skills.rest
NodeJSmith
NodeJSmithCOMMUNITY
Explore all skills by this author →
GitHub

mine.skill-eval

Automate evaluation and comparison of AI skill variants via structured grading and statistical analysis.

1|Updated Feb 22, 2026
Search Tags:#automation#testing#quality assurance#comparison#agentic workflow#skill evaluation
One-click install
npx skills add https://github.com/NodeJSmith/Claudefiles --skill mine-skill-eval
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: mine.skill-eval
Source: https://github.com/NodeJSmith/Claudefiles/tree/main/skills/mine.skill-eval
Command: npx skills add https://github.com/NodeJSmith/Claudefiles --skill mine-skill-eval

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill automates the complex and time-consuming process of evaluating the quality and consistency of other AI skills, providing objective data for improvement.

Core Features & Use Cases

  • Skill Variant Comparison: Directly compare two versions of a skill side-by-side to identify which performs better.
  • Automated Grading: Runs skills against predefined test cases and grades their outputs based on specific assertions.
  • Use Case: You've just refactored the mine.audit skill. Use mine.skill-eval to run both the old and new versions against a set of codebases, then compare the quality of their audit reports to ensure the changes were beneficial.

Quick Start

Use the mine.skill-eval skill to evaluate the mine.audit skill using the current project.

Frequently Asked Questions about mine.skill-eval

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

FAQPage Schema
How do I automate testing and comparison of AI skill variants?▼

Automate AI skill variant comparison by running structured grading, blind comparisons, and statistical analysis against defined assertions and user-provided prompts to generate objective performance reports.

How does automated grading work for evaluating skill quality?▼

Automated grading works by executing skills against predefined test cases and using parallel agentic sub-processes to evaluate their outputs based on specific, user-defined assertions.

Can I run a side-by-side comparison of two versions of a skill?▼

Yes, you can run a side-by-side comparison of two skill versions by executing both against a set of prompts and using blind comparison to identify which variant performs better.

What's the best way to evaluate if a refactored skill performs better than the original?▼

Evaluate refactored skills by running both the old and new versions against a shared set of inputs, then compare the statistical analysis and grading results to ensure changes were beneficial.

Do I need specific commands to run agentic workflow skill evaluations?▼

Yes, the evaluation process requires internal execution of `skill-eval-run` and `skill-eval-aggregate` commands to facilitate setup, execution, grading, and reporting.

Does skill evaluation support parallel processing for grading and comparison?▼

Yes, skill evaluation uses parallel agentic sub-processes specifically for grading and comparison tasks to automate the statistical analysis and reporting of skill performance.

Related Skills

Explore More Software Engineering Skills→
nibojniboj
1

skill-construire-evaluations

Generates evaluation cases and an evals.json for a target skill.

Community
Intermediate
terrylicaterrylica
62

multi-agent-performance-profiling

Profile data pipeline stages to identify performance bottlenecks and prioritize optimizations.

Community
Advanced
flutterflutter
2.8k

run-evals

Orchestrate automated evaluation of agent skills using isolated subagents and rubrics.

Official
Advanced
tech-with-sethtech-with-seth
4

create-unit-test

Create Vitest unit tests co-located with source files using project templates.

Official
Intermediate