How does this skill help reduce subjective bias?

By requiring explicit, observable descriptors for every score level, the skill forces reviewers to base assessments on evidence rather than vague personal impressions.

Can I use this for pass/fail compliance checks?

Yes, the skill includes a Checklist (Binary) pattern specifically designed for compliance gates and minimum quality requirements.

Can this skill help with team alignment?

Yes, it includes calibration techniques to ensure different team members interpret the rubric the same way, achieving high inter-rater reliability.

What is the benefit of using an analytic rubric?

Analytic rubrics break performance down into specific dimensions, providing detailed diagnostic feedback that helps evaluatees understand exactly where they need to improve.

How many criteria should a good rubric have?

A typical effective rubric includes 4-8 criteria. This provides enough detail for actionable feedback without becoming overly complex or time-consuming to use.

Evaluation Rubrics

Name: Evaluation Rubrics
Author: lyndonkl

bylyndonkl

•

生産性とワークフロー

Designs and applies structured scoring frameworks to ensure objective, consistent, and actionable quality assessments across any domain.

The Evaluation Rubrics skill empowers Claude to create professional-grade assessment frameworks that replace subjective opinions with objective data. It guides users through the entire rubric lifecycle—from identifying measurable criteria to writing distinct performance descriptors and calibrating for inter-rater reliability. Whether you are conducting code reviews, evaluating job candidates, or grading technical documentation, this skill provides the templates and logic needed to define high-quality standards and deliver transparent, developmental feedback that reduces bias and aligns teams.

主な機能

01Support for Analytic, Holistic, Single-Point, and Checklist rubric patterns

028 GitHub stars

03Inter-rater reliability guidance to ensure scoring consistency across multiple reviewers

04Automated generation of performance descriptors for multi-level scales

05Structured methodology for defining observable and measurable quality criteria

06Weighting systems to prioritize critical dimensions like security or performance

ユースケース

01Providing structured, actionable feedback for content creation and documentation

02Standardizing code reviews and technical design critiques for engineering teams

03Objectively evaluating vendor proposals, grant applications, or hiring candidates

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add lyndonkl/claude evaluation-rubrics

For use in Claude.ai and ChatGPT

Download Skill