skill-creator

Draft, test, and iteratively improve AI skills with SKILL.md and eval sets.

6|2|Updated May 11, 2026
One-click install
npx skills add https://github.com/pax-k/war-gov-uap-wiki --skill skill-creator-pax-k
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/pax-k/war-gov-uap-wiki/tree/main/.skills/skill-creator
Command: npx skills add https://github.com/pax-k/war-gov-uap-wiki --skill skill-creator-pax-k

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires yaml.

What problem does it solve?

Skill creators often face a fragmented, manual process to draft SKILL.md, run evaluations, and iterate improvements. This guide provides a structured workflow to streamline skill creation, testing, and refinement with repeatable steps and tooling.

Core Features & Use Cases

  • Structured workflow for creating new skills, editing SKILL.md, and tracking iterations.
  • Built-in eval-economy loop: prompts, holdout sets, evaluation results, and automated improvements.
  • Optional bundling of resources (scripts/references/assets) to scale skill automation and reuse.

Quick Start

Create a new skill, run the evaluation loop, and iterate on the description until triggering is optimized.

Frequently Asked Questions about skill-creator

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

FAQPage Schema
How do I create and iterate on AI skills with a structured workflow?

To create and iterate on AI skills, use a structured workflow to draft SKILL.md, run prompt evaluations against holdout sets, and apply improvement loops to optimize triggering and validate outcomes.

What is the best way to manage SKILL.md and evaluation sets during development?

The best way to manage SKILL.md and evaluation sets is to follow a repeatable workflow that tracks iterations, runs eval loops on prompts, and measures results to refine skill descriptions.

Do I need YAML to build and test skills?

You need YAML to build and test skills because it is a required dependency for managing the structured configuration files and metadata that drive the evaluation and iteration loops.

Can I bundle external resources like scripts and references into a skill?

You can bundle optional resources like scripts, references, and assets into a skill to scale automation and reuse, integrating them directly into the structured workflow and evaluation loops.

Why does my skill triggering fail during the evaluation loop?

Skill triggering fails during the evaluation loop when the SKILL.md description is not optimized, requiring you to iterate on prompts and validate outcomes against holdout sets until triggering improves.