What triggers can I use for Hookify rules?

Rules can trigger on bash commands (bash), file edits (file), when the agent attempts to finish a task (stop), or upon user input (prompt).

Do I need to restart Claude for rule changes to take effect?

No, Hookify rules are read dynamically. Changes to your rule files take effect immediately upon the next tool use or event trigger.

Where are Hookify rules stored in a project?

Hookify rules are stored in the '.claude/' directory as markdown files using the naming convention 'hookify.{rule-name}.local.md'.

Can I prevent Claude from running specific bash commands?

Yes, by setting the event to 'bash' and utilizing the 'block' action in the YAML frontmatter, you can stop dangerous commands from executing.

Writing Hookify Rules

Name: Writing Hookify Rules
Author: greatxrider

bygreatxrider

0•

Productividad y Flujo de Trabajo

Standardizes and automates the creation of custom Hookify rules to monitor agent behavior and enforce development best practices.

This skill provides specialized guidance for creating Hookify rules, which are markdown-based configuration files used to intercept, monitor, and guide Claude's actions within a workspace. By defining specific patterns for bash commands, file edits, and session events, users can implement custom safety guardrails, prevent common coding errors, and enforce project-specific workflows. It simplifies the complex syntax of YAML frontmatter and advanced regex conditions, ensuring that rules are correctly formatted to trigger exactly when needed without disrupting the developer's productivity.

Características Principales

01Automated rule file generation with correct YAML frontmatter structure

02Regex pattern optimization to ensure accurate triggering and avoid false positives

03Advanced multi-condition logic for complex file path and content monitoring

04Standardized templates for warning messages and corrective suggestions

05Comprehensive guidance for bash, file, stop, and prompt event triggers

060 GitHub stars

Casos de Uso

01Blocking dangerous terminal commands like 'rm -rf' or unauthorized privilege escalations

02Implementing automated completion checklists that trigger when a session is about to end

03Enforcing code quality by warning against debug logs, eval statements, or insecure innerHTML

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add greatxrider/nomanuai writing-rules

For use in Claude.ai and ChatGPT