Day 1: Install Claude Code and run your first slash command

Claude Code is not a chat window with a terminal attached. It is two products fused at the prompt: a conversation, and a command palette. Until you find the keystroke that surfaces the second product, nothing else in the tool makes sense.

What we tried

We followed the install docs, ran claude inside a project, and typed a paragraph describing what we wanted. It answered. Then we stalled. There was no obvious way to ask it to run tests, open a file, or do anything beyond chat.

What happened

A single keystroke fixed the mental model. Typing / opens the palette of everything Claude Code can do in this repo, right now: built-in commands, commands from installed plugins, and every skill you (or your team) have registered. It is the tool's on-ramp, its documentation, and its diff against the chat-in-a-browser experience all at once.

How the pieces fit together

Everything in the right column touches the same working directory. That is why running claude from your home folder is almost always wrong: you get the tool, but not the context.

What we learned

• Install from the official docs at https://docs.claude.com/en/docs/claude-code. The one-line installer is the supported path; everything else is a workaround.
• Start claude inside a project directory, never in ~. Claude Code reads files relative to where you launched it, and your home folder is the one place you do not want it reading.
• Type / to see every command available in the current repo, including ones shipped by plugins and skills. /help is the one to memorise first.
• Your first real action is almost always /init. It scaffolds a CLAUDE.md, the single most important piece of configuration you will write in this tool. Day 2 covers it.

Next

• Day 2. Your first CLAUDE.md.