careful

Warn before destructive commands via a PreToolUse hook.

Updated Apr 1, 2026
One-click install
npx skills add https://github.com/whd4/gstack --skill careful-whd4
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: careful
Source: https://github.com/whd4/gstack/tree/main/careful
Command: npx skills add https://github.com/whd4/gstack --skill careful-whd4

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

  • Destructive-command detection and warning before execution.
  • Granular allow/override controls per command.
  • Auditable usage analytics and session safety.

Quick Start

Enable careful mode before running potentially destructive commands to see warnings and decide whether to proceed.

Frequently Asked Questions about careful

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

FAQPage Schema
How do I get a warning before running destructive bash commands like rm -rf?

Careful mode intercepts destructive bash commands like rm -rf via a PreToolUse hook, returning a warning and prompt instead of executing. It prevents accidental data loss during production maintenance and debugging in shared environments.

Can I override the safety warning if I need to force push or reset git?

Yes, you can override the safety warning on a per-command basis. When careful mode intercepts a destructive action like git reset --hard, it prompts you, providing granular allow controls to proceed if you decide the operation is safe.

What destructive operations does careful mode detect in production environments?

Careful mode detects destructive operations including rm -rf, DROP TABLE, force-push, git reset --hard, and kubectl delete. It scans for these patterns during production maintenance to prevent accidental data loss.

Does this safety guardrail work for database commands like DROP TABLE?

Yes, the safety guardrail works for database commands like DROP TABLE. It implements guardrails via a PreToolUse hook that scans for destructive SQL patterns, returning a warning and prompt to prevent accidental data loss.

How do I enable safety mode when debugging live systems?

Enable safety mode before running potentially destructive commands when debugging live systems. The guardrail activates via a PreToolUse hook to scan for destructive actions, ensuring you receive warnings and can decide whether to proceed.

When should I not use destructive command guardrails in bash?

You should not use destructive command guardrails if you require fully automated, uninterrupted execution of scripts containing operations like kubectl delete. The per-command override prompt requires manual intervention, halting automated workflows.