skill-creator

Create, modify, and evaluate Claude Skills with SKILL.md frontmatter.

34|Updated Oct 30, 2023
One-click install
npx skills add https://github.com/daichan132/Youtube-Live-Chat-Fullscreen --skill skill-creator-daichan132
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/daichan132/Youtube-Live-Chat-Fullscreen/tree/main/.agents/skills/skill-creator
Command: npx skills add https://github.com/daichan132/Youtube-Live-Chat-Fullscreen --skill skill-creator-daichan132

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

The Skill Creator helps teams design, implement, and refine Claude skills by guiding the drafting of SKILL.md, validating frontmatter, organizing evals, and coordinating resources to accelerate skill development.

Core Features & Use Cases

  • Capture intent: structure and capture a skill's purpose, trigger contexts, and outputs.
  • Write SKILL.md: frontmatter (name + description) plus a detailed body with progressive disclosure and links to scripts, references, and assets.
  • Plan and run evals: generate evaluation prompts, manage train/test splits, spawn with-skill and baseline runs, and coordinate benchmarking.
  • Benchmark and improve: aggregate results, compare skill variants, and guide iterative improvements.
  • Package and deploy: validate skills and package into portable .skill files including dependencies and components.

Quick Start

Run the evaluation loop to draft a skill and iterate toward improvements.

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 validate Claude Skills from scratch?

To create Claude Skills, you draft a SKILL.md file with frontmatter containing name and description, then validate the frontmatter and organize optional scripts, references, and assets for deployment.

What is the best way to evaluate and benchmark Claude Skill performance?

Evaluating Claude Skills involves generating evaluation prompts, managing train/test splits, spawning with-skill and baseline runs, and aggregating results to compare skill variants for iterative improvements.

How do I package a Claude Skill for deployment?

Packaging a Claude Skill requires validating the skill components and bundling them into a portable .skill file that includes dependencies like pyyaml alongside scripts, references, and assets.

Can I optimize skill descriptions to improve triggering accuracy?

Optimizing skill descriptions for triggering accuracy is supported by benchmarking performance across skill variants, aggregating evaluation results, and guiding iterative improvements to the SKILL.md frontmatter.

Do I need pyyaml to build and run skill evaluation workflows?

Pyyaml is required as a dependency for the skill creation and evaluation workflow, supporting frontmatter validation and structured output generation during the skill development process.

How does progressive disclosure work when structuring a SKILL.md file?

Progressive disclosure in SKILL.md involves writing a detailed body with links to scripts, references, and assets, allowing Claude to access layered context based on the task's complexity.