careful

Detect destructive Bash commands and prompt for confirmation before execution.

Updated Apr 1, 2026
One-click install
npx skills add https://github.com/viraatdas/vstack --skill careful-viraatdas
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: careful
Source: https://github.com/viraatdas/vstack/tree/main/careful
Command: npx skills add https://github.com/viraatdas/vstack --skill careful-viraatdas

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Destructive commands can cause data loss or system damage in shared or production environments. This Skill adds safety guardrails by inspecting Bash commands in real time and prompting before potentially destructive actions.

Core Features & Use Cases

  • Detects dangerous patterns like rm -rf, git reset --hard, kubectl delete, and similar commands.
  • Prompts for confirmation or allows override to proceed with explicit consent.
  • Logs attempts to ~/.gstack/analytics for auditing and improvement.

Quick Start

Enable careful mode in your shell to receive warnings before destructive commands are executed.

Frequently Asked Questions about careful

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

FAQPage Schema
How do I add safety guardrails to prevent destructive bash commands from running?

Safety guardrails for destructive bash commands are added by implementing pre-run hooks that analyze tool input for dangerous patterns, prompt for confirmation, and log usage analytics before execution proceeds.

Can I get a warning before executing rm -rf or git reset --hard in an interactive shell?

Yes, warnings before executing rm -rf or git reset --hard in interactive shells are triggered by real-time command inspection hooks that detect dangerous patterns and prompt for explicit confirmation to proceed.

How do destructive command warning hooks handle automation workflows?

Destructive command warning hooks handle automation workflows by inspecting bash tool input for dangerous commands, requiring explicit consent overrides, and logging attempts to analytics files for auditing shared environments.

Does this command safety guardrail support kubectl delete operations?

Yes, this command safety guardrail supports kubectl delete operations by applying pre-run pattern detection to identify potentially destructive Kubernetes commands before they impact production environments.

Where are destructive command attempts logged for auditing?

Destructive command attempts are logged to the ~/.gstack/analytics directory for auditing purposes, recording usage analytics to track safety warnings and confirmation prompts across shared environments.