careful

Detects destructive shell commands and prompts for confirmation via a Bash pre-execution hook.

Updated Mar 31, 2026
One-click install
npx skills add https://github.com/ComputerConnection/zach-pack --skill careful-computerconnection
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: careful
Source: https://github.com/ComputerConnection/zach-pack/tree/main/skills/gstack-careful
Command: npx skills add https://github.com/ComputerConnection/zach-pack --skill careful-computerconnection

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Safety guardrails for destructive commands. Warns before rm -rf, DROP TABLE, force-push, git reset --hard, kubectl delete, and similar destructive operations. User can override each warning. Use when touching prod, debugging live systems, or working in a shared environment. Use when asked to "be careful", "safety mode", "prod mode", or "careful mode".

Core Features & Use Cases

  • Prevents accidental destructive actions by signaling warnings before execution.
  • Allows override with explicit confirmation to preserve workflow when needed.
  • Works as a safety layer in prod, staging, and shared environments, via a PreToolUse Bash hook.

Quick Start

Enable careful mode in your shell session and confirm each destructive command to proceed only when truly intended.

Frequently Asked Questions about careful

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

FAQPage Schema
How do I prevent destructive bash commands like rm -rf from running accidentally?

You can prevent destructive bash commands by using a pre-execution check hook that identifies operations like rm -rf in real time and prompts for confirmation before execution. This safety guardrail intercepts the command before it runs.

Can I override the safety warning when I intentionally need to run git reset --hard?

Yes, you can override the safety warning when you intentionally need to run git reset --hard. The guardrail prompts for explicit confirmation, allowing you to bypass the block and preserve your workflow when the destructive action is truly intended.

Does this destructive command guardrail work for kubectl delete operations in production?

Yes, this destructive command guardrail works for kubectl delete operations in production. It acts as a safety layer applicable in development, production touchpoints, and shared environments where destructive actions may occur.

What is a pre-execution check hook for bash and how does it stop force-push?

A pre-execution check hook for bash is a safety mechanism that scans commands before they run. It stops force-push by identifying the destructive operation in real time, warning you, and requiring explicit confirmation or override to proceed.

How do I enable safety mode for debugging live systems in a shared environment?

You enable safety mode for debugging live systems by activating the shell session guardrail. This implements a PreToolUse Bash hook that signals warnings before destructive operations, prompting you to confirm each action to proceed only when truly intended.

Are there limitations when using bash guardrails to warn about DROP TABLE commands?

The bash guardrail warns about DROP TABLE commands by scanning shell input, but its scope is limited to bash commands. It provides audit-ready logging and override capabilities, though it does not block execution if you explicitly confirm the destructive action.