careful

Intercept destructive Bash commands via PreToolUse hook and return permission decisions.

Updated Mar 29, 2026
One-click install
npx skills add https://github.com/filipnyquist/caidos --skill careful-filipnyquist
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: careful
Source: https://github.com/filipnyquist/caidos/tree/main/.claude/skills/gstack/careful
Command: npx skills add https://github.com/filipnyquist/caidos --skill careful-filipnyquist

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

  • Pre-execution checks for risky Bash commands.
  • Interactive warnings and optional overrides to proceed after confirmation.
  • Logging of safety events and integration with Claude Skill workflow.

Quick Start

Enable careful mode and run a potentially destructive command to see the warning prompt.

Frequently Asked Questions about careful

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

FAQPage Schema
How do I add safety warnings before running destructive Bash commands like rm -rf?

You can guard destructive Bash commands by intercepting risky inputs before they run. This functionality uses a PreToolUse hook with a Bash matcher to detect patterns like rm -rf and return a permission decision, warning you before execution.

Can I override the safety warning when executing a risky git reset --hard?

Yes, interactive warnings prompt you before risky commands execute, providing optional user overrides to proceed after confirmation. This allows you to bypass the guardrail if you determine the destructive action is safe for your current environment.

What destructive operations does the Bash safety hook intercept?

The safety hook intercepts destructive operations including rm -rf, DROP TABLE, force-push, git reset --hard, and kubectl delete. It detects these risky patterns before execution and returns a permission decision to prevent accidental damage.

Does the Bash command guard work across development and production environments?

Yes, the Bash command guard applies across development and production environments to warn before dangerous actions. It is specifically useful when touching prod, debugging live systems, or working in a shared environment where caution is required.

How do I enable careful mode to protect shared environments from force-push?

You enable careful mode to activate safety guardrails for destructive commands. Once enabled, running a potentially destructive operation like a force-push triggers an interactive warning prompt to confirm before execution proceeds.

Are safety events logged when a kubectl delete command is intercepted?

Yes, logging of safety events is integrated into the workflow when risky commands like kubectl delete are intercepted. The PreToolUse hook detects the pattern and logs the event before returning a permission decision.