skill-creator

Create and refine Claude skills with SKILL.md frontmatter and eval workflows.

2|Updated Mar 18, 2026
One-click install
npx skills add https://github.com/nattergabriel/oasis-greenhouse --skill skill-creator-nattergabriel
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/nattergabriel/oasis-greenhouse/tree/main/.claude/skills/skill-creator
Command: npx skills add https://github.com/nattergabriel/oasis-greenhouse --skill skill-creator-nattergabriel

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps teams create, evolve, and benchmark Claude skills by generating structured SKILL.md files, outlining evaluation workflows, and coordinating iterative improvements.

Core Features & Use Cases

  • Draft and validate SKILL.md frontmatter (name, description) and the instructional body.
  • Automate test prompt generation, eval plan scaffolding, and workspace iteration management (iteration-1, iteration-2, etc.).
  • Enforce best practices for dependencies, components, and safety checks; generate JSON eval schemas and tracking.
  • Use cases include building a new skill from scratch, updating existing skills, running evals and benchmarking, and refining triggering descriptions.

Quick Start

Draft a complete SKILL.md (frontmatter + body) and initiate the evaluation loop to test and iterate on the 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 a Claude skill from scratch?

To create a Claude skill, you draft the SKILL.md file by defining the frontmatter metadata and instructional body, then scaffold an evaluation workspace to generate test prompts and iterate on the design.

What is SKILL.md frontmatter and why is it needed for skill development?

SKILL.md frontmatter is structured metadata that defines the skill's name and description, serving as the triggering mechanism to ensure the skill activates appropriately when invoked by Claude.

How do I set up evaluation workflows to test and refine skills?

You set up evaluation workflows by scaffolding an iteration workspace, generating automated test prompts, organizing eval sets, and running benchmark tests to track improvements across multiple iterations.

Can I update an existing skill and run benchmarks on it?

Yes, you can update existing skills by modifying the SKILL.md content, organizing new eval sets, and running the evaluation loop to benchmark the refined triggering descriptions and instructional body.

What security checks are enforced during skill generation?

During skill generation, the system flags potential security risks, enforces structure for dependencies and components, and validates optional assets like scripts and references to ensure safe skill deployment.

Does skill creation work without external dependencies?

Skill creation operates entirely without external dependencies, using internal scaffolding to draft content, organize eval sets, and manage workspace iteration natively.