Can I block Claude from running certain commands?

Yes, by setting the 'action' field to 'block' and specifying a 'bash' event with a matching regex pattern, you can prevent specific commands from executing.

What is a Hookify rule?

A Hookify rule is a markdown file with YAML frontmatter stored in the .claude/ directory that instructs Claude how to react to specific patterns in commands, file edits, or session prompts.

What event types are supported by Hookify?

Hookify supports 'bash' for commands, 'file' for edits/writes, 'stop' for session endings, 'prompt' for user inputs, and 'all' for universal triggers.

Where are Hookify rules stored?

Rules are stored in the project's .claude/ folder and should follow the naming convention '.claude/hookify.{name}.local.md'.

Do I need to restart Claude to apply new rules?

No, Hookify rules are read dynamically and take effect on the very next tool use after the rule file is created or updated.

Hookify Rule Writer

Name: Hookify Rule Writer
Author: himmelreich-it

byhimmelreich-it

0•

개발자 도구

Standardizes and automates the creation of custom guardrails and workflow rules for Claude Code using markdown-based syntax.

The Hookify Rule Writing skill empowers developers to define custom behaviors, warnings, and restrictions for Claude Code through a simple yet powerful rule system. By configuring specific triggers for bash commands, file edits, and session completions, users can enforce coding standards, prevent dangerous operations like unauthorized deletions, and ensure project-specific checklists are followed. This skill provides comprehensive syntax guidance, regex patterns, and implementation best practices to help teams maintain high-quality codebases and secure development environments while working with AI agents.

주요 기능

01Advanced conditional logic using regex, substring, and prefix matching

02Customizable intervention levels with warning and blocking actions

03Project-specific guardrails for sensitive files and dangerous CLI operations

04Markdown-based rule configuration with YAML frontmatter for easy management

05Multi-event support for bash commands, file edits, and agent stop triggers

060 GitHub stars

사용 사례

01Preventing accidental execution of destructive shell commands like 'rm -rf'

02Displaying a mandatory completion checklist before the AI agent stops its task

03Enforcing a 'no console.log' or 'no debugger' policy in production-bound code

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add himmelreich-it/agent-skill-converter writing-rules

For use in Claude.ai and ChatGPT

Download Skill