guard

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

Updated Apr 5, 2026
One-click install
npx skills add https://github.com/bizzybae/qstack --skill guard-bizzybae
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: guard
Source: https://github.com/bizzybae/qstack/tree/main/gstack-original/guard
Command: npx skills add https://github.com/bizzybae/qstack --skill guard-bizzybae

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Full safety mode for shell operations that warns before destructive commands and blocks edits outside a defined boundary.

Core Features & Use Cases

  • Destructive command warnings — alerts before executing rm -rf, DROP TABLE, or other risky commands.
  • Edit boundary enforcement — restricts edits to a user-specified directory and blocks edits elsewhere.
  • Coordination with safety tools — relies on sibling /careful and /freeze scripts to implement guardrails in production-like sessions.

Quick Start

Enable guard mode and specify the directory that edits should be restricted to.

Frequently Asked Questions about guard

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

FAQPage Schema
How do I prevent accidental data loss from destructive shell commands?

To prevent accidental data loss, this tool intercepts and warns before destructive shell commands like rm -rf or DROP TABLE execute. It provides a fail-safe boundary in interactive sessions to stop mistakes before they happen.

Can I restrict shell edits to a specific directory and block changes elsewhere?

Yes, you can restrict shell edits to a specific directory by enforcing an edit boundary. The tool blocks any modifications attempted outside your user-specified directory, ensuring production environments remain untouched.

How do destructive command warnings work in interactive shell sessions?

Destructive command warnings work by monitoring interactive shell sessions and alerting you before risky commands execute. The system coordinates with sibling safety scripts to enforce guardrails and provide reliable fail-safe behavior.

Do I need sibling safety tools to enforce guardrails in production-like sessions?

Yes, you need sibling careful and freeze safety tools to enforce guardrails in production-like sessions. The guard coordinates with these scripts to implement reliable, fail-safe behavior and maintain strict operational boundaries.

What is the best way to block destructive commands like DROP TABLE in a shell environment?

The best way to block destructive commands like DROP TABLE is to enable a full safety mode that warns before execution. It applies directly to interactive sessions and isolates edits to protect against accidental data loss.