ع
Learn Tracks Reference Guides Saved
feature

Slash commands

Type a forward slash to run a saved, named action instead of writing the whole request.

Typing / in the chat shows a menu of built-in commands — and any you've saved yourself. They're shortcuts for things you do often, so you stop re-typing the same long prompt. This works the same in Claude Desktop and the terminal: just type the slash in the chat box. You can write your own by dropping a prompt in a file and giving it a name.

There are two kinds. Built-in commands ship with Claude Code — /review, /plan, /commit, /clear and more — and cover the moves you make constantly. Custom ones are yours: any prompt you keep re-typing can become a named command your whole team shares. You reach all of them the same way — type / in the chat — whether you’re in Claude Desktop or the terminal.

Making one is deliberately small. A custom command is just a prompt saved in a file under .claude/commands/ — the filename becomes the command name. Drop in the instruction you keep giving (“summarize what changed since yesterday and draft a standup update”), and from then on /standup runs it. Commands can take arguments too, so one saved prompt handles many specific cases.

Use a command when you decide the moment — it’s an action you trigger. That’s the difference from a skill, which Claude applies on its own when it’s relevant. A good rule of thumb: the third time you paste the same long instruction, stop and save it as a command.

the slash menu, for real

In Claude Code, typing / opens this exact menu. Try it — type, arrow through, and hit enter to learn what each one does.

claude
  • /clearWipe the conversation and start with a clean context window.
  • /compactSummarize the conversation so far to free up room — keeping the gist.
  • /initHave Claude study your project and write a CLAUDE.md for it.
  • /modelSwitch which Claude model is doing the work.
  • /helpList the slash commands available in your session.
  • /reviewHave Claude review a pull request or your pending changes.
  • /agentsCreate and manage subagents — specialized helpers for focused jobs.
  • /memoryOpen your CLAUDE.md memory files to add lasting instructions.
  • /mcpView and manage the MCP servers connected to this session.
  • /configOpen Claude Code settings — theme, model, permissions, and more.
  • /costShow how many tokens this session has used, and the cost so far.
  • /loginSign in, or switch which account Claude Code uses.
  • /statusShow version, account, and your current setup at a glance.
  • /vimToggle vim-style editing in the input prompt — a Power-Track comfort for terminal users.
  • /permissionsSee and change what Claude is allowed to do without asking each time.
  • /doctorRun a health check on your Claude Code install, version, and config to find what's wrong.
  • /terminal-setupTeach your terminal the shortcut for a newline (Shift+Enter) — a Power-Track setup step.
  • /exportSave the current conversation to a file or your clipboard.
  • /bugReport a problem with Claude Code to Anthropic, with context attached.
  • /logoutSign out of your Anthropic account in Claude Code.
  • /add-dirGive Claude access to another folder beyond the one you started in.
  • /pr-commentsPull the review comments from a GitHub pull request into the session.
  • /output-styleChange how Claude talks back — its tone and level of detail.
  • /hooksSet up commands that run automatically at certain moments.
  • /loopRun a prompt or slash command on a repeating interval until you stop it.
  • /scheduleSchedule a cloud agent to run once or on a recurring cron — even when you're not around.
  • /effortSet how hard Claude thinks before it answers — low for quick edits, high for tricky problems.
  • /fastToggle fast mode — trades a bit of cost for roughly 2.5× the response speed.
  • ! <command>Run a shell command from inside the chat and see the output land directly in the conversation.
  • /goalSet a lasting goal and let Claude keep working and re-checking until it's met.
  • /workflowsView, manage, and resume the multi-agent workflows running in your session.
  • /autofix-prLet Claude watch your open pull request in the cloud and push fixes until it's green.
moveexplainescclear

why it helps Turn a paragraph you keep typing into a one-word command your whole team can reuse.

examples
Built-in commands you already have
/review · /plan · /commit · /clear
Run a custom command you saved
/standup
tips & best practices
  • Type / and pause — the menu shows everything available, including your own.
  • If you keep pasting the same long instruction, that's a sign it should become a command.
  • Custom commands are just a saved prompt in a file — start one the moment a workflow repeats.

Questions people ask

How do I see what slash commands are available?
Type `/` in the chat and pause — a menu shows every command available, including the built-in ones and any you've saved yourself. It's the same in Claude Desktop and the terminal.
Can I make my own slash command?
Yes. A custom command is just a prompt saved in a file under `.claude/commands/` — the filename becomes the command name. Drop in the instruction you keep re-typing and it becomes a one-word command your whole team can reuse.
What's the difference between a slash command and a skill?
A slash command runs when *you* call it — it's an action you trigger. A skill applies on its own when Claude decides it's relevant. Use a command for the moments you choose; use a skill for a way of working you want followed automatically.
Do I need to code to write a custom command?
No. A custom command is plain text — the prompt you'd otherwise type — saved in a file. No code required.
Do slash commands work in Claude Desktop?
Yes. Type `/` in the Desktop chat box and the same menu of built-in and custom commands appears, exactly as in the terminal. The shortcut is part of the chat, not the command line.