Does this skill help with agent performance?

It includes best practices for agent isolation, specialized tool allowlists, and patterns for parallel agent execution to process independent tasks simultaneously.

What is the primary purpose of this skill?

This skill provides a standardized framework for validating and optimizing Claude Code plugins, ensuring they follow core principles, correct component models, and proper directory structures.

How does it handle RFC 2119 compliance?

It enforces the mandatory use of specific terms like MUST, SHOULD, and MAY in plugin documentation to ensure instructions are clear and unambiguous for the AI.

Can it distinguish between different types of skills?

Yes, it provides specific guidance for 'Instruction-type' skills (user-invoked workflows) and 'Knowledge-type' skills (declarative references for agents).

Plugin Best Practices & Optimizer

Name: Plugin Best Practices & Optimizer
Author: FradSer

byFradSer

•

207

개발자 도구

Validates and optimizes Claude Code plugin structures to ensure compliance with architectural standards and RFC 2119 requirements.

소개

The Plugin Best Practices skill acts as a comprehensive quality assurance layer for developers building within the Claude Code ecosystem. It provides automated guidance on component modeling, distinguishing between instruction-type skills, knowledge-type skills, and autonomous agents to ensure optimal performance. By enforcing strict RFC 2119 terminology and validating directory structures, manifest declarations, and parallel execution patterns, this skill helps developers create robust, standardized, and high-performance plugins.

주요 기능

Differentiation logic for instruction-type vs. knowledge-type skills
Optimization patterns for parallel agent execution and result consolidation
Automated validation of plugin.json manifest declarations and component paths
Structural auditing for agent isolation and tool allowlist safety
207 GitHub stars
RFC 2119 compliance checking for standardized documentation terminology

사용 사례

Conducting a pre-release quality audit of a new Claude Code plugin
Optimizing complex workflows that require multiple agents running in parallel
Refactoring legacy commands into modern skill and agent architectures

소개

주요 기능

Differentiation logic for instruction-type vs. knowledge-type skills
Optimization patterns for parallel agent execution and result consolidation
Automated validation of plugin.json manifest declarations and component paths
Structural auditing for agent isolation and tool allowlist safety
207 GitHub stars
RFC 2119 compliance checking for standardized documentation terminology

사용 사례

Conducting a pre-release quality audit of a new Claude Code plugin
Optimizing complex workflows that require multiple agents running in parallel
Refactoring legacy commands into modern skill and agent architectures