The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/publish.py script uses subprocess.run to invoke external document processing tools like pandoc, typst, and LaTeX engines. The implementation uses argument lists rather than shell strings, which is a secure practice that prevents shell injection vulnerabilities.
[EXTERNAL_DOWNLOADS]: The skill's documentation provides clear and legitimate instructions for installing the necessary build tools (pandoc and typst) using official package managers such as winget and brew. These are well-known, trusted software sources.
[DATA_EXFILTRATION]: No network operations were detected in the skill's scripts or instructions. The skill performs all operations locally on the user's files without sending data externally.
[PROMPT_INJECTION]: The instructions in SKILL.md are focused on document formatting and hand-in guidance. There are no attempts to override agent safety protocols or hijack behavior.
[SAFE]: The script includes utility features like managing a .gitignore file to track build artifacts, which is documented and performed using standard library functions.

proposal-publish