careful

Intercept destructive shell commands with a PreToolUse hook to warn before execution.

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Intercepts destructive shell commands to warn before execution, reducing the risk of data loss in shared environments.

Core Features & Use Cases

  • Pre-run checks that detect dangerous commands like rm -rf, git reset --hard, kubectl delete, and similar patterns.
  • Optional override capability to proceed after a warning in prod, debugging, or collaborative workflows.
  • Works across Bash-based workflows, local development, and CI pipelines to prevent accidental damage.

Quick Start

Enable careful mode to prompt before executing dangerous commands.

Frequently Asked Questions about careful

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

FAQPage Schema
How do I intercept and warn before destructive bash commands execute?

To intercept destructive bash commands, you need a PreToolUse hook running a Bash script to detect dangerous patterns like rm -rf or git reset --hard, warning you before execution to prevent accidental data loss.

Can I add guardrails to prevent accidental data loss in CI pipelines?

Yes, you can add precheck guardrails to CI pipelines by implementing a Bash script hook that detects dangerous operations and prompts an override, preventing accidental damage to infrastructure or data.

Does this safety hook work for local development machines and shared environments?

This safety hook works across local development machines, shared environments, and CI pipelines, applying pre-run checks to Bash-based workflows to reduce the risk of destructive operations impacting data.

What dangerous shell patterns does the precheck detection look for?

The precheck detection looks for dangerous shell patterns like rm -rf, git reset --hard, kubectl delete, and similar destructive commands to provide a warning and optional override before execution.

Is there a way to override the destructive command warning during debugging?

Yes, there is an optional override capability allowing you to proceed after a warning during debugging, production, or collaborative workflows when a dangerous command is intentionally required.