How does the UserPromptSubmit hook function?

This hook proactively injects relevant skill guidance into the context before Claude sees the user's prompt, based on detected keywords and intent patterns.

How do I test a new skill trigger?

You can manually test triggers using the provided npx tsx commands in the terminal to simulate prompt submission or tool usage without running a full session.

What is the 500-line rule for Claude skills?

It is an Anthropic best practice that ensures SKILL.md files remain under 500 lines to maintain Claude's context efficiency, utilizing reference files for more granular documentation.

What are the different types of skill triggers?

Triggers include explicit keywords, implicit intent patterns (regex), specific file path globs, and technology-specific content detection within files.

Can I prevent Claude from making specific code edits?

Yes, by setting a skill's enforcement level to BLOCK, you can physically prevent tool execution until the developer addresses a specific guardrail requirement.

Skill Developer Guide

Name: Skill Developer Guide
Author: Microck

byMicrock

•

学习与文档

Streamlines the creation and management of custom Claude Code skills by enforcing Anthropic's best practices and automated activation patterns.

关于

The Skill Developer guide provides a comprehensive framework for building domain-specific capabilities within Claude Code, ensuring consistency through standardized YAML frontmatter, multi-hook architectures, and sophisticated trigger systems. It empowers developers to implement proactive suggestions via the UserPromptSubmit hook or critical guardrails through the PreToolUse hook, while strictly adhering to the 500-line rule and progressive disclosure patterns to maintain AI context efficiency and code quality.

主要功能

Two-hook architecture supporting proactive suggestions and post-response reminders
89 GitHub stars
Standardized templates for skill structure and YAML frontmatter
Advanced trigger configuration using keywords, intent patterns, and file paths
Configurable enforcement levels including BLOCK, SUGGEST, and WARN
Built-in session tracking to prevent repetitive notifications within the same workflow

使用场景

Debugging and testing skill activation patterns using CLI-based evaluation tools
Organizing complex domain knowledge into modular, auto-activating documentation
Developing custom guardrails to prevent common database or security errors

关于

主要功能

Two-hook architecture supporting proactive suggestions and post-response reminders
89 GitHub stars
Standardized templates for skill structure and YAML frontmatter
Advanced trigger configuration using keywords, intent patterns, and file paths
Configurable enforcement levels including BLOCK, SUGGEST, and WARN
Built-in session tracking to prevent repetitive notifications within the same workflow

使用场景

Debugging and testing skill activation patterns using CLI-based evaluation tools
Organizing complex domain knowledge into modular, auto-activating documentation
Developing custom guardrails to prevent common database or security errors