The Agent Skills Directory

[SAFE]: The skill facilitates virtual try-on tasks by securely communicating with reputable AI providers (fal.ai and OpenAI) through a platform-specific proxy.\n- [SAFE]: Image data is handled locally using base64 encoding or via public URLs, with appropriate validation for file existence, type, and size (10MB limit).\n- [SAFE]: Cost tracking is integrated via a dedicated module that logs usage to a standard internal directory, ensuring transparent resource monitoring.\n- [SAFE]: Security best practices are followed for credential management, utilizing environment variables for API keys instead of hardcoding sensitive information.\n- [SAFE]: Network traffic is restricted to the internal proxy or verified AI service domains, and file operations are limited to the workspace and specific output directories.

image-tryon