careful

Intercept Bash commands and warn before destructive actions via PreToolUse hook.

Updated Mar 13, 2026
One-click install
npx skills add https://github.com/soyasis/gstack-private --skill careful-soyasis
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: careful
Source: https://github.com/soyasis/gstack-private/tree/main/careful
Command: npx skills add https://github.com/soyasis/gstack-private --skill careful-soyasis

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Safety mode is active and prevents destructive commands by validating Bash inputs before execution, warning the user, and allowing override if necessary.

Core Features & Use Cases

  • Pre-execution detection of destructive patterns like rm -rf, git reset --hard, kubectl delete, and docker prune.
  • Interactive warnings with the option to proceed or cancel to safeguard production systems and shared environments.
  • Automatic logging of destructive-command attempts to a local analytics store for audit and improvement.

Quick Start

Enable careful mode and begin issuing commands as usual to receive a prompt before any destructive action.

Frequently Asked Questions about careful

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

FAQPage Schema
How do I prevent destructive shell commands from running automatically?

You can prevent destructive shell commands by intercepting Bash inputs before execution, which validates commands like rm -rf and git reset --hard, then issues an interactive warning to proceed or cancel.

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

Yes, enabling safety mode provides pre-execution detection for destructive patterns like rm -rf, git reset --hard, and kubectl delete, issuing an interactive warning that allows you to proceed or cancel.

Does this pre-command safety check work for production and shared environments?

Yes, the pre-command safety check applies across development and production workflows, safeguarding shared environments and sensitive data by intercepting destructive Bash commands before execution.

How do I log attempts to run destructive bash commands for auditing?

You can log destructive command attempts automatically by enabling safety mode, which records intercepted commands to a local analytics store for audit and improvement.

Can I override the guardrails if I need to run a required destructive command?

Yes, the interactive warnings include an override option, allowing you to proceed with required destructive actions while maintaining safe debugging in protected workflows.