The Agent Skills Directory

[SAFE]: The skill is designed for document processing and follows security best practices throughout its implementation.
[COMMAND_EXECUTION]: External tools such as soffice (LibreOffice) and git are called via subprocess.run with list-based arguments to perform document conversion and text comparison. These calls are well-restricted to specific tasks related to the skill's primary purpose.
[SAFE]: XML parsing is handled by the defusedxml library, providing protection against XML External Entity (XXE) and expansion attacks.
[SAFE]: The skill operates entirely within the agent's local workspace or temporary directories, with no evidence of sensitive file access or network-based data exfiltration.
[SAFE]: Proper XML escaping is implemented when handling user-provided metadata, such as author names, to prevent injection attacks.

docx