Skills | AI Engineering Curriculum

What Skills Are

Skills extend what Claude can do by giving it custom instructions it can follow on demand. You write a SKILL.md file, and Claude adds it to its toolkit.

Two ways they activate:

You invoke it - type /skill-name and Claude immediately follows the skill's instructions
Claude invokes it automatically - if your request matches the skill's description, Claude loads and applies it on its own

Custom slash commands (.claude/commands/) have been merged into skills. They're the same thing. Skills add more features on top.

File Structure

Each skill is a directory with SKILL.md as the required entrypoint:

my-skill/
├── SKILL.md           ← main instructions (required)
├── template.md        ← template for Claude to fill in
├── examples/
│   └── sample.md      ← example output showing expected format
└── scripts/
    └── validate.sh    ← script Claude can execute

Keep SKILL.md under 500 lines. Move large docs to separate files.

Where Skills Live

Location	Path	Applies to
Personal	`~/.claude/skills/<name>/SKILL.md`	All your projects
Project	`.claude/skills/<name>/SKILL.md`	This project only
Plugin	`<plugin>/skills/<name>/SKILL.md`	Where plugin is enabled
Enterprise	Managed settings	All org users

When skills share the same name, priority is: enterprise > personal > project.

Anatomy of a SKILL.md

YAML

---
name: explain-code
description: Explains code with visual diagrams and analogies. Use when explaining how code works or when the user asks "how does this work?"
---

When explaining code, always include:

1. **Start with an analogy** - compare to something from everyday life
2. **Draw a diagram** - use ASCII art to show flow or structure
3. **Walk through the code** - explain step by step
4. **Highlight a gotcha** - what's a common mistake?

Two parts:

YAML frontmatter (between ---) - tells Claude when and how to use it
Markdown body - the actual instructions Claude follows

All Frontmatter Fields

Field	Required	Description
`name`	No	Becomes the `/slash-command`. Uses directory name if omitted.
`description`	Recommended	What it does and when to use it. Claude uses this to decide when to auto-load it.
`argument-hint`	No	Hint shown in autocomplete. E.g. `[issue-number]`
`disable-model-invocation`	No	`true` = only YOU can invoke it. Claude won't auto-trigger it.
`user-invocable`	No	`false` = hidden from `/` menu. Only Claude can invoke it.
`allowed-tools`	No	Tools Claude can use without asking permission when this skill is active.
`model`	No	Override the model used when this skill runs.
`context`	No	Set to `"fork"` to run in an isolated subagent.
`agent`	No	Which subagent type to use when `context: fork` is set.
`hooks`	No	Hooks scoped to this skill's lifetime only.

Controlling Who Invokes a Skill

Frontmatter	You can invoke	Claude auto-invokes
(default)	✓	✓
`disable-model-invocation: true`	✓	✗
`user-invocable: false`	✗	✓

disable-model-invocation: true - use for skills with side effects you control: /deploy, /commit, /send-slack-message.

user-invocable: false - use for background knowledge Claude applies automatically but that isn't meaningful as a slash command.

Passing Arguments

YAML

---
name: fix-issue
disable-model-invocation: true
---

Fix GitHub issue $ARGUMENTS following our coding standards.

/fix-issue 123 → Claude receives "Fix GitHub issue 123 following our coding standards..."

Positional arguments:

YAML

Migrate the $0 component from $1 to $2.

/migrate-component SearchBar React Vue → $0=SearchBar, $1=React, $2=Vue

Dynamic Context Injection

The !`command` syntax runs a shell command before Claude sees anything. Output gets injected into the skill content.

YAML

---
name: pr-summary
context: fork
agent: Explore
allowed-tools: Bash(gh *)
---

## Pull request context
- PR diff: !`gh pr diff`
- PR comments: !`gh pr view --comments`

## Your task
Summarize this pull request clearly and concisely.

Claude only sees the rendered result with actual PR data - not the commands themselves.

Running Skills in a Subagent

Add context: fork to run in isolation - no access to conversation history, only the skill content as its task.

YAML

---
name: deep-research
description: Research a topic thoroughly using codebase exploration
context: fork
agent: Explore
---

Research $ARGUMENTS thoroughly:
1. Find relevant files using Glob and Grep
2. Read and analyze the code
3. Summarize findings with specific file references

agent options: Explore, Plan, general-purpose, or any custom subagent. Defaults to general-purpose.

Real Examples

/commit - structured git commit

YAML

---
name: commit
description: Create a structured git commit message
disable-model-invocation: true
allowed-tools: Bash(git *)
---

Review staged changes and create a commit:
1. Run `git diff --staged`
2. Write message: imperative verb, under 50 chars, explain the why in the body
3. Run `git commit -m "..."`

/standup - daily standup from git history

YAML

---
name: standup
disable-model-invocation: true
allowed-tools: Bash(git *)
---

Yesterday: !`git log --oneline --since="2 days ago" --until="yesterday" --author="$(git config user.name)"`
Today's staged: !`git diff --staged --stat`

Format as: Yesterday / Today / Blockers

/explain - explain code simply

YAML

---
name: explain
description: Explain code or a concept simply. Use when user asks "what does this do?"
---

Explain $ARGUMENTS as if to someone who doesn't code:
1. One sentence summary
2. A real-world analogy
3. What problem it solves
4. What would break if it didn't exist

String Substitutions

Variable	Description
`$ARGUMENTS`	Everything passed after the skill name
`$ARGUMENTS[N]`	Specific argument by 0-based index
`$N`	Shorthand for `$ARGUMENTS[N]`
`${CLAUDE_SESSION_ID}`	Current session ID

Power User Tips

/context - check if skills are hitting the context budget limit (descriptions always load; full content loads only on invocation)
Monorepo support - packages/frontend/.claude/skills/ is auto-discovered when working in that package
Restrict skill access - deny Skill in /permissions, or use Skill(deploy *) to block specific ones
ultrathink in skill content - enables extended thinking mode for that skill

Sources

Claude Code Skills Documentation (official docs)