Verified:GitHub

skill-creator

Coordinate end-to-end Claude skill design, evaluation, and deployment.

2|Updated May 19, 2024

Search Tags:#automation #workflow #markdown #evaluation #iteration #skill #frontmatter

One-click install

npx skills add https://github.com/YanBerdin/next-portfolio --skill skill-creator-yanberdin

Or copy as Structured Prompt for Agent▼

Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/YanBerdin/next-portfolio/tree/main/.github/skills/skill-creator
Command: npx skills add https://github.com/YanBerdin/next-portfolio --skill skill-creator-yanberdin

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires yaml.

What problem does it solve?

The Skill Creator enables teams to design, test, and iteratively improve Claude skills by providing a structured workflow that connects drafting SKILL.md, running evaluations, and benchmarking results to drive reliable skill deployment.

Core Features & Use Cases

Orchestrates the end-to-end skill development loop: drafting SKILL.md, wiring eval prompts, running evaluations, and analyzing results.
Supports iterative improvement by feeding evaluation outcomes back into skill descriptions and suggested optimizations.
Provides benchmarking and analytics hooks to compare skill configurations and measure trigger accuracy, latency, and resource usage.
Bundles resources (scripts, references, assets) into a deployable skill package and validates readiness before packaging.
Facilitates collaboration by documenting processes, test cases, and version history for skill evolution.

Quick Start

Draft an initial SKILL.md for your skill and run the guided workflow to iterate from draft to deployable 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 evaluate Claude skills end-to-end?▼

You can evaluate Claude skills by wiring eval prompts and running iterative evaluation loops that analyze results to improve trigger accuracy, latency, and resource usage for reliable deployment.

What is the best way to iterate on SKILL.md descriptions for better trigger accuracy?▼

You need YAML to parse SKILL.md frontmatter metadata, while additional scripts, references, and assets are bundled into the deployable skill package during the validation and packaging phase.

Does this skill creation workflow support benchmarking and version history?▼

Before packaging a skill, readiness validation checks the bundled resources including scripts, references, and assets to ensure the SKILL.md configuration is complete and deployable.

Can I bundle scripts and references into a deployable skill package?▼

You can bundle scripts, references, and assets into a deployable skill package after the workflow validates readiness, ensuring the SKILL.md and associated resources are prepared for deployment.