Slash Commands Reference

Complete reference for all built-in and custom slash commands in Claude Code.

Built-in Commands

Session Management

Command	Description
/clear	Clear conversation history
/exit	Exit Claude Code
/resume	Resume a previous conversation
/compact [instructions]	Compact conversation with optional focus
/rewind	Rewind conversation and/or code

Configuration

Command	Description
/config	Open Settings interface (Config tab)
/status	Open Settings interface (Status tab)
/model	Select or change the AI model
/permissions	View or update permissions
/privacy-settings	View and update privacy settings
/output-style [style]	Set output style
/vim	Enter vim mode

Tools & Integrations

Command	Description
/mcp	Manage MCP server connections
/hooks	Manage hook configurations
/plugin	Manage Claude Code plugins
/agents	Manage custom AI subagents
/ide	Manage IDE integrations

Project & Files

Command	Description
/init	Initialize project with CLAUDE.md
/memory	Edit CLAUDE.md memory files
/add-dir	Add working directories
/context	Visualize current context usage

Git & Code Review

Command	Description
/review	Request code review
/pr-comments	View pull request comments
/security-review	Security review of changes
/install-github-app	Setup GitHub Actions

Diagnostics

Command	Description
/doctor	Check installation health
/cost	Show token usage statistics
/usage	Show plan limits and rate status
/bashes	List background tasks
/todos	List current todo items

Other

Command	Description
/help	Get usage help
/bug	Report bugs to Anthropic
/export [filename]	Export conversation
/login	Switch Anthropic accounts
/logout	Sign out
/release-notes	View release notes
/sandbox	Enable sandboxed bash
/statusline	Setup status line UI
/terminal-setup	Install keybindings

Custom Slash Commands

Create custom commands as Markdown files.

Locations

Location	Scope
.claude/commands/	Project (team-shared)
~/.claude/commands/	Personal (all projects)

Syntax

---
allowed-tools: Bash(git add:*), Bash(git status:*)
description: Create a git commit
model: claude-sonnet-4-5
argument-hint: commit message
---

Create a commit with message: $ARGUMENTS

Arguments

Placeholder	Purpose
$ARGUMENTS	All arguments
$1, $2	Specific positions

Frontmatter Options

Option	Description
allowed-tools	Tools command can use
description	Command description
model	Model to use
argument-hint	Help text for arguments

Examples

Git Commit Command

.claude/commands/commit.md:

---
allowed-tools: Bash(git add:*), Bash(git commit:*)
description: Stage and commit changes
argument-hint: commit message
---

Stage all changes and create a commit with message: $ARGUMENTS

Usage:

/commit fix: resolve login issue

Deploy Command

.claude/commands/deploy.md:

---
allowed-tools: Bash(npm:*), Bash(vercel:*)
description: Deploy to production
---

Run npm build and deploy to Vercel production

Test Command

.claude/commands/test.md:

---
allowed-tools: Bash(npm test:*)
description: Run tests with coverage
---

Run the test suite with coverage report and summarize results

Bash Execution

Prefix with ! to run bash before command:

!/my-command

File References

Use @ to include file contents:

Review the changes in @src/main.ts

MCP Slash Commands

MCP servers expose prompts as slash commands:

/mcp__<server>__<prompt> [arguments]

Examples:

/mcp__github__list_prs
/mcp__notion__search_pages "project plans"
/mcp__memory__save_note "meeting notes"

Tips

1. Tab completion: Commands autocomplete
2. Help: Use /help for command list
3. Arguments: Most commands accept arguments
4. Chaining: Combine commands in prompts

Next Steps

• Hooks Reference - Automate with hooks
• MCP Guide - Connect external tools
• CLI Reference - Command-line options