guard

Warn on destructive commands and restrict edits to a specified directory.

Updated Mar 27, 2026
One-click install
npx skills add https://github.com/Jesse-L-M/scz-target-engine --skill guard-jesse-l-m
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: guard
Source: https://github.com/Jesse-L-M/scz-target-engine/tree/main/.claude/skills/gstack/guard
Command: npx skills add https://github.com/Jesse-L-M/scz-target-engine --skill guard-jesse-l-m

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Protects production environments from destructive commands by combining warning-based safety with directory-bound edit restrictions, preventing accidental data loss or unauthorized changes.

Core Features & Use Cases

  • Destructive command warnings: alerts before executing rm -rf, DROP TABLE, or other dangerous actions.
  • Directory-bound edits: restricts edits to a user-specified path to prevent changes outside the allowed scope.
  • Coordination with related safeguards: integrates with sibling safety skills to provide end-to-end protection.

Quick Start

Enable guard mode by specifying a directory to restrict edits to and run the guard command.

Frequently Asked Questions about guard

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I prevent destructive commands like rm -rf from executing in a production environment?

To prevent destructive commands in production, you can enforce a safety layer that warns before executing dangerous actions like rm -rf or DROP TABLE. This restricts accidental data loss by blocking unauthorized modifications before they run.

What is the best way to restrict edits to a specific directory during live system debugging?

The best way to restrict edits during live debugging is to enforce directory-bound edit restrictions. This confines modifications to a user-specified path, preventing changes outside the allowed scope and protecting the rest of the project.

How do pre-execution safety hooks work to block accidental data loss?

Pre-execution safety hooks work by intercepting dangerous actions before they run, issuing warnings and freezing edits. They coordinate with sibling safety skills to block destructive commands and enforce strict edit boundaries across project directories.

Do I need to configure an edit boundary to protect live systems from unauthorized modifications?

Yes, you need to configure an edit boundary by specifying a directory path to restrict modifications. This prevents unauthorized changes across project directories and adds a warning layer against destructive commands during live system configuration.

Can I use directory-bound edit restrictions alongside other safety modes in my DevOps workflow?

Yes, directory-bound edit restrictions can coordinate with sibling safety skills in your DevOps workflow. This integration provides end-to-end dual protection by combining pre-execution warnings with strict edit freezing to secure live systems.

Why should I use a warning and freeze mode instead of standard command execution for production environments?

Warning and freeze mode actively intercepts dangerous commands before execution, preventing irreversible data loss. Standard execution lacks this pre-execution safety layer, leaving live systems vulnerable to accidental destructive actions and out-of-scope edits.