Skill Creator

Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.

Overview

The Skill Creator is a specialized utility within the anthropics/skills repository designed to manage the full lifecycle of AI agent capabilities. Compatible with Claude and claude-code, this tool enables developers to build new skills from the ground up or refine existing ones through iterative improvements. It provides functional modules for running evaluations and benchmarking performance using variance analysis to ensure reliability. Additionally, the tool focuses on technical optimization by refining skill descriptions to enhance triggering accuracy during agent interactions. As part of the anthropics/skills project, which has gained significant community traction with over 150,000 stars, this skill supports diverse applications ranging from Python development and security reviews to data research and ROS integrations.

Use Cases

Generating new agent skills from scratch with optimized descriptions for improved triggering accuracy.
Benchmarking skill performance and conducting variance analysis to ensure consistent output quality.
Modifying and improving existing skill logic based on automated evaluation results and testing metrics.

Install Notes

# Review source first
open https://github.com/anthropics/skills/blob/main/skills/skill-creator/SKILL.md

Copy or clone the skill folder into your agent skills directory after reviewing its instructions and scripts.

Security Notes

When using Skill Creator to develop or modify agent capabilities, users should ensure that the resulting skills adhere to established security protocols for data handling and system access. Reviewing the logic of newly generated skills is essential to prevent unintended execution patterns, especially when the skills interact with external environments, browsers, or sensitive research data.

Related Skills

MCP Server Development Guide

anthropics/skills

Coding

Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).

CodexClaude
typescriptpython
150,001 starsSource linked

Building LLM-Powered Applications with Claude

anthropics/skills

Coding

This skill helps you build LLMpowered applications with Claude. Choose the right surface based on your needs, detect the project language, then read the relevant languagespecific documentation.

Claude CodeClaude
typescriptpython
150,001 starsSource linked

Improve Codebase Architecture

mxyhi/ok-skills

Coding

Find deepening opportunities in a codebase, informed by the domain language in CONTEXT.md and the decisions in docs/adr/. Use when the user wants to improve architecture, find refactoring opportunities, consolidate tightly-coupled modules, or make a codebase more testable and AI-navigable.

CodexClaude Code
designreview
423 starsApache-2.0

Karpathy Guidelines

mxyhi/ok-skills

Coding

Behavioral guidelines to reduce common LLM coding mistakes. Use when writing, reviewing, or refactoring code to avoid overcomplication, make surgical changes, surface assumptions, and define verifiable success criteria.

CodexClaude Code
review
423 starsApache-2.0

Vercel AI SDK — Build AI-Powered Apps in TypeScript

TerminalSkills/skills

Coding

You are an expert in the Vercel AI SDK, the TypeScript toolkit for building AIpowered applications. You help developers integrate LLMs (OpenAI, Anthropic, Google, Mistral, Ollama) with React Server Components, streaming UI, tool calling, structured output with Zod schemas, RAG pipelines, multistep agents, and edgecompa

CodexClaude Code
typescriptreact
71 starsApache-2.0