What is the difference between BLOCK and SUGGEST enforcement?

BLOCK prevents Claude from executing a tool (like Edit or Write) until specific conditions are met, whereas SUGGEST simply injects helpful guidance into the prompt without interrupting the workflow.

How can I prevent a skill from nagging me in every turn?

Skills use session tracking to store state in .claude/hooks/state/, allowing them to trigger once per session or until a specific file marker like @skip-validation is detected.

What is the 500-line rule for Claude skills?

The 500-line rule is a best practice recommending that any individual SKILL.md file stay under 500 lines to maximize Claude's remaining context window for the actual task at hand.

How do I trigger a skill based on specific files?

You can define glob patterns in the skill-rules.json file under the filePathTriggers section to automatically activate a skill when specific file types or directories are accessed.

Claude Skill Developer

Name: Claude Skill Developer
Author: Activer007

byActiver007

개발자 도구

Streamlines the creation, configuration, and management of custom Claude Code skills using standardized architectural patterns.

소개

The Claude Skill Developer is a specialized meta-skill designed to guide developers through the entire lifecycle of building custom capabilities for Claude Code. It provides a comprehensive framework for implementing auto-activation triggers, managing the skill-rules.json configuration, and leveraging the dual-hook architecture (UserPromptSubmit and PreToolUse). By enforcing Anthropic best practices—such as the 500-line rule and progressive disclosure—this tool ensures that custom skills are performant, context-efficient, and capable of providing either gentle suggestions or hard guardrails for critical development tasks.

주요 기능

Standardized skill scaffolding with YAML frontmatter and structured SKILL.md templates
Session-aware state management to prevent repetitive or intrusive AI suggestions
Advanced trigger management using keywords, regex intent patterns, and glob file paths
Automated testing workflows for validating skill activation and hook performance
0 GitHub stars
Dual-hook architecture implementation for proactive context injection and safety enforcement

사용 사례

Optimizing Claude's performance by refactoring large skills into progressive disclosure patterns
Building organizational guardrails to prevent common security vulnerabilities or database errors
Creating domain-specific documentation skills that auto-activate based on project file types

소개

주요 기능

Standardized skill scaffolding with YAML frontmatter and structured SKILL.md templates
Session-aware state management to prevent repetitive or intrusive AI suggestions
Advanced trigger management using keywords, regex intent patterns, and glob file paths
Automated testing workflows for validating skill activation and hook performance
0 GitHub stars
Dual-hook architecture implementation for proactive context injection and safety enforcement

사용 사례

Optimizing Claude's performance by refactoring large skills into progressive disclosure patterns
Building organizational guardrails to prevent common security vulnerabilities or database errors
Creating domain-specific documentation skills that auto-activate based on project file types