self-evaluation-patterns

What problem does it solve?

This Skill provides a structured framework for AI agents to perform self-evaluation before finalizing output, ensuring accuracy, completeness, and adherence to scope, thereby reducing errors and improving overall quality.

Core Features & Use Cases

• Completion Checklist: A step-by-step verification process against acceptance criteria, scope, consistency, build status, and side effects.
• Quality Metrics: Assesses output based on accuracy, completeness, and consistency.
• Failure Detection: Identifies early triggers for potential failures like premise collapse or scope creep.
• Uncertainty Levels: Quantifies the confidence in the output, from "Certain" to "Unable to Judge".
• Role-Specific Application: Tailors checks for Orchestrator, Planner, and Worker roles.
• Use Case: A worker agent completing a coding task will use this skill to verify its code against requirements, check for unintended side effects, and ensure it builds successfully before reporting completion.

Quick Start

Use the self-evaluation-patterns skill to check if the generated code meets all acceptance criteria and has no unintended side effects.