What is the 500-line rule in Claude Code?

It is a best practice to keep SKILL.md files under 500 lines to ensure Claude can process information efficiently, using reference files for deeper details.

What is the difference between Block and Suggest enforcement?

Block physically prevents tool execution (like editing a file) until criteria are met, while Suggest injects a reminder into the prompt without stopping the workflow.

How does session tracking prevent notification fatigue?

It uses a state file to record when a skill has been triggered in a specific session, allowing the system to skip the reminder on subsequent actions in that same session.

What are the primary hook types in this framework?

The framework uses UserPromptSubmit for proactive suggestions before the user prompt is processed and PreToolUse for enforcing guardrails during tool execution.

Skill Developer Framework

Name: Skill Developer Framework
Author: Microck

byMicrock

•

開発者ツール

Orchestrates the creation, management, and enforcement of custom Claude Code skills following Anthropic's best practices.

概要

The Skill Developer framework provides a comprehensive toolkit for building specialized capabilities within Claude Code. It implements a sophisticated two-hook architecture—using UserPromptSubmit for proactive suggestions and PreToolUse for critical guardrails—allowing developers to define domain-specific rules, trigger patterns, and enforcement levels. By adhering to the 500-line rule and progressive disclosure patterns, it ensures that Claude remains efficient while providing deep, context-aware guidance for complex coding tasks.

主な機能

81 GitHub stars
Session-aware state management to prevent repetitive notifications
Built-in testing utilities for validating skill activation and hook logic
Configurable enforcement levels (Block, Suggest, Warn) for workflow control
Two-hook architecture for proactive suggestions and guardrail enforcement
Support for multiple trigger types including keywords, intent patterns, and file paths

ユースケース

Creating custom guardrails to prevent common database or security errors
Managing complex documentation via progressive disclosure and reference files
Implementing domain-specific style guides and architectural patterns for teams

概要

主な機能

81 GitHub stars
Session-aware state management to prevent repetitive notifications
Built-in testing utilities for validating skill activation and hook logic
Configurable enforcement levels (Block, Suggest, Warn) for workflow control
Two-hook architecture for proactive suggestions and guardrail enforcement
Support for multiple trigger types including keywords, intent patterns, and file paths

ユースケース

Creating custom guardrails to prevent common database or security errors
Managing complex documentation via progressive disclosure and reference files
Implementing domain-specific style guides and architectural patterns for teams