How is the completeness score calculated?

Each of the eight categories is scored on a scale of 0 to 3 based on specific criteria, resulting in a total raw score out of 24 which is converted to a percentage.

Does this skill follow official Anthropic patterns?

Yes, the entire evaluation checklist and scoring rubric are derived directly from patterns observed in Anthropic's official skills repository.

What categories does the completeness audit cover?

The audit evaluates skills across eight specific categories: Preparation, Progression, Verification, Scripts, Examples, Anti-Patterns, References, and Assets.

Where are the audit reports saved?

Reports are automatically generated as Markdown files within the .claude/audits/ directory of your local project.

Audit Skill Completeness

Name: Audit Skill Completeness
Author: Jamie-BitFlight

byJamie-BitFlight

•

Developer Tools

Evaluates Claude Code skills against Anthropic's official quality standards to ensure marketplace readiness and high-quality implementation.

This skill provides a comprehensive auditing framework for AI-enhanced workflows, measuring individual skills against eight critical quality categories derived from Anthropic's official repository. It systematically evaluates preparation, progression, verification, scripts, examples, anti-patterns, references, and assets to identify gaps in functionality or documentation. By generating detailed markdown reports with actionable recommendations and a standardized completeness score, it helps developers benchmark their work against industry best practices and prepare skills for public distribution.

Key Features

01Validates skill directory structure, frontmatter, and required assets

02Calculates a standardized completeness percentage and scoring rubric

03Generates detailed Markdown reports with file-specific evidence

0426 GitHub stars

05Provides actionable recommendations and prioritized improvement tasks

06Automated 8-category quality assessment based on Anthropic standards

Use Cases

01Pre-marketplace publication review to verify quality standards

02Benchmarking local skills against official Anthropic implementation patterns

03Identifying documentation and automation gaps in newly developed skills

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jamie-bitflight/claude_skills audit-skill-completeness

For use in Claude.ai and ChatGPT

Download Skill