guard

Restrict edits to a defined directory and warn before destructive commands.

Updated Apr 12, 2026
One-click install
npx skills add https://github.com/adryanmoldokkr32-pixel/gstack --skill guard-adryanmoldokkr32-pixel
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: guard
Source: https://github.com/adryanmoldokkr32-pixel/gstack/tree/main/guard
Command: npx skills add https://github.com/adryanmoldokkr32-pixel/gstack --skill guard-adryanmoldokkr32-pixel

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Full safety mode combines destructive command warnings with directory-scoped edit restrictions to prevent accidental or malicious damage.

Core Features & Use Cases

  • Combines /careful and /freeze safety checks to warn before destructive commands and to lock edits outside a defined directory.
  • Suitable for production maintenance, live debugging, and any operation requiring strict change boundaries across a project.

Quick Start

Activate guard mode to enable destructive-command warnings and edit boundaries by selecting a restricted directory.

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 during production maintenance?

Preventing destructive commands during production maintenance requires enabling guard mode to warn before executing destructive actions and lock edits to a defined directory. It combines pre-checks and freeze boundaries to restrict tool usage.

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

Restricting edits to a specific directory during live debugging involves activating guard mode to enforce a freeze boundary. This locks file modifications outside your selected directory while warning against destructive commands.

Can I combine pre-checks and freeze boundaries to lock down project changes?

Yes, you can combine pre-checks and freeze boundaries by enabling guard mode. It relies on sibling safety mechanisms to perform pre-checks and enforce a directory-scoped freeze boundary before tool usage, ensuring strict change boundaries.

Does guard mode work without defining a restricted directory?

Guard mode requires selecting a restricted directory to function properly. Without defining this boundary, it cannot enforce directory-scoped edit restrictions or apply its safety checks to prevent accidental damage during operations.

When should I not use unrestricted edits during prod maintenance?

You should not use unrestricted edits during prod maintenance when strict change boundaries are required. Guard mode applies warnings for destructive commands and locks edits outside a defined directory to prevent accidental or malicious damage.