skill-creator

Draft SKILL.md files, run evals, and iterate on Claude skills.

10|3|Updated Apr 2, 2026
One-click install
npx skills add https://github.com/raphaelfh/newsweep --skill skill-creator-raphaelfh
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/raphaelfh/newsweep/tree/main/.claude/skills/skill-creator
Command: npx skills add https://github.com/raphaelfh/newsweep --skill skill-creator-raphaelfh

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

The Skill Creator helps teams design, test, and iterate Claude skills—from initial drafting to evaluation and optimization—so that skills trigger reliably and deliver measurable value.

Core Features & Use Cases

  • End-to-end skill creation: draft SKILL.md, run evals, review results, and iterate based on feedback.
  • Evals and benchmarking: leverage built-in evaluation tooling, reviewer scripts, and optional benchmarks to quantify triggering accuracy and performance.
  • Trigger optimization: refine descriptions and prompts to improve discovery and reduce under/over-trigger incidents.

Quick Start

Use the Skill Creator to start a new skill project, write an initial SKILL.md, and begin the loop of evaluation and improvement.

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 evaluate Claude skills end to end?

To create and evaluate Claude skills, draft a SKILL.md file, run built-in evaluations, review results, and iterate based on feedback to improve triggering accuracy and overall performance.

What is the best way to optimize Claude skill triggering and reduce over-triggering?

The best way to optimize Claude skill triggering is to refine skill descriptions and prompts, then run evaluations and benchmarks to quantify triggering accuracy and reduce under or over-trigger incidents.

How does evaluation and benchmarking work for Claude skill creation?

Evaluation and benchmarking for Claude skill creation works by leveraging built-in evaluation tooling, reviewer scripts, and optional benchmarks to measure triggering accuracy and iterate on skill performance.

Can I use YAML to manage SKILL.md drafting and automation scripts?

Yes, you can use YAML to manage SKILL.md drafting, as the skill creation workflow requires YAML dependencies to support templates, scripts, and automation for measurement and evaluation.

What do I need to start a new Claude skill project from scratch?

To start a new Claude skill project, you need to set up tooling for prompts and evaluation, then use the workflow to write an initial SKILL.md and begin the loop of evaluation and improvement.