careful

Warn before executing destructive Bash commands via a PreToolUse hook.

1|Updated Mar 28, 2026
One-click install
npx skills add https://github.com/shhubbh/flowstate --skill careful-shhubbh
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: careful
Source: https://github.com/shhubbh/flowstate/tree/main/.claude/skills/gstack/careful
Command: npx skills add https://github.com/shhubbh/flowstate --skill careful-shhubbh

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Destructive commands can cause irreversible damage. This skill provides safety guardrails that warn before executing dangerous operations such as rm -rf, DROP TABLE, git reset --hard, kubectl delete, and similar actions.

Core Features & Use Cases

  • Pre-run warnings for potentially destructive Bash commands.
  • Optional override to proceed after a warning when safety allowances are granted.
  • Local analytics logging to surface safety reviews without exposing sensitive data.

Quick Start

Enable careful mode before running risky commands and confirm the warning to proceed or cancel.

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 Bash commands like rm -rf in production?

You can prevent accidental execution of destructive Bash commands by enabling guardrails that warn before running risky actions. This skill acts as a PreToolUse hook to inspect commands like rm -rf and block them unless explicitly overridden.

What is the best way to add safety guardrails for git push --force and kubectl delete operations?

The best way to add safety guardrails for git push --force and kubectl delete is using a pre-run warning system. It detects these destructive patterns, blocks execution by default, and logs the safety events for auditing.

Can I override the warning and proceed if I intentionally need to run a DROP TABLE command?

Yes, you can override the warning to proceed with a DROP TABLE command when safety allowances are granted. The system prompts you to confirm the warning, allowing you to cancel or explicitly approve the destructive action.

Does this destructive command blocker work with Git, DevOps, and Kubernetes workflows?

Yes, this destructive command blocker works with Git, DevOps, and Kubernetes workflows. It guards risky actions across these systems, including git reset --hard and kubectl delete, by intercepting unsafe Bash commands before execution.

Why should I use a command warning hook instead of manually checking scripts for destructive operations?

You should use a command warning hook because it automates safety inspections and logs events for auditing, eliminating human error. Manually checking scripts is error-prone, whereas this hook reliably blocks destructive patterns like rm -rf unless overridden.

When should I not use automated guardrails for shell operations?

You should not use automated guardrails for shell operations when running fully trusted, automated pipelines where human confirmation prompts would cause execution to hang. The override mechanism is designed for interactive debugging, not silent background tasks.