The Agent Skills Directory

[SAFE]: The skill performs an audit of other skills' configurations and state metadata to detect potential cost inefficiencies. It only accesses file system metadata (modification times) and its own state files.
[SAFE]: No network operations, data exfiltration, or access to sensitive credentials were found in the provided files.
[SAFE]: The remediation utility in audit.py follows a safe pattern by providing instructions to the user rather than performing automated modifications to system or application configuration files.
[SAFE]: The script uses yaml.safe_load() for parsing its own state files, following security best practices for deserialization.

cron-hygiene