20 Official Claude Code Commands You Should Know

Claude Code has a set of built in slash commands that make it easier to manage projects, conversations, models, permissions, and development tasks.

These commands start with / and can be entered directly in Claude Code. Here are 20 commands currently listed in Anthropic’s official documentation.

Essential Claude Code Commands

1. /add-dir

Adds additional working directories to your Claude Code session. This is useful when your project needs access to files outside the main working directory.

2. /agents

Lets you manage custom AI subagents in Claude Code. Subagents can be used for specific tasks within a larger coding workflow.

3. /bug

Lets you report a bug to Anthropic. The command can send relevant conversation information along with your report.

4. /clear

Clears the current conversation history. It’s useful when you want to start a new task without the previous conversation taking up context.

5. /compact

Compacts the current conversation by summarizing its context. This can help when a long coding session starts using a large amount of context.

6. /config

Opens Claude Code configuration settings. You can use it to view or change available configuration options.

7. /cost

Shows token usage information for your Claude Code session. It can help you keep track of how much context and usage your session is consuming.

8. /doctor

Checks your Claude Code installation and environment for common problems. It’s useful when Claude Code isn’t working as expected.

9. /help

Shows help information and available Claude Code commands. If you’re unsure what a command does, this is one of the first commands to try.

10. /init

Sets up a CLAUDE.md file in your project. This file can contain instructions and information that Claude Code should follow when working on the project.

11. /login

Lets you log in to Claude Code or switch between accounts. It’s useful when you need to change the account currently connected to Claude Code.

12. /logout

Signs you out of your current Claude Code account.

13. /mcp

Manages MCP server connections. MCP allows Claude Code to connect with external tools and services that support the Model Context Protocol.

14. /memory

Helps manage Claude Code memory files, including CLAUDE.md files used to store project instructions and useful context.

15. /model

Lets you view or change the Claude model being used for your session.

16. /permissions

Lets you view and update Claude Code permissions. This controls what actions Claude Code is allowed to perform.

17. /pr_comments

Displays comments associated with a pull request. This can be useful when working on code changes that are being reviewed.

18. /review

Requests a code review from Claude. It can help identify potential problems and areas that may need attention before changes are finalized.

19. /status

Shows the current Claude Code status, including relevant account and system information.

20. /terminal-setup

Helps configure terminal key bindings for Claude Code. This can make interacting with Claude Code from the terminal more convenient.

These Claude Code slash commands cover many of the tasks developers commonly need during a coding session. Some are simple commands, such as /help, /clear, and /status, while others, including /mcp, /agents, and /permissions, are designed for more advanced workflows.

Claude Code is updated regularly, so commands and features can change over time. If you’re using a newer version, checking /help inside Claude Code is a quick way to see which commands are available in your installation.