skills.rest
barikata1984
barikata1984COMMUNITY
Explore all skills by this author →
GitHub

skill-creator

Draft SKILL.md files, generate test prompts, and run with-skill versus baseline evaluations.

Updated Aug 27, 2026
Search Tags:#benchmarking#packaging#scripts#skill-creation#viewer#evals#description-optimization
One-click install
npx skills add https://github.com/barikata1984/.claude --skill skill-creator-barikata1984
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/barikata1984/.claude/tree/main/skills/skill-creator
Command: npx skills add https://github.com/barikata1984/.claude --skill skill-creator-barikata1984

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

Reduces the friction of authoring, testing, and iterating Claude skills by providing a repeatable workflow for drafting SKILL.md, generating evals, running with-skill and baseline comparisons, grading outputs, and optimizing the skill description for reliable triggering.

Core Features & Use Cases

  • Drafting & Authoring: Guided structure and examples for writing SKILL.md frontmatter and operational instructions.
  • Eval & Benchmarking: Create test prompts, run parallel with-skill and baseline runs, capture timing/tokens, grade outputs, and aggregate benchmark statistics.
  • Iterate & Improve: Automate an improve loop that drafts assertions, analyzes results, updates the skill, and optimizes the skill description for better trigger accuracy.
  • Utilities: Includes scripts for aggregating benchmarks, generating a review viewer, packaging skills, and description-optimization helpers.

Quick Start

Tell the assistant the task you want automated and ask it to draft SKILL.md, propose 2–3 realistic test prompts, run an initial evaluation comparing with-skill and without-skill, and summarize the results and next improvements.

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 with reproducible benchmarks?▼

To create and evaluate Claude skills, you draft a SKILL.md file, generate test prompts, and run paired with-skill and baseline evaluations to capture timing, tokens, and grading statistics for reproducible benchmarking.

What is the best way to optimize a skill description for reliable triggering?▼

Optimizing a skill description involves running an automated improve loop that analyzes evaluation results, updates the skill instructions, and applies description-optimization helpers to refine trigger accuracy for reliable Claude skill activation.

How do I run baseline comparisons when testing Claude skills?▼

Running baseline comparisons requires generating test prompts and executing parallel with-skill and without-skill runs to capture performance metrics, grade outputs, and aggregate benchmark statistics for direct comparison.

Do I need Python and PyYAML to package and benchmark skills?▼

Yes, packaging and benchmarking skills requires a Python environment with anthropic and pyyaml dependencies to support bundled scripts for aggregation, grading, viewer generation, and automated description optimization workflows.

Can I automate the iteration loop for drafting and refining SKILL.md files?▼

Yes, you can automate the iteration loop by drafting assertions, analyzing evaluation results, updating SKILL.md instructions, and optimizing the skill description automatically to continuously improve Claude skill performance.

What format is required for the SKILL.md frontmatter to start authoring a skill?▼

Authoring a skill requires a SKILL.md file containing frontmatter with at minimum the name and description fields, which enables the evaluation, benchmarking, and description optimization workflows to function correctly.

Related Skills

Explore More Software Engineering Skills→
emliunixemliunix

skill-creator

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

Community
Intermediate
koukokouko
6

skill-creator-advance

Create and iteratively improve AI skills with eval-driven development.

Community
Advanced
erichowenserichowens
181

skill-coach

Guide creation, auditing, and improvement of Claude Skills with anti-patterns and progressive disclosure.

Community
Advanced
RHEcosystemAppEngRHEcosystemAppEng
10

run-evals

Automate execution and grading of AI skill evaluations for CI pipelines.

Official
Advanced