Can I use hooks to block dangerous commands?

Yes, PreToolUse hooks allow you to inspect tool inputs and return a 'deny' decision to prevent the execution of unauthorized or risky operations before they occur.

How do I persist environment variables across a session?

You can use a SessionStart hook to write export commands to the $CLAUDE_ENV_FILE, which makes those variables available to Claude throughout the entire active session.

Are multiple hooks executed sequentially or in parallel?

All matching hooks for a specific event are executed in parallel by Claude Code. This means they run independently and do not have access to each other's specific outputs.

What is the difference between prompt-based and command hooks?

Prompt-based hooks use natural language reasoning to make decisions based on context, while command hooks execute bash scripts for faster, deterministic operations like file system checks.

Claude Code Hook Development

Name: Claude Code Hook Development
Author: brixtonpham

bybrixtonpham

0•

開発者ツール

Facilitates the creation, implementation, and management of event-driven hooks for Claude Code plugins to automate workflows and enforce safety policies.

This skill provides a comprehensive framework for developing Claude Code hooks, enabling developers to intercept and respond to lifecycle events like tool usage, session starts, and task completion. By supporting both advanced prompt-based hooks for nuanced AI reasoning and high-performance bash command hooks for deterministic checks, it allows for sophisticated validation, context enrichment, and security enforcement. Whether you are building guardrails against dangerous commands or automating project-specific setup tasks, this skill ensures your Claude Code plugins are robust, secure, and context-aware.

主な機能

01Prompt-based hook implementation for context-aware AI reasoning and validation.

02Support for the full lifecycle of events including PreToolUse, PostToolUse, and Stop.

030 GitHub stars

04Advanced configuration guidance for both plugin-specific and user-level settings.

05Security best practices for preventing path traversal and credential leakage.

06Deterministic bash command hooks for high-performance file and system operations.

ユースケース

01Implementing custom session initialization scripts to load project-specific environment variables and context.

02Creating quality assurance hooks that verify task completion and test success before Claude stops a session.

03Building automated safety guardrails to block or modify risky terminal commands before execution.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add brixtonpham/dot-claude hook-development

For use in Claude.ai and ChatGPT

Download Skill