skills.rest
MagicBowen
MagicBowenCOMMUNITY
Explore all skills by this author →
GitHub

skill-creator

Create, test, and refine Claude skills from SKILL.md drafting through evaluation.

4|Updated Nov 14, 2025
Search Tags:#automation#workflow#evaluation#iteration#benchmark#skill-creation#project-management
One-click install
npx skills add https://github.com/MagicBowen/dogent --skill skill-creator-magicbowen
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/MagicBowen/dogent/tree/main/claude/skills/skills/skill-creator
Command: npx skills add https://github.com/MagicBowen/dogent --skill skill-creator-magicbowen

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

Streamlines the creation and evolution of Claude skills by guiding you from initial SKILL.md drafting through evaluation, benchmarking, and iterative improvements to ensure reliable triggering and performance.

Core Features & Use Cases

  • Structured SKILL.md templates with mandatory frontmatter (name and description) and a detailed body for operational instructions.
  • End-to-end evaluation loop: run eval prompts, collect results, and refine descriptions and test prompts based on outcomes.
  • Packaging, benchmarking, and history tracking: validate, package into a .skill, and compare iterations with benchmark reports.
  • Use case: a team creates a new skill for a product workflow, iterates on the triggering description, and measures improvements over time.

Quick Start

Use the skill-creator to initialize a new skill directory, populate SKILL.md, and run the evaluation/improvement loop to iterate toward a well-triggered skill.

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 refine Claude skills with evaluation loops?▼

To create and refine Claude skills with evaluation loops, draft a SKILL.md file, execute test prompts, and iteratively refine descriptions based on evaluation outcomes. This systematic process ensures reliable triggering and measurable performance improvements.

What is a SKILL.md file and why is frontmatter needed for workflow automation?▼

A SKILL.md file is a structured template defining Claude skill operational instructions, where frontmatter containing the name and description is required to ensure accurate triggering and workflow automation within project management tasks.

How do I benchmark and package a Claude skill for project management?▼

You benchmark and package a Claude skill by validating the iterated SKILL.md, packaging it into a .skill format, and generating comparison benchmark reports to track performance improvements across different iterations.

Does the skill-creator workflow require any specific dependencies or setup?▼

The skill-creator workflow requires the yaml dependency to parse and manage the structured SKILL.md frontmatter. You need this environment setup to properly initialize the skill directory and execute the evaluation scripts.

Why does my Claude skill fail to trigger accurately during automation workflows?▼

A Claude skill fails to trigger accurately when its SKILL.md description lacks specificity. You fix this by running evaluation prompts, analyzing the benchmark results, and iteratively refining the description to match expected automation triggers.

Related Skills

Explore More Product & Management Skills→
emliunixemliunix

skill-creator

Extract mandatory frontmatter from SKILL.md to define a Skill Unit.

Community
Intermediate
ginmp8ginmp8

skill-benchmark

Generate standardized benchmark reports for reusable agent skill packages.

Community
Advanced
RefoundAIRefoundAI
1.2k

ai-evals

Design rubrics, test cases, and metrics for AI model evaluations.

Official
Intermediate
santychuycomsantychuycom

skill-creator

Create and modify AI skills with SKILL.md and YAML frontmatter.

Official
Advanced