hook-create

Creates deterministic PreToolUse and PostToolUse hooks for Claude Code tool executions.

Updated Jan 16, 2026
One-click install
npx skills add https://github.com/hyxklee/AI-DEV-ToolKit --skill hook-create
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: hook-create
Source: https://github.com/hyxklee/AI-DEV-ToolKit/tree/main/claude/skills/hook-create
Command: npx skills add https://github.com/hyxklee/AI-DEV-ToolKit --skill hook-create

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Provides deterministic hooks to enforce automation and safety in tool executions, preventing unintended deviations by LLMs.

Core Features & Use Cases

  • Create PreToolUse and PostToolUse hooks to control actions, block dangerous commands, and inject context.
  • Support session-level and user-prompt level hooks for consistent behavior across interactions.
  • Scenario: automatically format outputs, validate inputs, and guard sensitive operations in Claude Code workflows.

Quick Start

Configure and apply Claude Code hooks to enforce determinism for tool executions within your project.

Frequently Asked Questions about hook-create

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate deterministic safety checks before tool use in Claude Code?

Deterministic hooks in Claude Code enforce guaranteed formatting, input validation, and safety checks by configuring PreToolUse and PostToolUse events to prevent unintended deviations by LLMs during tool interactions.

Can I block dangerous commands and validate tool interactions automatically?

Yes, you can block dangerous commands and validate tool interactions by configuring PreToolUse hooks that intercept tool executions, guard sensitive operations, and enforce safety protocols automatically within your workflows.

How do I configure session-level hooks for consistent formatting in Claude Code workflows?

You configure session-level and user-prompt level hooks in Claude Code to maintain consistent behavior and automatically format outputs across interactions by applying deterministic settings to your project environment.

What is the best way to inject context automatically before tool use?

The best way to inject context automatically is by setting up PreToolUse hooks that enforce deterministic automation, ensuring the right context is applied before any tool interaction occurs in your Claude Code workflow.

Why do I need deterministic hooks for tool use instead of relying on LLM judgment?

Deterministic hooks are needed to prevent unintended deviations by LLMs, ensuring guaranteed formatting, validation, and safety checks execute reliably before and after tool use without relying on probabilistic model behavior.