Why is the 'set -euo pipefail' header important for hooks?

This header ensures that a hook script fails immediately if any command within it errors or if an unset variable is used, preventing Claude from proceeding under potentially unsafe or unpredictable conditions.

What is a Claude Code hook?

Hooks are automated actions triggered at specific points in a Claude Code session—such as before a tool is used or after a response finishes—allowing for validation, monitoring, and control of AI actions via bash commands or LLM prompts.

How does this skill improve hook security?

It identifies dangerous command patterns like recursive deletes, hard resets, or remote code execution and ensures scripts use the correct exit code 2 to successfully block unsafe AI actions.

Can this skill fix my hook scripts automatically?

Yes, when run with the --fix flag, the skill can automatically inject safety headers like 'set -euo pipefail', add missing exit codes, and generate required metadata frontmatter.

Hook Auditor & Safety Guard

Name: Hook Auditor & Safety Guard
Author: avifenesh

byavifenesh

•

311

•

セキュリティとテスト

Analyzes and fixes Claude Code hook scripts to ensure security, proper error handling, and lifecycle compliance.

The enhance-hooks skill provides an automated auditing and optimization framework for Claude Code hooks, ensuring your session automations are safe, efficient, and correctly configured. It scans hook scripts for dangerous patterns like 'rm -rf' or 'curl | sh', validates YAML frontmatter for metadata consistency, and enforces shell script best practices such as fail-fast headers. By verifying hook lifecycle events and timeout configurations, it helps developers build robust, reliable workflows that extend Claude's capabilities without compromising system security or project integrity.

主な機能

01Validation of hook lifecycle events and tool matcher regex syntax

02Frontmatter verification to ensure hooks have required names and descriptions

03311 GitHub stars

04Automated safety scanning for dangerous shell commands and destructive actions

05Auto-fix capabilities for missing safety headers and incorrect exit codes

06Output schema validation for PreToolUse and Stop event decision-making

ユースケース

01Standardizing project hooks with consistent error handling and timeout configurations

02Enforcing security policies by blocking dangerous commands via PreToolUse hooks

03Auditing third-party automation scripts before deployment in local environments

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add avifenesh/awesome-slash hooks

For use in Claude.ai and ChatGPT

Download Skill