ck:skill-creator

Automate Claude skill creation and refinement through eval-driven iteration.

Updated Apr 9, 2026
One-click install
npx skills add https://github.com/DatTran26/KienTruThiHanh --skill ck-skill-creator-dattran26
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ck:skill-creator
Source: https://github.com/DatTran26/KienTruThiHanh/tree/main/.agents/skills/skill-creator
Command: npx skills add https://github.com/DatTran26/KienTruThiHanh --skill ck-skill-creator-dattran26

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This skill unit enables teams to systematically create and refine Claude skills using an eval-driven loop, reducing guesswork and accelerating deployment.

Core Features & Use Cases

  • Eval-driven iteration: automate test, grade, compare, and optimize skill prompts and structures.
  • Progressive disclosure: metadata → SKILL.md → bundled resources for efficiency.
  • Packaging & benchmarking: validate structure, package skills, and generate benchmark reports.

Quick Start

Run scripts/init_skill.py ck:skill-name --path <path> to initialize a new skill directory with SKILL.md and starter resources.

Frequently Asked Questions about ck:skill-creator

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

FAQPage Schema
How do I automate Claude skill creation with an eval-driven loop?

Automate Claude skill creation by using an eval-driven loop to test, grade, compare, and optimize skill prompts iteratively. This systematically reduces guesswork and accelerates the deployment of new skills.

What is the best way to package and validate Claude skills for benchmarking?

Package and validate Claude skills through deterministic benchmarking scripts that validate directory structures, bundle resources, and generate benchmark reports to ensure skills meet design patterns.

How do I initialize a new Claude skill directory with starter resources?

Initialize a new Claude skill directory by running `scripts/init_skill.py ck:skill-name --path <path>`, which automatically generates a SKILL.md file and starter resources for progressive disclosure.

Does the skill creation workflow require the anthropic dependency?

Yes, the skill creation workflow requires the anthropic dependency to automate prompt evaluation, apply iterative refinements, and align skill structures with the Claude framework.

What is progressive disclosure in Claude skill development?

Progressive disclosure in Claude skill development structures resources sequentially from metadata to SKILL.md and bundled references, optimizing efficiency for prompt evaluation and deployment.

Can I optimize existing skill scripts and descriptions using evals?

Yes, you can optimize existing skill scripts, references, and descriptions by applying eval-driven iteration to benchmark performance, grade outputs, and tune descriptions systematically.