careful

Inspect Bash input and warn before destructive commands like rm, git, and kubectl.

Updated Apr 18, 2026
One-click install
npx skills add https://github.com/algorithmbasics/gstack --skill careful-algorithmbasics
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: careful
Source: https://github.com/algorithmbasics/gstack/tree/main/careful
Command: npx skills add https://github.com/algorithmbasics/gstack --skill careful-algorithmbasics

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". (gstack)

Core Features & Use Cases

  • Provides real-time warnings before executing dangerous shell actions (rm -rf, git push --force, kubectl delete, etc.)
  • Supports per-command overrides to continue after a warning
  • Logs guardrail events locally for auditing and improvement

Quick Start

Enable careful mode and run your destructive command to receive a warning with an optional override.

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 enabling a safety mode that inspects shell input and issues a pre-run warning before executing risky actions like rm -rf. This prevents accidental data loss by requiring an explicit override to proceed.

Can I override the warning when force-pushing or dropping tables in a shared environment?

Yes, safety warnings for destructive commands support per-command overrides, allowing you to continue executing actions like git push --force or DROP TABLE after acknowledging the risk in shared environments.

Does this safety guardrail work with kubectl delete and git reset --hard operations?

Yes, the safety guardrails cover kubectl delete, git reset --hard, and similar destructive operations by detecting risky patterns through pre-run hooks during shell interactions across sensitive repos and prod environments.

What is the best way to prevent accidental data loss when debugging live prod systems?

The best way to prevent accidental data loss during prod debugging is to enable careful mode, which applies pre-run hooks to inspect bash input, warn about destructive actions, and log guardrail events locally for auditing.

Are guardrail events logged locally when a destructive command warning is triggered?

Yes, the safety guardrails log guardrail events locally for auditing and improvement whenever a destructive command warning is triggered, ensuring you have a record of risky shell interactions and any applied overrides.