careful

Warn before executing destructive shell and deployment commands via a PreToolUse hook.

Updated Feb 24, 2026
One-click install
npx skills add https://github.com/gaaschk/gaasch-family --skill careful-gaaschk
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: careful
Source: https://github.com/gaaschk/gaasch-family/tree/main/.claude/skills/gstack/careful
Command: npx skills add https://github.com/gaaschk/gaasch-family --skill careful-gaaschk

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Safety guardrails for destructive commands to prevent accidental damage in development, testing, and production environments.

Core Features & Use Cases

  • Warns before destructive commands like rm -rf, DROP TABLE, git reset --hard, kubectl delete, and more, with an override option to proceed if necessary.
  • Logs usage events for auditing and guardrail improvement.
  • Suitable for collaborative environments to reduce risk during debugging and live-system work.

Quick Start

Enable careful mode to warn before destructive commands and allow override.

Frequently Asked Questions about careful

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

FAQPage Schema
How do I prevent accidental damage from destructive commands like rm -rf or git reset --hard?

To prevent accidental damage from destructive commands, you need safety guardrails that warn before execution. This Skill intercepts shell and deployment operations like rm -rf and git reset --hard, prompting for confirmation before proceeding.

Can I override destructive command warnings when I intentionally need to run them?

Yes, you can override destructive command warnings for authorized actions. The Skill provides an explicit override option, allowing you to proceed with necessary destructive operations while still logging the event for auditing.

Does this safety guardrail work with database and Kubernetes operations like DROP TABLE and kubectl delete?

Yes, this safety guardrail works with database and Kubernetes operations. It detects destructive patterns like DROP statements and kubectl delete commands across services, returning a warning to prevent accidental data loss in production environments.

How do shell safety guardrails analyze commands before they execute?

Shell safety guardrails analyze commands using a PreToolUse hook that inspects the incoming command pattern. If a destructive operation is detected, it returns a permission decision requesting user confirmation before the command executes.

What is the best way to audit destructive command usage in a collaborative development environment?

The best way to audit destructive command usage is by logging override events locally. This Skill records usage analytics for guardrail events, providing an audit trail to reduce risk during collaborative debugging and live-system work.

Are there limitations when using command override options for authorized destructive actions?

A limitation of using the command override option is that authorized destructive actions will still execute after the warning. While the guardrail logs the event for auditing, it cannot reverse operations like git push --force once the override is applied.