Can hooks be used for code formatting?

Yes, the PostToolUse hook is the standard way to trigger formatters like Prettier or linters like ESLint immediately after Claude finishes writing or editing a file.

Where do I configure my Claude Code hooks?

Hooks are configured in your .claude/settings.json file, which can be located in your project root for project-specific hooks or in your home directory for global hooks.

What are Claude Code hooks?

Hooks are automated scripts or commands that execute at specific points in the Claude Code workflow, such as before a tool runs, after a file is edited, or when a session ends.

Why use the claude-hooks-sdk instead of plain bash?

The SDK provides type safety, built-in logging, and a structured manager to handle complex JSON input/output from Claude, making your hooks more maintainable and less error-prone.

How do I block a dangerous command in Claude Code?

You can use a PreToolUse hook to inspect the tool input and return an exit code of 2 or a JSON object with a 'deny' decision to block the command from executing.

Claude Code Hooks SDK

Name: Claude Code Hooks SDK
Author: hgeldenhuys

byhgeldenhuys

•

생산성 및 워크플로

Implements automated lifecycle hooks and script extensions to customize and secure Claude Code's execution behavior.

소개

This skill provides a comprehensive implementation guide and a type-safe TypeScript SDK for creating Claude Code hooks, allowing for deep customization of the agent's workflow. It enables developers to execute custom bash commands or scripts at 9 specific lifecycle points, such as before tool execution (validation), after file writes (formatting), or during prompt submission (context injection). By utilizing the provided SDK, users can build robust, production-grade automations that improve developer productivity, enforce security guardrails, and ensure consistent code quality across AI-assisted development sessions.

주요 기능

Security validation to intercept and block dangerous terminal commands
3 GitHub stars
Support for all 9 Claude Code hook events including PreToolUse and SessionStart
Automated code formatting and linting integration following AI-driven edits
Type-safe TypeScript SDK for building complex hook logic with Bun
Context injection capabilities to enhance Claude's awareness of session metadata

사용 사례

Enforcing security policies by blocking destructive shell commands during agent execution
Pre-loading environment variables and project-specific context at the start of every session
Triggering Prettier or ESLint automatically after Claude modifies source code

소개

주요 기능

Security validation to intercept and block dangerous terminal commands
3 GitHub stars
Support for all 9 Claude Code hook events including PreToolUse and SessionStart
Automated code formatting and linting integration following AI-driven edits
Type-safe TypeScript SDK for building complex hook logic with Bun
Context injection capabilities to enhance Claude's awareness of session metadata

사용 사례

Enforcing security policies by blocking destructive shell commands during agent execution
Pre-loading environment variables and project-specific context at the start of every session
Triggering Prettier or ESLint automatically after Claude modifies source code