What are Claude Code hooks?

Hooks are user-defined shell commands that execute at specific points in Claude Code's lifecycle, allowing you to intercept, modify, or block tool usage and automate workflows.

Where is the configuration for Claude Code hooks stored?

Configuration can be stored globally in ~/.claude/settings.json, or at the project level in .claude/settings.json or .claude/settings.local.json.

How do I block a tool from running using a hook?

You can use a PreToolUse hook and return an exit code of 2 from your command script to block the tool from executing.

Can I use hooks to auto-format code?

Yes, by using a PostToolUse hook with a matcher for Write or Edit tools, you can trigger formatters like Prettier or Black every time Claude modifies a file.

What environment variables are available in hooks?

Hooks have access to CLAUDE_PROJECT_DIR (the project root) and CLAUDE_PLUGIN_ROOT (the plugin installation path).

Claude Code Hooks Reference

Name: Claude Code Hooks Reference
Author: Dunc4nJ

byDunc4nJ

•

Herramientas para Desarrolladores

Provides a comprehensive guide and reference for implementing lifecycle hooks to automate and control Claude Code behaviors.

The Hooks Reference skill empowers developers to master the Claude Code lifecycle by providing detailed documentation on event triggers, configuration schemas, and control flow logic. It enables the creation of custom automation scripts—such as auto-formatting on file writes, blocking sensitive file edits, or logging bash commands—ensuring a deterministic and secure AI coding environment. This reference is essential for users looking to integrate Claude Code into their existing local workflows or enforce specific guardrails during autonomous operations through shell commands and JSON configuration.

Características Principales

01Configuration guidance for user-level, project-level, and local settings files

021 GitHub stars

03Matcher pattern syntax for targeting specific tools like Bash, Write, or Edit

04Control flow logic documentation using exit codes and JSON output formats

05Detailed event lifecycle reference including PreToolUse, PostToolUse, and SessionStart

06Ready-to-use patterns for auto-formatting, security blocking, and command logging

Casos de Uso

01Setting up security guardrails to prevent Claude from accessing or modifying sensitive files like .env or .git

02Creating custom logging systems to track and audit every bash command Claude executes during a session

03Implementing automated linting or formatting scripts that run immediately after Claude edits a file

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add dunc4nj/entropy-plugins hooks-reference

For use in Claude.ai and ChatGPT

Download Skill