skills.rest
starrank-soft
starrank-softOFFICIAL
Explore all skills by this author →
GitHub

agent-self-evaluation

Evaluate AI agent output quality using a structured 5-axis rubric.

Updated Jun 24, 2026
Search Tags:#quality assurance#code review#debugging#productivity#evaluation#self-reflection
One-click install
npx skills add https://github.com/starrank-soft/PixelArraySkill --skill agent-self-evaluation
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-self-evaluation
Source: https://github.com/starrank-soft/PixelArraySkill/tree/main/skills/agent-self-evaluation
Command: npx skills add https://github.com/starrank-soft/PixelArraySkill --skill agent-self-evaluation

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This skill addresses the common issue of AI overconfidence and output inconsistency by forcing a structured, evidence-based reflection after complex tasks.

Core Features & Use Cases

  • 5-Axis Scoring: Evaluates output across Accuracy, Completeness, Clarity, Actionability, and Conciseness.
  • Evidence-Based Improvement: Requires the agent to cite specific proof for every score, ensuring improvements are grounded in reality.
  • Use Case: After writing a complex multi-file refactor, use this skill to catch missed edge cases or unverified assumptions before presenting the final code to the user.

Quick Start

Run the evaluate script by providing your task description and the generated output to receive a structured quality scorecard.

Frequently Asked Questions about agent-self-evaluation

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

FAQPage Schema
How do I evaluate AI code generation quality automatically?▼

To evaluate AI code generation quality, provide your task description and generated output to a heuristic-based script that generates a standardized scorecard, identifying accuracy and completeness gaps.

What is the best way to review multi-file refactoring for unverified assumptions?▼

Reviewing multi-file refactoring for unverified assumptions requires a structured 5-axis rubric that forces evidence-based reflection to catch missed edge cases before presenting the final code.

How does a 5-axis rubric evaluate technical writing accuracy?▼

A 5-axis rubric evaluates technical writing accuracy by scoring output across Accuracy, Completeness, Clarity, Actionability, and Conciseness, requiring specific proof citations for every score.

Do I need Python 3 to run heuristic-based output analysis?▼

Yes, you need Python 3 installed to execute the heuristic-based analysis script that processes your complex coding or architectural design tasks to generate a quality scorecard.

Can I use self-reflection evaluation for architectural design tasks?▼

Yes, you can apply self-reflection evaluation to architectural design tasks requiring high-fidelity verification, addressing AI overconfidence by forcing a structured, evidence-based reflection after complex tasks.

Related Skills

Explore More Software Engineering Skills→
wyattjohwyattjoh
1

task-planner

Decompose feature designs into validated JSON task graphs and Markdown plan files.

Community
Advanced
jsamuelsen11jsamuelsen11

writing-quality

Apply structured writing standards to technical documentation for clarity and accuracy.

Community
Intermediate
opmauopmau
6

review

Review staged git diffs against CLAUDE.md rules before commit.

Community
Intermediate
hb9397hb9397

multi-review

Analyze code changes across security, performance, maintainability, and testing perspectives.

Community
Advanced