Command Development FAQs

Question 1

What is the purpose of YAML frontmatter in command files?

Accepted Answer

Frontmatter allows you to define command metadata, such as descriptions for the /help menu, specific tool permissions (Read, Write, Bash), and the preferred AI model (Sonnet, Opus, or Haiku) for that specific task.

Question 2

How do dynamic arguments work in custom slash commands?

Accepted Answer

You can use the $ARGUMENTS variable to capture a user's entire input, or positional variables like $1, $2, and $3 to map specific words into your prompt, allowing for highly flexible and reusable command structures.

Question 3

Where should I store my custom Claude Code commands?

Accepted Answer

Store project-specific commands in the '.claude/commands/' directory of your repository, or save personal, cross-project commands in your home directory at '~/.claude/commands/'.

Question 4

What is the Command Development skill for Claude Code?

Accepted Answer

The Command Development skill enables developers to create, configure, and organize custom slash commands (like /review or /deploy) that automate complex prompts and multi-step workflows directly within the Claude Code CLI.

Question 5

Can I inject real-time project data into my commands?

Accepted Answer

Yes. By using the [BANG] prefix with backticks, you can execute bash commands (like 'git status' or 'ls') to gather real-time context that is automatically injected into the prompt before Claude processes it.

Command Development

Command Development

主な機能

ユースケース

主な機能

ユースケース