guard

Destructive command warnings alert you when commands like rm -rf, DROP TABLE, or force-push are attempted in your environment. This prevents unsafe changes by integrating pre-tool hooks that execute safety checks before any tool usage.

Yes, you can enforce an edit boundary that restricts file modifications to a user-selected directory. This prevents accidental or malicious changes to files outside that defined scope, which is essential for critical environment safety.

Pre-tool hooks work by integrating with sibling safety skills to ensure safety checks are performed before any tool usage. They enforce destructive-command warnings and directory-scoped edit restrictions to prevent unsafe changes in critical environments.

The best way to lock down edits in live systems is by combining destructive-command warnings with directory-scoped edit restrictions. This dual approach prevents both unsafe command execution and out-of-scope file modifications.

Yes, enforcing edit boundaries and destructive warnings requires an explicit dependency on sibling careful and freeze skills. These sibling skills provide the foundational hooks and boundary mechanisms to enforce warnings and restrictions.

Directory-scoped edit restrictions are not suitable for workflows requiring modifications across multiple directories outside a single sanctioned scope. They are designed specifically for workflows where safety around production or live systems is essential.