skills.rest
bingran-you
bingran-youCOMMUNITY
Explore all skills by this author →
Website𝕏@bingran_bryGitHub

skill-creator

Draft, evaluate, and iteratively improve Claude skills with built-in benchmarking.

3|Updated Jan 14, 2022
Search Tags:#automation#evaluation#iteration#creation#benchmark#skill
One-click install
npx skills add https://github.com/bingran-you/bingran-you --skill skill-creator-bingran-you
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/bingran-you/bingran-you/tree/main/repo-skills/skill-creator
Command: npx skills add https://github.com/bingran-you/bingran-you --skill skill-creator-bingran-you

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill provides a structured, end-to-end workflow for creating, evaluating, and iterating Claude skills. It centralizes drafting, testing, evaluating, and refining a skill into a repeatable process, reducing time-to-trigger accuracy and ensuring consistent improvement through quantifiable feedback.

Core Features & Use Cases

  • End-to-end skill workflow: draft, run prompts, evaluate results, and iteratively improve the skill.
  • Evaluation and benchmarking: integrate evals, an eval viewer, and optional benchmarking to quantify progress.
  • Descriptor optimization: coordinate with the description optimizer to improve triggering accuracy and coverage.
  • Packaging for deployment: bundle skills into portable .skill files for sharing and installation.

Quick Start

Create an initial SKILL.md, then run the loop to draft, test, and iterate your skill using the built-in evaluator and viewer.

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 systematically?▼

To create and evaluate Claude skills, you can use a structured workflow that drafts a SKILL.md, runs design prompts, and evaluates performance with a built-in evaluator. This process enables iterative improvement through quantifiable feedback before deployment.

What is the best way to benchmark changes to a Claude skill?▼

The best way to benchmark changes to a Claude skill is by integrating the built-in evaluator and eval viewer. This allows you to quantify progress and measure triggering accuracy across iterative refinement cycles.

How do I package Claude skills for sharing and installation?▼

You package Claude skills for sharing and installation by bundling them into portable .skill files. This packaging workflow integrates directly with the refinement loop to deliver ready-to-deploy skills.

Do I need pyyaml to automate skill creation and iteration?▼

Yes, you need pyyaml installed to automate skill creation and iteration. This dependency supports the scripts that orchestrate the end-to-end drafting, evaluation, and refinement workflow.

How does descriptor optimization improve triggering accuracy for Claude skills?▼

Descriptor optimization improves triggering accuracy by coordinating with a description optimizer to refine skill metadata. This ensures consistent improvement and broader coverage when the skill is selected for execution.

Related Skills

Explore More Software Engineering Skills→
1molchuan1molchuan
1

oh-my-ccg-init

Verify MCP, Codex, and Gemini endpoints and initialize RPI state.

Community
Intermediate
timurgaleevtimurgaleev
5

canary

Monitor post-deploy web apps for defects and performance regressions.

Community
Advanced
yashs33244yashs33244
1

skill-creator

Draft SKILL.md files and run trigger evaluations for Claude skills.

Community
Advanced
mneves75mneves75
74

writing-skills

Apply Test-Driven Development to create and test AI skills.

Community
Intermediate