safety-guard

Preventing destructive operations on production systems requires command interception and directory locks to monitor autonomous agents. This Skill restricts unintended system modifications and data loss by requiring PreToolUse hooks to intercept destructive commands before execution.

Locking file edits to a specific directory tree during a migration requires a freeze mode that restricts modifications to a defined path. This Skill provides directory locking mechanisms to ensure critical files outside the target tree remain untouched during autonomous operations.

Combining cautionary warnings with directory restrictions is supported through a dedicated guard mode. This mode merges careful warnings and freeze locks to provide maximum production protection by intercepting destructive commands while simultaneously locking the directory tree.

Preventing unintended system modifications does require PreToolUse hooks for command interception. These hooks monitor and restrict destructive operations on production systems, enabling cautionary warnings that halt unintended system modifications before they execute.

Directory locking mechanisms ensure autonomous agent safety by restricting file edits to a specific directory tree. This prevents unintended system modifications and data loss by blocking agents from altering critical files outside the locked directory scope.