skill-creator

Draft SKILL.md content and run trigger-based evals to optimize Claude skill descriptions.

1|Updated May 12, 2026
One-click install
npx skills add https://github.com/MoxyWolfLLC/moxywolf-plugins --skill skill-creator-moxywolfllc
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/MoxyWolfLLC/moxywolf-plugins/tree/main/skill-bundles/moxywolf-skills/skills/skill-creator
Command: npx skills add https://github.com/MoxyWolfLLC/moxywolf-plugins --skill skill-creator-moxywolfllc

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill helps you create new skills or improve existing ones by turning vague ideas into clear instructions, then validating those instructions with trigger-based evals so Claude actually uses the skill when it should.

Core Features & Use Cases

  • Skill drafting and iteration: Turn a user’s goal into a working SKILL.md, then refine it based on evaluated results.
  • Test-case creation and evaluation: Build a set of realistic prompts, run Claude with and without the skill, and compare outcomes.
  • Quantitative + qualitative feedback loop: Produce benchmarks, view pass-rate and variance, and use a viewer to guide improvements.
  • Description optimization for triggering: Generate a set of trigger/non-trigger eval queries and automatically improve the SKILL.md description to reduce under-triggering.
  • Suitable use cases: Use when you want to start from scratch, edit/optimize an existing skill, and run evals/benchmarks to measure performance and triggering accuracy.

Quick Start

Ask the Skill to create a new skill for your desired task and include what the skill should do, what phrases should trigger it, and what output format you want.

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 test new AI skills for Claude Code?

To create and test AI skills for Claude Code, draft SKILL.md instructions, define eval prompts, run trigger-based comparisons with and without the skill, and optimize descriptions for accurate invocation using quantitative benchmarks and qualitative review.

What is trigger optimization for AI skill instructions?

Trigger optimization for AI skill instructions involves generating trigger and non-trigger eval queries to automatically improve the SKILL.md description. This reduces under-triggering and ensures Claude accurately invokes the skill only when appropriate.

How do I benchmark Claude skill performance using evals?

Benchmarking Claude skill performance involves running Claude with access to the skill versus baselines across realistic test prompts. You compare outcomes using variance reporting, pass-rate metrics, and a viewer to guide iterative instruction improvements.

Can I refine an existing AI skill instead of starting from scratch?

Yes, you can refine an existing AI skill instead of starting from scratch. The process evaluates current trigger-based performance, generates benchmark comparisons, and iteratively optimizes the SKILL.md description and instructions to improve results.

What's the best way to measure AI skill triggering accuracy?

The best way to measure AI skill triggering accuracy is to generate specific trigger and non-trigger eval queries, run baseline comparisons, and use variance and pass-rate benchmarks to quantify how reliably the skill activates.