing technique for aesthetic refinement rather than a security injection.\n- [SAFE]: No obfuscation, Base64 encoding, or hidden control characters were detected in any of the analyzed files. All license files for the included fonts are standard and point to legitimate, well-known open-source repositories.

canvas-design