The Agent Skills Directory

[Data Exposure & Exfiltration] (SAFE): The scripts operate exclusively on the local filesystem. No network operations, API calls, or hardcoded credentials were detected. Path handling is performed using the robust pathlib module.
[Unverifiable Dependencies & Remote Code Execution] (SAFE): All imports (sys, zipfile, pathlib, os, re) are part of the Python standard library. No external or untrusted packages are required.
[Indirect Prompt Injection] (SAFE): The quick_validate.py script includes a security-positive check that prevents the use of angle brackets (<, >) in descriptions, which mitigates potential injection or rendering issues in downstream UIs.
[Command Execution] (SAFE): No instances of os.system, subprocess, eval, or exec were found. The code is limited to file I/O, path manipulation, and regex validation.
[Privilege Escalation] (SAFE): The scripts do not attempt to gain administrative privileges or modify system-level configurations.

skill-creator