The Agent Skills Directory

[SAFE]: The skill is a defensive tool that enhances the security posture of the agent. It implements hooks to detect and warn against destructive command patterns such as recursive deletions (rm -rf), forced git pushes, and SQL data destruction commands.
[SAFE]: The 'Lockdown Mode' provides a robust security boundary by enforcing directory-level restrictions on file editing and writing tools. It uses standard path resolution and validation techniques to prevent accidental or malicious modifications outside of a defined workspace.
[SAFE]: No evidence of malicious intent, data exfiltration, obfuscation, or unauthorized privilege escalation was found. The skill's behavior is transparently documented and aligned with its stated purpose of protecting the user's environment.

safe-mode