skill-creator

Create, test, and iterate Claude skills with SKILL.md drafting and eval prompts.

31|Updated Sep 29, 2023
One-click install
npx skills add https://github.com/minima-js/minimajs --skill skill-creator-minima-js
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/minima-js/minimajs/tree/main/.agents/skills/skill-creator
Command: npx skills add https://github.com/minima-js/minimajs --skill skill-creator-minima-js

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Creating robust Claude skills is a manual, error-prone process requiring careful planning, testing, and iteration. This Skill provides a repeatable blueprint to draft SKILL.md, structure evaluation workflows, and progressively improve skills based on feedback.

Core Features & Use Cases

  • Guided SKILL.md drafting: define name, description, and metadata to enable discovery.
  • Evaluation scaffolding: generate prompts, run with-skill and baseline tests, and collect results for comparison.
  • Iteration loop: apply improvements, re-run tests, and optionally optimize the triggering description.
  • Reference tooling: integrates optional scripts, references, and assets to support large skills.

Quick Start

Create a new skill folder with SKILL.md, capture the core metadata in frontmatter, and follow the steps to draft, test, and iterate.

Frequently Asked Questions about skill-creator

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

FAQPage Schema
How do I create a SKILL.md file with the correct YAML frontmatter and markdown instructions?

To create SKILL.md files, draft the core metadata in YAML frontmatter and define markdown instructions to establish reusable AI skills. This structures your skill's name, description, and logic for proper discovery and triggering.

What is the best way to evaluate and test Claude skills for triggering accuracy?

Evaluating Claude skills requires generating test prompts to run with-skill and baseline comparisons. This evaluation scaffolding collects results to measure triggering accuracy and identify areas for iterative improvement.

How do I structure an iteration loop to refine skill descriptions and optimize performance?

Optimize skill performance by applying improvements, re-running baseline tests, and adjusting the triggering description. This iteration loop progressively refines the skill based on direct evaluation feedback.

Can I integrate external scripts and reference assets into large skills?

You can integrate optional scripts, references, and assets to support large skills. This reference tooling ensures complex skill requirements are properly defined alongside the SKILL.md markdown instructions.

What metadata is required to enable discovery for reusable AI skills?

Discovery for reusable AI skills requires defining the name, description, and core metadata in YAML frontmatter. Capturing this metadata ensures the skill is properly indexed and triggered across domains.