The Agent Skills Directory

[SAFE]: The skill uses official SDKs for OpenAI, Google Gemini, and Replicate. All external network communications are directed to these well-known and reputable services.
[SAFE]: Command execution for image processing (via ImageMagick and rembg) is implemented with security in mind. Scripts such as ThumbnailText.ts include specific character escaping to prevent potential command injection vulnerabilities.
[SAFE]: Secret management follows best practices by using environment variables and .env files. There are no hardcoded credentials or unauthorized data exfiltration patterns.
[SAFE]: File system interactions are localized to expected areas such as the user's Downloads folder and specific asset directories.

do-art