Introduction to Claude Skills

What are Claude Skills?

A skill is a SKILL.md file that gives Claude a set of instructions, a checklist, or a multi-step procedure. Claude loads the skill when it's relevant, or you invoke it directly by typing /skill-name.

Skills work across Claude.ai, Claude Code, and the API — write once, use everywhere. They follow the open Agent Skills standard, so they're portable across AI tools.

When should you create a skill?

Create a skill when you notice any of these patterns:

Skills vs CLAUDE.md

Anatomy of a SKILL.md file

Every skill is a directory named after the command you want to type. Inside that directory, SKILL.md is the only required file. It has two parts:

Example: a skill that summarises uncommitted git changes and flags risks

The !`git diff HEAD` line is called dynamic context injection: Claude Code runs the command before sending the skill to Claude, so Claude always sees real, live data — not a static placeholder.

How to invoke a skill

There are two ways to run a skill:

Where skills live

Where you store a skill determines who can use it:

Bundled skills — available immediately

Claude Code ships with a set of ready-to-use bundled skills. Unlike fixed commands, these are prompt-based — they give Claude detailed instructions and let it orchestrate the work. You invoke them the same way as custom skills.

Create your first skill in 3 steps

Here's the simplest possible skill — a personal commit helper that stages all changes and writes a commit message.

mkdir -p ~/.claude/skills/commit

---
name: commit
description: Stage all changes and write a descriptive commit message.
disable-model-invocation: true
allowed-tools: Bash(git add *) Bash(git commit *)
---

Stage all changes and write a commit message that follows this format:

<type>(<scope>): <summary>

Types: feat, fix, chore, docs, refactor, test
Keep the summary under 72 characters.
Write a short body if the change needs context.

Ways to build a skill

Quick Reference