package-evaluator

Evaluate Claude Code packages across six weighted quality dimensions.

310|45|Updated Feb 22, 2026

Search Tags:#audit #metrics #evaluation #quality #packages #frontmatter

Ready to run

npx skills add https://github.com/Mathews-Tom/armory --skill package-evaluator

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: package-evaluator
Source: https://github.com/Mathews-Tom/armory/tree/main/skills/package-evaluator
Command: npx skills add https://github.com/Mathews-Tom/armory --skill package-evaluator

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Evaluates and certifies Claude Code packages for activation readiness by scoring quality across structured dimensions and surfacing actionable findings.

Core Features & Use Cases

Comprehensive, multi-dimension quality evaluation across all package types (skills, agents, hooks, rules, commands, utilities, presets) with a standardized rubric.
Supports Quick Audit (single-package) and Full Audit (all packages with ranking) to guide deployment decisions.
Generates severity-classified findings, concrete recommendations, and references the evaluation rubric for calibration.

Quick Start

Run the skill on a single package folder to generate a full audit report.

Frequently Asked Questions about package-evaluator

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I evaluate the quality of Claude Code packages for activation readiness?▼

To evaluate Claude Code packages, audit them across six weighted quality dimensions to quantify readiness, enforce a standardized rubric, and surface actionable findings. This generates a structured score report to guide deployment decisions.

Can I audit multiple packages in a repository and rank them by quality?▼

Yes, you can perform a full repository audit to evaluate all packages and generate a ranked table by quality score. This ranks all package types and workflows to guide overall deployment decisions.

What metrics are checked during a Claude Code package audit?▼

Package audits check frontmatter validity, references integrity, and apply type-specific signals. The evaluation enforces a standardized rubric across six weighted dimensions to generate robust scores.

Does the package evaluation rubric support different package types like skills and hooks?▼

Yes, the evaluation rubric supports all package types including skills, agents, hooks, rules, commands, utilities, and presets. It applies type-specific signals to ensure robust scoring across different workflows.

How do I generate a quick audit report for a single Claude Code package?▼

Run the audit on a single package folder to generate a quick audit report. This evaluates the package across six dimensions, checks frontmatter and references, and outputs severity-classified findings with concrete recommendations.