Which event types can trigger a rule?

Rules can be triggered by five event types: 'bash' for command-line tools, 'file' for edits and writes, 'stop' for session completion, 'prompt' for user input, or 'all' for every event.

Can I block Claude from performing an action?

Yes, by setting the 'action' field to 'block' in the rule's frontmatter, you can prevent the operation (like a dangerous bash command) from executing.

Where should Hookify rule files be stored?

Rules must be stored in the `.claude/` directory of your project root using the naming convention `.claude/hookify.{rule-name}.local.md`.

What are Hookify rules in Claude Code?

Hookify rules are markdown files with YAML frontmatter that define patterns Claude should monitor and specific messages it should display when those patterns are matched during tasks.

How do I test a new regex pattern for a rule?

You can test patterns using a simple Python command: `python3 -c "import re; print(re.search(r'your_pattern', 'test text'))"` or by using online regex testers with Python flavor support.

Hookify Rule Architect

Name: Hookify Rule Architect
Author: akiojin

byakiojin

•

开发者工具

Simplifies the creation of automated safety checks and workflow enforcement rules for Claude Code environments.

The Hookify Rule Architect skill provides comprehensive guidance for authoring and managing Hookify rules—specialized Markdown files that monitor and govern Claude's behavior during development. By leveraging regex patterns and YAML frontmatter, this skill enables users to configure Claude to detect dangerous bash commands, prevent the commit of debug logs, and enforce project-specific best practices. Whether you need to block sensitive file edits or create completion checklists, this skill ensures your AI-driven development remains safe, consistent, and aligned with your team's standards.

主要功能

01Advanced multi-condition logic for granular control over project-specific rules.

02Real-time file monitoring to prevent shipping debug code or sensitive credentials.

03Automated detection of dangerous bash commands like 'rm -rf' or privilege escalation.

04Customizable event triggers for tool usage, session completion, and user prompts.

052 GitHub stars

06Dynamic rule loading for immediate testing and iteration without environment restarts.

使用场景

01Enforcing pre-deployment checklists when Claude attempts to conclude a task.

02Warning developers when sensitive configuration files or environment variables are modified.

03Preventing the accidental inclusion of 'console.log' or API keys in production-bound code.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add akiojin/playfab-mcp-server writing-rules

For use in Claude.ai and ChatGPT

Download Skill