The Agent Skills Directory

[SAFE]: No malicious instructions or prompt injection patterns were detected. The skill workflow is transparent and follows its stated purpose.\n- [SAFE]: Dependency analysis shows use of standard, reputable libraries including google-genai, cairosvg, and Pillow. External assets like Tailwind CSS and React are loaded from well-known CDNs.\n- [SAFE]: Credential management follows best practices, requiring users to provide their own API keys via a local environment file (.env) rather than hardcoding or attempting to exfiltrate secrets.\n- [SAFE]: Remote operations are restricted to official API calls and standard image processing. The option for a custom API base URL is documented as a user-configurable feature for proxy support.\n- [SAFE]: While the skill interpolates user-provided text into a prompt for image generation, this is an intended primary function and the risk surface is consistent with the skill's legitimate purpose.

logo-generator