Build Better AI Agent Skills With Skill Creator v2 from Anthropic
In this video, I walk through Anthropic’s Skill Creator v2 and show how it works and then use it to review and improve a real skill inside I created. It's a long video but it's pretty impressive what the new skill creator skill does. You’ll see how to: install Skill Creator from skills.sh inspect what’s included in the skill run it against an existing README Wizard skill I created use sub-agents, evals, and benchmark comparisons review results in the HTML eval viewer optimize the skill description for better triggering accuracy This is a practical look at how to make your agent skills more reliable, measurable, and easier to improve over time. If you’re building custom skills for GitHub Copilot, Claude Code, or other coding agents, this workflow is worth adding to your toolbox. 0:00 Intro and Find Skill Creator on skills.sh 1:29 Install Skill Creator 3:09 Explore what the skill includes 5:28 Test it on a real README skill 8:59 Review problems and planned fixes 12:54 Run evals and sub-agents 14:10 Compare benchmark results 18:58 Open the HTML eval viewer 23:45 Optimize the skill description Links mentioned in the video: skills.sh https://skills.sh/anthropics/skills/s... npx skills add https://github.com/anthropics/skills --skill skill-creator #VSCode #GitHubCopilot #AIEngineering #AgentSkills

Claude Code Skills Just Got a MASSIVE Upgrade

How to Get and Evaluate Startup Ideas | Startup School

I Got Laid Off From Block — My Second Layoff in Months | The Reality of Tech in 2026

Don't Build Agents, Build Skills Instead – Barry Zhang & Mahesh Murag, Anthropic

or how programming is all about creativity

What Are Skills? Build Your First AI Agent Skill in Minutes

Is RAG Still Needed? Choosing the Best Approach for LLMs

How Anthropic Engineers ACTUALLY Prompt Claude Code

The complete guide to Agent Skills

Model Context Protocol (MCP) Explained for Beginners: AI Flight Booking Demo!

Agentic Testing and Browser Automation with Playwright CLI Skill Live Stream

Agent Skills vs MCP: What’s the difference?

The Best Local Agentic Coding Workflow (Complete Guide)

The Multi-Agent Architecture That Actually Ships — Luke Alvoeiro, Factory

How to Use Playwright CLI Skill for Agentic Testing

Android 17 sucks. So I put Linux on a phone.

Claude Code best practices | Code w/ Claude

Building AI Agents that actually work (Full Course)

Vintage Pink Peony Painting | Frame TV Art 4K Screensaver (3 Hours)

