Does this support automatic code formatting?

Absolutely. You can configure a PostToolUse hook to trigger tools like Prettier, Black, or ESLint immediately after Claude finishes editing a file.

How do I block a tool from running?

You can use a PreToolUse hook with a shell command that exits with code 2. This prevents the tool from executing and allows you to provide specific feedback to the Claude agent.

Where are hook configurations stored?

Hooks can be stored globally in ~/.claude/settings.json for all projects or locally within a project's .claude/settings.json file for repository-specific behavior.

What are Claude Code hooks?

Hooks are specialized scripts or prompts that trigger during specific lifecycle events, such as starting a session or before a tool is used, to customize or restrict agent behavior.

Can I use LLMs to evaluate hooks?

Yes, 'prompt' type hooks allow you to use natural language instructions and context to decide whether a tool should proceed or if a session should stop.

Hook Creator

Name: Hook Creator
Author: GzuPark

byGzuPark

•

개발자 도구

Automates and customizes Claude Code behavior by creating shell-based or LLM-driven lifecycle hooks for tasks like formatting, security, and logging.

Hook Creator empowers developers to extend Claude Code's functionality by configuring hooks that execute during specific lifecycle events. Whether you need to auto-format code with Prettier after an edit, block access to sensitive files like .env, or log command history to external files, this skill provides a structured workflow for implementation. It supports both shell-based commands for performance and prompt-based LLM evaluations for complex decision-making, allowing for deep integration between Claude and your local development environment.

주요 기능

01Automated formatting and linting after file edits

02Tool execution logging and session monitoring

031 GitHub stars

04File protection and access control for sensitive data

05Prompt-based hooks for LLM-driven event logic

06Auto-approval workflows for permission requests

사용 사례

01Creating comprehensive audit trails of all CLI commands executed by the AI agent.

02Enforcing security policies by blocking edits to environment files or protected directories.

03Streamlining developer workflows by automatically running tests or formatters after Claude modifies code.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add gzupark/claude-plugin-pack hook-creator

For use in Claude.ai and ChatGPT

Download Skill