skill-creator

Create and refine Claude Skills with structured workflows for testing and iteration.

23|2|Updated May 27, 2026

Search Tags:#automation #markdown #cli #tooling #evaluation #loop #skill-creator

Ready to run

npx skills add https://github.com/zhouguoqing/QianYuan.AIAgenticFramework --skill skill-creator-zhouguoqing

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/zhouguoqing/QianYuan.AIAgenticFramework/tree/main/.agents/skills/skill-creator
Command: npx skills add https://github.com/zhouguoqing/QianYuan.AIAgenticFramework --skill skill-creator-zhouguoqing

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The Skill Creator addresses the complexity of designing, validating, and iterating reusable Skills by providing a structured workflow for frontmatter, body instructions, and ramped loading of resources.

Core Features & Use Cases

Progressive loading: metadata, the full SKILL.md body, and optional bundles load only when needed during activation.
End-to-end skill iteration: run evaluations, capture results, and automatically generate improvement suggestions to converge on reliable triggering behavior.
Reusable templates and standards: consistent frontmatter, eval prompts, and benchmarking outputs to accelerate multi-skill development.
Use cases: teams building modular AI agents can prototype, test, and optimize skills, then package for deployment.

Quick Start

Initialize a new skill with SKILL.md, run an evaluation loop, review results, and implement improvements until triggering is reliable.

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 reusable Claude Skills with a structured workflow?▼

To create and validate reusable Claude Skills, use a structured workflow that scaffolds frontmatter, body instructions, and resource loading, then runs evaluation loops to test and benchmark prompting behavior for reliable deployment.

What is progressive loading for skill metadata and how does it work?▼

Progressive loading for skill metadata is a mechanism where metadata, the full SKILL.md body, and optional asset bundles load only when needed during activation, optimizing resource usage and skill triggering behavior across projects.

Can I automate prompt evaluation and improvement iterations for AI agent skills?▼

Yes, you can automate prompt evaluation and improvement iterations by running an evaluation loop that captures results, benchmarks outputs, and automatically generates improvement suggestions to converge on reliable triggering behavior.

How do I scaffold frontmatter and benchmarking templates for multi-skill development?▼

Scaffold frontmatter and benchmarking templates for multi-skill development by initializing a new skill with SKILL.md, applying reusable standards, and running evaluation loops to validate consistent prompting behavior across modular AI agents.

Does this skill creation workflow require external dependencies or components?▼

No, this skill creation workflow requires no external dependencies or components, allowing teams to independently prototype, test, optimize, and package modular AI agent skills for deployment across projects.

What's the best way to package and deploy modular AI agent skills across projects?▼

The best way to package and deploy modular AI agent skills is to prototype, validate with evaluation loops, optimize triggering behavior, and use reusable templates and standards to automate skill deployment across projects.