skill-creator

Create, test, and optimize skills for Claude Code.

Updated May 28, 2026
One-click install
npx skills add https://github.com/SensLiao/Claude-code-setting --skill skill-creator-sensliao
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: skill-creator
Source: https://github.com/SensLiao/Claude-code-setting/tree/main/skills/skill-creator
Command: npx skills add https://github.com/SensLiao/Claude-code-setting --skill skill-creator-sensliao

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires eval-viewer, scripts.run_loop, scripts.aggregate_benchmark, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill enables the creation, editing, and optimization of skills for Claude Code. It helps users write, test, and evaluate skills to improve their performance.

Core Features & Use Cases

  • Skill Creation: Draft and define new skills, including their behavior, triggers, and outputs.
  • Skill Evaluation: Run test cases, analyze performance, and iterate on skill improvements.
  • Skill Optimization: Optimize skill descriptions for better triggering accuracy.
  • Use Case: For a user looking to create a skill for data analysis, the Skill Creator can guide them through the process, including defining the skill's behavior, setting up test cases, and iterating on the skill based on feedback.

Quick Start

To create a new skill, run the following commands:

skill-creator create --name "my-data-analysis-skill"
skill-creator run-tests --name "my-data-analysis-skill" --prompt "Analyze sales data"

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 test custom skills for Claude Code?

To create custom skills for Claude Code, define the skill's behavior, triggers, and outputs, then run test cases and iterate on improvements using the skill testing environment. This handles skill definition, performance analysis, and description optimization.

What is the best way to optimize Claude Code skill descriptions for better triggering?

Optimizing Claude Code skill descriptions involves running test cases, analyzing performance metrics, and iteratively refining the definition to improve triggering accuracy. This process ensures the subagents invoke the correct skill.

Can I evaluate Claude Code skills by running specific prompt test cases?

Yes, you can evaluate Claude Code skills by running specific prompt test cases against your defined skill. The system analyzes the performance results and guides you through iterative improvements based on the feedback.

Do I need to interact with subagents to build multi-step skills in Claude Code?

Yes, building complex, multi-step skills in Claude Code requires interaction with the subagents and the scripting environment. This integration handles skill design, testing, and iterative improvement throughout the creation process.

Why does my Claude Code skill trigger incorrectly during testing?

Incorrect skill triggering during testing indicates a need for description optimization. You must analyze the benchmark performance, refine the skill definition, and rerun the tests to achieve better triggering accuracy.

Does skill-creator work with the Claude Code scripting environment?

Yes, skill-creator works directly with the Claude Code scripting environment. It requires interaction with subagents and scripts to handle complex, multi-step tasks involving skill design and iterative improvement.