guard

Activate destructive command warnings and an edit boundary for protected directories.

Updated Apr 11, 2026
One-click install
npx skills add https://github.com/akiyoshi/gstack-copilot-jp --skill guard-akiyoshi
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: guard
Source: https://github.com/akiyoshi/gstack-copilot-jp/tree/main/.github/skills/guard
Command: npx skills add https://github.com/akiyoshi/gstack-copilot-jp --skill guard-akiyoshi

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Activates a comprehensive safety layer by combining destructive-command warnings with an edit boundary to prevent accidental or harmful changes during critical operations.

Core Features & Use Cases

  • Destructive command warnings: alerts before executing dangerous commands.
  • Edit boundary enforcement: limits file edits to a user-specified directory.
  • Safe single-command activation: quickly enable full safety without configuring multiple tools.

Quick Start

Ask the user which directory should be protected, then enable guard mode to enforce the edit boundary and warn on destructive commands.

Frequently Asked Questions about guard

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

FAQPage Schema
How do I prevent accidental destructive commands during critical deployment workflows?

To prevent accidental destructive commands during critical deployment workflows, you can activate a safety layer that intercepts risky operations and issues warnings before execution. This ensures irreversible edits are avoided.

How do I set an edit boundary to restrict file modifications to a specific directory?

To set an edit boundary that restricts file modifications, you specify a protected directory when enabling guard mode. The safety layer then enforces this boundary, preventing edits outside the user-defined directory.

What is the best way to enable full workflow protection for DevOps operations without configuring multiple tools?

The best way to enable full workflow protection without configuring multiple tools is to use a single-command safety mode. This combines destructive command warnings with edit boundary enforcement for immediate DevOps protection.

Does guard mode require any sibling hooks to enforce edit boundaries and warnings?

Yes, guard mode requires integration with sibling careful and freeze hooks to properly enforce the edit boundary and destructive command warnings during critical development and deployment operations.

When should I use an edit boundary instead of a general destructive command warning system?

An edit boundary is ideal when you must restrict changes to a specific directory during critical workflows. Destructive command warnings alone do not prevent unauthorized file edits outside the protected scope.