What are the pass/fail criteria for assessment?

An assessment is considered 'Excellent' at 90-100 points and 'Failing' if the score is below 60 or if critical structural issues are present.

How is the compliance score calculated?

The skill uses a 100-point system where issues are deducted based on severity: critical (-20), high (-10), medium (-5), low (-2), and warning (-1). A score of 90+ is considered production-ready.

What does this skill actually validate?

It validates plugin structures (plugin.json, hooks.json), hook protocol compliance (stdin/stdout patterns), routing coverage, and adherence to Anthropic's engineering patterns for progressive disclosure.

Does this skill require manual script execution?

While it can be used for manual review, it is designed to programmatically execute Python scripts within your repository to verify implementation logic against defined standards.

Anthropic Compliance & Standards Assessor

Name: Anthropic Compliance & Standards Assessor
Author: jrc1883

byjrc1883

0•

Security & Testing

Validates Claude Code plugins against official Anthropic engineering standards and hook protocols using automated validation scripts.

This skill provides a rigorous, automated framework for ensuring that plugins and workflows adhere to Claude Code's architectural patterns and Anthropic's engineering best practices. By executing a suite of Python-based validation scripts and checking against machine-readable JSON schemas, it evaluates plugin structures, hook protocols, and agent routing coverage to provide a consistent compliance score. It is an essential tool for developers building production-ready Claude Code extensions who require objective, data-driven feedback on implementation quality and integration reliability.

Key Features

01Evaluates agent routing coverage and progressive disclosure efficiency

02Machine-readable JSON schema compliance checks for standards alignment

03Comprehensive assessment reports with prioritized fix recommendations

04Standardized 100-point scoring system with severity-based deductions

05Automated validation scripts for plugin structure and hook protocols

060 GitHub stars

Use Cases

01Ensuring a new Claude Code plugin meets production standards before deployment

02Auditing agent routing logic to prevent keyword coverage gaps or conflicts

03Debugging hook protocol errors or JSON stdin/stdout inconsistencies

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jrc1883/popkit-claude pop-assessment-anthropic

For use in Claude.ai and ChatGPT

Download Skill