Are Hookify rules shared across different projects?

Hookify rules ending in '.local.md' are typically project-specific and should be added to your .gitignore to keep them local to your current workspace.

What events can be used to trigger a rule?

Rules can be triggered by 'bash' commands, 'file' edits (writing or modifying files), 'stop' events (when the agent finishes), or 'prompt' events (user input).

Can Hookify rules prevent Claude from running specific commands?

Yes, by setting the 'action' field to 'block' and defining a pattern for the command, you can stop Claude from executing operations like 'rm -rf' or unauthorized script runs.

What is a Hookify rule in Claude Code?

A Hookify rule is a configuration file stored in the .claude/ directory that uses regex patterns to monitor Claude's behavior and trigger specific warning messages or blocks.

How do I create a new Hookify rule?

Create a markdown file named '.claude/hookify.{name}.local.md' with YAML frontmatter defining the name, event type, and pattern to match.

Hookify Rule Assistant

Name: Hookify Rule Assistant
Author: Activer007

byActiver007

0•

開発者ツール

Streamlines the creation and configuration of Hookify rules to enforce coding standards and prevent dangerous operations in Claude Code.

The Hookify Rule Assistant skill provides a comprehensive framework for creating specialized monitoring rules within Claude Code. By defining patterns in YAML frontmatter and custom markdown messages, users can intercept Claude's actions across bash commands, file edits, and session completions. This skill helps developers implement automated safety checks, prevent the accidental commit of sensitive data, and enforce project-specific best practices by providing real-time feedback and warnings directly within the agent's workflow.

主な機能

01Generates structured Hookify rule files with correct YAML frontmatter syntax

02Enables advanced condition matching using regex, substring, and prefix/suffix checks

030 GitHub stars

04Supports multiple trigger events including bash, file, stop, and prompt

05Provides templates for common safety warnings like blocking dangerous shell commands

06Facilitates project-specific 'Definition of Done' checklists for session completion

ユースケース

01Blocking debug code such as 'console.log' or sensitive credentials from being saved to files

02Preventing the execution of destructive commands like 'rm -rf' or privilege escalation

03Enforcing mandatory pre-stop checklists to ensure tests are run and documentation is updated

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add activer007/ordinary-claude-skills rule-identifier

For use in Claude.ai and ChatGPT

Download Skill