Introduction to Agent Skills

Stop repeating yourself to Claude Code

Your coding standards, PR format, and review checklist should live in Skills, not in every chat you open. Write them once. Let Claude reach for them automatically.

Taught by Edwin Ashdown. Founder of AIFirstPartner, 12 production AI systems shipped.

What You'll Learn

You're a developer or technical team lead using Claude Code who wants to stop repeating the same instructions in every conversation. By the end of this course you'll be able to:

Course Structure

1

Understanding Skills

Free

CLAUDE.md, subagents, hooks, MCP. Each one solves a different problem. Pick wrong and you fight the tool for the rest of the project.

3 lessons

Start free →
2

Building Skills

Free

Frontmatter that matches how real users phrase requests, tool restrictions for read-only review flows, and the progressive-disclosure pattern that lets you put a long reference behind a short SKILL.md.

4 lessons

3

Deploying and Maintaining Skills

Free

Commit to Git, publish a plugin, or enforce via enterprise settings. Plus the subagent gotcha that bites everyone: Skills don't auto-propagate into delegated work. Fix it here.

4 lessons

About the Instructor

Edwin Ashdown founded AIFirstPartner and has shipped 12 production AI systems for clients in government, legal, and enterprise. He is fractional CTO at Speculo, a cyber risk software company, and manages cloud and data platform engineering for a large Australian government department supporting 20+ agencies across the portfolio. He publishes his build process at 150days.aifirstpartner.com.

This course is the playbook he uses when onboarding a team onto Claude Code Skills.

Read more about the instructor →

Frequently Asked Questions

Why use this when the Anthropic docs are free?

The docs tell you how Skills work. This course shows you which Skill to write first, how to word a description so it actually triggers, and how to share Skills across a team without breaking anyone's setup. If you'd rather read reference material and experiment, the docs are genuinely good. If you want the shortest path to a working Skill your team picks up on git pull, that's what this course is for.

How long does the course take to complete?

The video content is about 30 minutes total. With the two exercises and the capstone project, plan for 1.5 to 2 hours end to end.

Will I build real Skills during the course?

Yes. Two hands-on exercises where you create and debug Skills for your own workflow, plus a capstone where you design a team skills architecture you could hand to a colleague on Monday.

What's the difference between Skills and CLAUDE.md?

CLAUDE.md loads into every conversation. It's for always-on project standards. Skills load on demand when Claude matches your request to a skill description. If your instructions only apply to specific tasks (like PR reviews or debugging), they belong in a Skill.

Do I need prior experience with Claude Code?

Yes. This course assumes you're already using Claude Code and are comfortable with the terminal, Git, and markdown. If you're new to Claude Code, start with Claude Code 101 first.

Is this affiliated with Anthropic?

No. Introduction to Agent Skills is independent. This course is not an Anthropic product and has not been reviewed or endorsed by them.

Built by Edwin Ashdown, maker of TranscriptIQ and AI IQ, tools that analyse Claude Code sessions. About the instructor →

Ready to build your first Skill?

Module 1 is free and takes about 10 minutes. Have your first Skill running by the end of the day.