careful

Assess destructive commands and prompt for confirmation before execution.

Updated Apr 7, 2026
One-click install
npx skills add https://github.com/zz8011/harnesstest --skill careful-zz8011
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: careful
Source: https://github.com/zz8011/harnesstest/tree/main/.claude/skills/gstack/careful
Command: npx skills add https://github.com/zz8011/harnesstest --skill careful-zz8011

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Safety risk from executing destructive commands by mistake; it provides warnings and guardrails to prevent accidental data loss.

Core Features & Use Cases

  • Warnings before destructive commands (rm -rf, git reset --hard, kubectl delete, etc.)
  • User override to proceed after warning
  • Safe-exceptions handling for common build artifacts (node_modules, dist, pycache)

Quick Start

Enable careful mode before running any destructive command to receive a warning and an override option.

Frequently Asked Questions about careful

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

FAQPage Schema
How do I prevent accidental data loss from destructive bash commands like rm -rf?

Preventing data loss from destructive bash commands requires a pre-run hook that analyzes command inputs and prompts for confirmation before execution. This mechanism intercepts risk-prone operations to ensure intentional actions.

Can I configure a whitelist to skip warnings for safe build artifacts?

Configuring a whitelist skips warnings for safe build artifacts by avoiding blocking safe targets. This safe-exceptions handling specifically targets directories like node_modules, dist, and __pycache__ to streamline debugging.

Does this command safety guardrail work with kubectl and git operations?

This command safety guardrail works with kubectl and git operations by assessing risk-prone actions across development, operations, and production environments. It guards commands like git reset --hard and kubectl delete during deployment.

What is the best way to add a confirmation prompt before executing risky shell commands?

The best way to add a confirmation prompt for risky shell commands is implementing a pre-run hook that filters destructive actions. It warns before execution and exposes configurable override options to proceed after warning.

Why do I need a warning system for production environment deployments?

A warning system for production environment deployments is needed to guard against destructive operations executed by mistake. It provides risk prevention by prompting for confirmation before potentially harmful actions impact production.

How to override a destructive command warning when I intentionally want to run it?

To override a destructive command warning intentionally, you use the exposed configurable override options to proceed after the warning. The pre-run hook prompts for confirmation, allowing you to bypass the guardrail.