careful

Warn users before executing destructive shell commands via a PreToolUse hook.

Updated Apr 12, 2026
One-click install
npx skills add https://github.com/2300031147/clawopencode --skill careful-2300031147
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: careful
Source: https://github.com/2300031147/clawopencode/tree/main/skills/careful
Command: npx skills add https://github.com/2300031147/clawopencode --skill careful-2300031147

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Safety guardrails for destructive commands warn before rm -rf, DROP TABLE, force-push, git reset --hard, kubectl delete, and similar operations. Users can override each warning in risky contexts.

Core Features & Use Cases

  • Destructive command detection and warnings for dangerous patterns like rm -rf, git reset --hard, kubectl delete, and more.
  • Safe exceptions allow certain trusted patterns to proceed without warnings.
  • Audit logging of usage to ~/.gstack/analytics for traceability.

Quick Start

Enable careful mode before risky operations and run your destructive command through the tool to receive a warning.

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 like rm -rf?

You can prevent accidental data loss from rm -rf by using a PreToolUse hook that detects destructive command patterns and issues a warning before execution. This applies safety guardrails across development, deployment, and production environments.

Can I override safety warnings for trusted git reset --hard operations?

Yes, you can override safety warnings for trusted git reset --hard operations using supported override workflows. The system allows certain trusted patterns to proceed without warnings while still logging the incident for traceability.

Does this destructive command warning system work with kubectl delete and docker prune?

The destructive command warning system works with kubectl delete, docker prune, and similar risky shell actions. It functions as a PreToolUse hook for Bash-based workflows to intercept dangerous patterns before they execute.

How do I audit and log incidents when risky shell commands are executed?

You can audit risky shell command incidents by logging usage to the analytics store at ~/.gstack/analytics. This provides full traceability of destructive command warnings and overrides within your Bash-based workflows.

What is the best way to block force-push and DROP TABLE commands in deployment scripts?

The best way to block force-push and DROP TABLE commands is implementing a PreToolUse hook that detects risky patterns and warns users before execution. Users can then override warnings for trusted operations while maintaining an audit log.