guard

You can prevent accidental data loss by enforcing guard rails on dangerous commands. This mechanism warns before executing risky operations like rm -rf or DROP TABLE, restricting edits to a user-selected directory to block unintended changes in both development and production environments.

You enforce an edit boundary by selecting a specific directory to protect, then enabling guard mode. This policy restricts edits to your chosen boundary, automatically blocking any unauthorized file modifications outside that directory.

A destructive-command warning system alerts you before executing risky operations like rm -rf or DROP TABLE. You need this mechanism when protecting critical systems from accidental data loss and preventing unauthorized edits in development or production environments.

Yes, enabling runtime checks for edit boundaries requires integration with the careful and freeze hooks. These hooks perform the pre-use safety checks and boundary enforcement necessary to verify your guard rails are active before executing commands.

Yes, you can apply directory-bound edit policies in development contexts to prevent unintended changes. The guard mechanism enforces these boundaries, restricting edits to your selected directory and warning on destructive commands to protect critical systems.

The primary limitation is that guard rails require integration with the careful and freeze hooks to perform runtime checks. Without these hooks active, the boundary enforcement and destructive-command warnings cannot verify or block risky operations effectively.