skill-creator

Automates creation, evaluation, and refinement of Claude Skills.

Updated Mar 24, 2026
One-click install
npx skills add https://github.com/france0414/icb-v2 --skill skill-creator-france0414
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/france0414/icb-v2/tree/main/.agents/skills/skill-creator
Command: npx skills add https://github.com/france0414/icb-v2 --skill skill-creator-france0414

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This skill helps teams design, implement, and continuously improve Claude Skills by providing a repeatable process for crafting frontmatter, drafting SKILL.md bodies, and running iterative evaluations and benchmarks.

Core Features & Use Cases

  • Structured frontmatter validation and metadata extraction to improve discovery and triggering reliability.
  • Guided SKILL.md drafting with progressive disclosure patterns and evaluation hooks.
  • End-to-end improve loop: run evaluations, analyze results, and refine the skill description and prompts.
  • Packaging and dependency management for distribution (scripts/, references/, assets/).

Quick Start

Draft a basic skill, run an initial eval cycle, and iterate until results improve.

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 frontmatter validation?

To automate Claude skill creation, this skill validates frontmatter metadata and drafts the SKILL.md body, ensuring triggering accuracy and reliable discovery before deployment.

What's the best way to evaluate and refine Claude skills iteratively?

The best way to refine Claude skills is running iterative evaluation loops that analyze benchmark results, update descriptions, and optimize prompts until the skill's triggering accuracy improves.

How does guided SKILL.md drafting work with progressive disclosure patterns?

Guided SKILL.md drafting uses progressive disclosure patterns and evaluation hooks to structure the skill body, balancing comprehensive instructions with clean context management for the model.

Do I need pyyaml and anthropic dependencies to package Claude skills?

Yes, you need the anthropic and pyyaml dependencies to manage packaging and handle the YAML frontmatter validation required for distributing Claude skills with their associated scripts and references.

Can I include scripts and references for deterministic tasks in my skill package?

Yes, you can include scripts, references, and assets in the skill package to support deterministic tasks, provide documentation, and manage dependencies for end-to-end deployment preparation.

When should I not use an automated evaluation loop for skill refinement?

You should avoid automated evaluation loops when a skill lacks deterministic outputs to benchmark, as the improve loop requires analyzable results to effectively refine descriptions and prompts.