skill-creator

Create and refine Claude Code skills with SKILL.md drafting and eval design.

Updated Mar 29, 2026
One-click install
npx skills add https://github.com/zizhenchen47-oss/claude-skills --skill skill-creator-zizhenchen47-oss
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/zizhenchen47-oss/claude-skills/tree/main/skill-creator
Command: npx skills add https://github.com/zizhenchen47-oss/claude-skills --skill skill-creator-zizhenchen47-oss

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill helps you turn an idea into a working Claude skill and improve it over time, so you can move from vague intent to a tested, triggerable, packaged skill with less guesswork.

Core Features & Use Cases

  • Capture intent and scope: Clarify what the skill should do, when it should trigger, and what output it should produce.
  • Draft and refine SKILL.md: Write a focused skill body, improve the trigger description, and keep the instructions lean and usable.
  • Create and run evals: Build realistic test prompts, compare skill outputs against baselines, and review pass rates, timing, and token usage.
  • Iterate with feedback: Use benchmark results and reviewer comments to tighten the workflow, add scripts when repeated work appears, and improve reliability.
  • Package for reuse: Validate the skill and ship it as a distributable file when it is ready.

Quick Start

Tell Claude what you want the skill to enable, then use this skill to draft the SKILL.md, create test prompts, run evaluations, and refine the description until the skill triggers and performs well.

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 Claude skills from scratch?▼

To create Claude skills, you capture intent, draft SKILL.md, generate test prompts, run evals, and refine trigger descriptions. This structured workflow moves from vague ideas to tested, triggerable, packaged skills with reliable quality control.

What is skill benchmarking and how does it evaluate prompt engineering outputs?▼

Skill benchmarking compares outputs against baselines using realistic test prompts. It reviews pass rates, timing, and token usage to evaluate prompt engineering quality, helping iterate workflows based on concrete benchmark results.

How do I write a trigger description for a SKILL.md file?▼

Writing a trigger description for SKILL.md involves clarifying what the skill does and when it triggers. You draft a focused body, keep instructions lean, and iteratively tune the description until the skill triggers reliably.

Do I need pyyaml installed to package and validate Claude skills?▼

Yes, pyyaml is required for this skill creation workflow. It supports the scripts and validation processes needed to package and ship tested Claude skills as distributable files.

What is the best way to refine an existing Claude skill workflow?▼

The best way to refine a Claude skill is using benchmark results and reviewer comments to tighten workflows. You add scripts for repeated work, improve reliability, and validate the skill before packaging.

When should I not use an automated skill creation workflow?▼

Avoid automated skill creation workflows when tasks lack clear scope or measurable outputs. Skill creation requires defining intent, generating test prompts, and running evals, which fails without concrete success criteria.