skill-creator

Guides teams through SKILL.md drafting and iterative eval cycles for Claude skills.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/basabecode/SomosTecnicos --skill skill-creator-basabecode
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/basabecode/SomosTecnicos/tree/main/.agents/skills/skill-creator
Command: npx skills add https://github.com/basabecode/SomosTecnicos --skill skill-creator-basabecode

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires pyyaml.

What problem does it solve?

Drafting and refining high-quality Claude skills is time-consuming and error-prone. The Skill Creator guides end-to-end development—from initial SKILL.md drafting to iterative evals—so teams can produce robust skills with clear triggering behavior.

Core Features & Use Cases

  • Guided SKILL.md authoring with frontmatter validation and structured instructions.
  • Iterative evaluation loop: generate prompts, run evals, review results, and improve descriptions.
  • Supports optional bundles (scripts/references/assets) for on-demand tooling and documentation.

Quick Start

Draft your initial SKILL.md with a clear name and description, then run the evaluation loop to iteratively refine triggering.

Frequently Asked Questions about skill-creator

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

FAQPage Schema
How do I write and structure a SKILL.md file for Claude skills?

To write a SKILL.md file for Claude skills, you draft the frontmatter and structured instructions, then validate the formatting to ensure clear triggering behavior before deployment.

What is the best way to test and evaluate Claude skill triggering behavior?

The best way to test Claude skill triggering is to run an iterative evaluation loop: generate test prompts, execute evals, review results, and refine the skill descriptions to improve accuracy.

Can I include external scripts and references when I create Claude skills?

Yes, you can include optional bundles like scripts, references, and assets when you create Claude skills to support on-demand tooling and documentation within the skill directory.

Do I need pyyaml to run iterative eval cycles for skill creation?

Yes, you need pyyaml installed to run iterative eval cycles for skill creation, as the evaluation tooling and frontmatter parsing depend on this library.

How do I improve Claude skill descriptions that fail to trigger correctly in production?

To improve Claude skill descriptions that fail to trigger correctly, use the evaluation loop to review test results and iteratively refine the frontmatter description text.

What are the limitations of using structured interviews for skill authoring?

Structured interviews for skill authoring depend on a well-scaffolded skill directory to support discovery and deployment, requiring manual coordination of frontmatter and test prompts.