careful

Intercept bash commands matching destructive patterns and issue user warnings.

Updated May 9, 2026
One-click install
npx skills add https://github.com/kk20300113-png/my-claude-skills --skill careful-kk20300113-png
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: careful
Source: https://github.com/kk20300113-png/my-claude-skills/tree/main/careful
Command: npx skills add https://github.com/kk20300113-png/my-claude-skills --skill careful-kk20300113-png

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Running accidental destructive commands like rm -rf, git reset --hard, or kubectl delete can cause irreversible data loss, service downtime, or lost uncommitted work, especially when operating in production environments or shared development setups where a single typo has major consequences.

Core Features & Use Cases

  • Destructive command interception: Automatically scans all bash commands for high-risk patterns including recursive file deletions, database drops, force git pushes, and container deletions before they execute.
  • Safe exception handling: Allows non-destructive common deletions (like removing node_modules or build folders) to run without unnecessary warnings.
  • Use case example: When debugging a live production server, if you accidentally type a command to delete critical application data, the skill will prompt you to confirm before proceeding, preventing accidental outages.

Quick Start

Enable careful mode before running any commands on your production infrastructure or shared team codebase to get automatic warnings for risky operations.

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 in production?

To prevent accidental data loss from destructive bash commands, a safety skill can intercept your bash tool inputs and match them against high-risk patterns like recursive deletions before execution. It issues a warning and allows user override for intentional operations.

Can I get a warning before running git reset hard or kubectl delete?

Yes, you can get warnings before running git reset hard or kubectl delete. The skill automatically scans all bash commands for high-risk patterns including force git pushes and container deletions, prompting you for confirmation before they execute.

Does command safety interception block safe deletions like removing node_modules?

Command safety interception does not block safe deletions like removing node_modules or build folders. It features safe exception handling that allows non-destructive common deletions to run automatically without issuing unnecessary warnings.

What is the best way to protect shared development environments from unintended command execution?

The best way to protect shared development environments from unintended command execution is enabling a careful mode that intercepts destructive system commands. This prevents service disruption and irreversible data loss by requiring explicit user confirmation for matched high-risk operations.

How do I override a false positive warning for a destructive command I intend to run?

You can override a false positive warning for a destructive command you intend to run by using the user override feature. The skill issues warnings for matched commands but explicitly allows you to bypass them to proceed with intentional operations.