skill-creator

Create, test, and refine Claude skills with structured SKILL.md workflows.

Updated Feb 24, 2023

Search Tags:#automation #workflow #evaluation #benchmark #skill #frontmatter #creator

Ready to run

npx skills add https://github.com/HyperLee/Leetcode_folder --skill skill-creator-hyperlee

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/HyperLee/Leetcode_folder/tree/main/leetcode_946/.github/skills/skill-creator
Command: npx skills add https://github.com/HyperLee/Leetcode_folder --skill skill-creator-hyperlee

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill streamlines the end-to-end lifecycle of Claude skills, enabling teams to draft, test, evaluate, and iteratively improve skills to improve triggering accuracy and performance.

Core Features & Use Cases

End-to-end skill lifecycle: from drafting SKILL.md to evaluating with evals, benchmarking, and refining descriptions.
Structured workflow: supports frontmatter validation, detailed SKILL.md bodies, and optional bundles (scripts, references, assets) for reproducible results.
Tooling and packaging: provides utilities for running evaluations, viewing results, and packaging skills for distribution.
Use case: a team wants to create a new skill, run a loop of evaluations, analyze results, and iterate until triggering is reliable across diverse prompts.

Quick Start

Use this skill to guide the entire lifecycle of skill creation: draft, test prompts, run evals, review results, and iterate until triggering accuracy improves.

Frequently Asked Questions about skill-creator

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

FAQPage Schema

How do I automate Claude skill creation and evaluation at scale?▼

You can automate Claude skill creation and evaluation by drafting SKILL.md files, running structured evals, benchmarking performance, and iterating until triggering accuracy improves across diverse workflows.

What is the best way to validate frontmatter and structure for a new skill?▼

The best way to validate skill frontmatter and structure is using an automated creation workflow that checks YAML configurations, enforces structured SKILL.md bodies, and bundles optional scripts or references for reproducible results.

How do I test and refine skill triggering accuracy for diverse prompts?▼

You test and refine skill triggering accuracy by running an evaluation loop with diverse prompts, analyzing benchmark results, and iterating on descriptions until the skill triggers reliably across different workflows.

Do I need YAML dependencies to package and review Claude skills?▼

Yes, you need YAML dependencies to validate frontmatter configurations when packaging and reviewing Claude skills, as the automation workflow relies on YAML parsing for structured skill definitions.

Can I bundle scripts and references with my SKILL.md for distribution?▼

Yes, you can bundle scripts, references, and assets alongside a structured SKILL.md body to ensure reproducible results, with tooling provided for packaging these components for distribution.

What are the limitations of automating skill lifecycle workflows?▼

The automation focuses on frontmatter validation, SKILL.md structuring, and eval workflows, meaning manual intervention may still be required for complex logic design outside of triggering accuracy and description refinement.