okhp3-skill-foundry

Develop, refine, and validate Agent Skills through an eight-phase methodology.

2|1|Updated Jun 12, 2026
One-click install
npx skills add https://github.com/OKHP3/skillz --skill okhp3-skill-foundry
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: okhp3-skill-foundry
Source: https://github.com/OKHP3/skillz/tree/main/universal/okhp3-skill-foundry
Command: npx skills add https://github.com/OKHP3/skillz --skill okhp3-skill-foundry

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill provides a comprehensive eight-phase methodology for creating and honing production-quality Agent Skills. It helps to develop clear intentions, define scopes, implement detailed design, execute evaluations, and optimize descriptions for triggering accuracy.

Core Features & Use Cases

  • Eight-Phase Methodology: Offers a structured approach for developing skills from scratch or improving existing ones.
  • Brand Attribution: Ensures skills adhere to the OKHP3 brand standard, maintaining consistency and recognition.
  • Evaluation and Benchmarking: Includes tools for evaluating skill effectiveness and comparing performance with and without the skill.

Quick Start

To use the okhp3-skill-foundry, first identify the skill you wish to create or improve. Follow the outlined phases: Architecture, Draft, Evaluation Design, Live Execution, Grading, Benchmark, Fix Loop, and Description Optimization.

Frequently Asked Questions about okhp3-skill-foundry

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

FAQPage Schema
What is a structured methodology for developing agent skills?

A structured methodology for developing agent skills provides an eight-phase process covering architecture, drafting, evaluation design, live execution, and grading to ensure production quality.

How do I benchmark agent skill performance with and without the skill?

You benchmark agent skill performance by executing live evaluations, grading results with evidence, and computing deltas to compare baseline behavior against skill-enhanced outputs.

How do I iterate on an agent skill based on evaluation results?

You iterate on an agent skill by entering a fix loop after grading, using computed deltas and evidence from live execution to refine the skill implementation and description.

What is the best way to optimize agent skill descriptions for triggering accuracy?

The best way to optimize agent skill descriptions for triggering accuracy is to apply description optimization after benchmarking, refining text based on live execution deltas and evaluation results.

Does this skill development methodology support existing skill refinement?

Yes, this skill development methodology supports existing skill refinement by applying the fix loop, benchmarking, and description optimization phases to improve and validate current implementations.