guard

To prevent accidental destructive commands during production debugging, you need a safety mode that warns before executing risky operations like rm -rf or DROP TABLE. This Skill enforces those warnings and restricts edits to a chosen directory to block unintended damage.

You can restrict file edits to a specific directory boundary during risky deployments by configuring an edit boundary. This Skill enforces directory-bound edits, ensuring changes stay within a specified path and blocking any modifications outside that defined directory.

Full safety mode for live system maintenance is a protective state that warns on destructive commands and restricts edits to a chosen directory. It works by using hooks to verify safety boundaries before tasks run, preventing accidental damage in production environments.

Yes, directory-bound edits can integrate with sibling risk-mitigation tools like careful and freeze behaviors. This Skill combines edit boundaries with destructive command warnings for end-to-end safety during risky DevOps operations, debugging, and deployments.

To set up edit boundaries before running tasks on live systems, run the guard command and specify the directory to restrict edits to. The Skill relies on user input to set the boundary and uses hooks to verify safety boundaries before any tasks execute.

You should not use full safety mode for directory-bound edits when you need unrestricted access to modify files across multiple directories. This Skill imposes strict boundaries and destructive command warnings, which may interfere with tasks requiring broad system changes outside a single chosen path.