careful

Prevent destructive shell command execution without explicit confirmation.

Updated Apr 2, 2026
One-click install
npx skills add https://github.com/jonkiky/ccdi-federation-ai --skill careful-jonkiky
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: careful
Source: https://github.com/jonkiky/ccdi-federation-ai/tree/main/.agents/careful
Command: npx skills add https://github.com/jonkiky/ccdi-federation-ai --skill careful-jonkiky

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Safety guardrails prevent accidental execution of destructive shell commands. Users can override warnings to proceed when necessary.

Core Features & Use Cases

  • Detect dangerous commands such as rm -rf, git reset --hard, kubectl delete, and docker prune.
  • Prompt with a warning and require explicit confirmation before proceeding.
  • Allow safe exceptions for common build artifacts and test data.
  • Log usage analytics locally for auditing and improvement.

Quick Start

Enable careful mode and run a dangerous command to see the warning and decision prompt.

Frequently Asked Questions about careful

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

FAQPage Schema
How do I prevent accidental execution of destructive shell commands like rm -rf?

To prevent destructive shell commands like rm -rf, you need safety guardrails that enforce pre-run checks and require explicit confirmation before execution. This intercepts dangerous commands via a PreToolUse hook to stop data loss.

Can I add safe exceptions for build artifacts when blocking destructive commands?

Yes, you can add safe exceptions for common build artifacts and test data. This allows specific non-critical files to bypass the destructive command warnings while still protecting essential system data.

Does the shell safety guardrail work with git reset --hard and kubectl delete?

Yes, the shell safety guardrail works with git reset --hard and kubectl delete. It detects these specific destructive commands, prompts a warning, and requires explicit confirmation before allowing them to proceed.

What is the best way to add a pre-hook for destructive command confirmation in bash?

The best way to add a pre-hook for destructive commands in bash is to enforce pre-run checks using a PreToolUse hook. This mechanism prompts explicit confirmation and logs usage locally for auditing.

Why do I need to confirm destructive commands during production debugging?

You need to confirm destructive commands during production debugging because commands like docker prune could cause irreversible data loss. Safety guardrails ensure you explicitly approve these actions before execution.

Can I override the destructive command warning if I need to proceed?

Yes, you can override the warning to proceed when necessary. The safety guardrails prompt you with a warning and require explicit confirmation, allowing you to bypass the block if you intentionally need to run the command.