What is a Claude Code Skill?

A Claude Code Skill is a specialized capability that provides domain-specific guidance, patterns, and best practices to enhance Claude's behavior during specific coding tasks.

What is the 500-line rule in Claude Code development?

It is an Anthropic best practice recommending that main SKILL.md files stay under 500 lines to maintain Claude's performance, using reference files for more detailed documentation.

Can I prevent Claude from nagging me with the same skill suggestions?

Yes, the skill-developer guide covers session tracking and skip conditions (like file markers or environment variables) to ensure guidance is timely but not repetitive.

How does the hook architecture work?

It uses a two-hook system: UserPromptSubmit for proactive suggestions before a prompt is processed, and PreToolUse for enforcing critical guardrails before a tool (like Edit) is executed.

What are the different enforcement levels for skills?

Skills can be set to BLOCK (preventing action until a condition is met), SUGGEST (injecting helpful context), or WARN (providing low-priority advisory information).

Claude Code Skill Developer

Name: Claude Code Skill Developer
Author: claudiodearaujo

byclaudiodearaujo

•

开发者工具

Guides developers through creating, managing, and debugging specialized skills for Claude Code using Anthropic's official best practices.

This skill provides a comprehensive framework for building custom Claude Code capabilities, enabling developers to implement domain-specific guidance and safety guardrails within their terminal environment. It covers the complete skill lifecycle, from configuring the two-hook architecture (UserPromptSubmit and PreToolUse) to managing complex rule sets in skill-rules.json. By adhering to the 500-line rule and progressive disclosure patterns, it ensures that Claude remains efficient and context-aware, helping teams enforce coding standards, prevent runtime errors, and streamline project-specific workflows.

主要功能

01Standardized skill creation following Anthropic's 500-line rule

021 GitHub stars

03Hook architecture management for proactive and reactive reminders

04Enforcement level implementation including Block, Suggest, and Warn modes

05Debugging tools for skill activation and session-aware state management

06Configuration guidance for skill-rules.json with keyword and intent triggers

使用场景

01Establishing organizational coding standards and safety guardrails

02Implementing progressive disclosure for large documentation sets to maintain context window efficiency

03Automating skill activation based on specific file paths or regex patterns

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add claudiodearaujo/sistema-de-narra-o-de-livro skill-developer

For use in Claude.ai and ChatGPT

Download Skill