AIv2.1BEGINNERClaude Code Slash Commands
Claude Code slash commands cheat sheet — complete reference for all built-in CLI commands with syntax, options, and usage examples.
8 min read
claudeclaude-codeslash-commandsclicommands
Other AI Sheets
Sign in to mark items as known and track your progress.
Sign inCore Commands
Essential commands for managing your session
/help
Display all available commands
bash
/help💡 Lists all available slash commands
⚡ Shows built-in, custom, and plugin commands
🔍 Project commands labeled "(project)", user commands labeled "(user)"
🎯 Run anytime to see what commands are available
helpcoredocumentation
/clear
Clear conversation history and start fresh
bash
/clear💡 Removes entire conversation history
⚡ Starts fresh while staying in same session
🔍 Does not exit the CLI
🎯 Use /resume to restore previous conversation
clearcoreconversation
/exit
Exit Claude Code CLI
bash
/exit💡 Closes Claude Code CLI session
⚡ Conversation auto-saved for later
🔍 Can also use Ctrl+C or Ctrl+D
🎯 Use /resume next time to continue conversation
exitcoresession
/status
View current status and configuration
bash
/status💡 Opens Settings interface showing status
⚡ Check version, model, account, connectivity
🔍 Verify your current configuration
🎯 Quick way to see what's active
statuscoreconfig
Conversation Management
Manage conversation state, history, and context
/compact
Compress conversation to reduce token usage
bash
/compact
/compact [focus instructions]💡 Compresses conversation to save tokens
⚡ Add focus instructions to preserve specific content
🔍 Helpful when approaching context limits
🎯 Reduces costs while keeping important info
compactconversationtokens
/context
Visualize current context usage
bash
/context💡 Shows visual grid of context usage
⚡ Color-coded to show usage levels
🔍 Helps monitor conversation size
🎯 Know when to compact or start fresh
contextconversationtokens
/resume
Restore previous conversation
bash
/resume💡 Restores previous conversation from last session
⚡ Continues where you left off
🔍 Works after /clear or /exit
🎯 Conversations auto-save when exiting
resumeconversationhistory
/rewind
Return conversation to earlier state
bash
/rewind💡 Returns conversation and code to earlier state
⚡ Interactive selection of rewind point
🔍 Can revert both messages and code changes
🎯 Use to undo mistakes or try different approaches
rewindconversationundo
/export
Export conversation to file or clipboard
bash
/export
/export [filename]💡 Saves conversation to file or clipboard
⚡ Specify filename or let it copy to clipboard
🔍 Exports as formatted Markdown
🎯 Great for documentation or sharing
exportconversationsave
Project & Files
Project initialization and file management
/init
Initialize project with CLAUDE.md
bash
/init💡 Creates CLAUDE.md for project-specific instructions
⚡ Stores persistent context and guidelines
🔍 Shared with team via git
🎯 Claude reads this automatically each session
initprojectsetup
/memory
Edit CLAUDE.md memory file
bash
/memory💡 Opens CLAUDE.md for editing
⚡ Define coding standards and project context
🔍 Claude reads this automatically
🎯 Reduces need to repeat instructions
memoryprojectcontext
/add-dir
Add supplementary working directories
bash
/add-dir <directory-path>💡 Adds additional directories to Claude's workspace
⚡ Useful for monorepos and shared libraries
🔍 Claude can read/write files in added dirs
🎯 Expand scope beyond current working directory
add-dirprojectworkspace
Code Review & Quality
Code analysis, review, and security checks
/review
Request code review from Claude
bash
/review💡 Claude reviews your code for quality and issues
⚡ Checks best practices, bugs, performance
🔍 Reviews recent changes and context
🎯 Use before committing or deploying
reviewcode-qualitybest-practices
/security-review
Comprehensive security audit
bash
/security-review💡 Complete security audit of pending changes
⚡ Checks for common vulnerabilities
🔍 Reviews authentication, authorization, data exposure
🎯 Essential before production deployments
securityauditvulnerabilities
/bug
Report bugs to Anthropic
bash
/bug💡 Submit bug report to Anthropic
⚡ Includes conversation context for debugging
🔍 Helps improve Claude Code
🎯 Use for crashes, errors, unexpected behavior
bugsupportreport
Configuration & Settings
Customize Claude Code behavior and preferences
/config
Open Settings interface
bash
/config💡 Opens full Settings interface
⚡ Configure model, output style, integrations
🔍 Central place for all preferences
🎯 Also use specific commands like /model
configsettingspreferences
/model
Switch AI model
bash
/model💡 Switch between Sonnet, Opus, and Haiku
⚡ Choose based on task complexity
🔍 Sonnet is default and best for most tasks
🎯 Opus for complex work, Haiku for simple tasks
modelaisettings
/output-style
Set response output style
bash
/output-style
/output-style explanatory💡 Customize how Claude responds
⚡ Built-in styles: default, explanatory, learning
🔍 Interactive menu or set directly
🎯 Create custom styles with /output-style:new
output-stylepreferencesresponse
Tools & Integration
Manage plugins, MCP servers, and integrations
/mcp
Manage Model Context Protocol servers
bash
/mcp💡 Manage Model Context Protocol server connections
⚡ Connect external tools and services to Claude
🔍 OAuth authentication support
🎯 Exposes custom slash commands per server
mcpintegrationprotocol
/plugin
Manage Claude Code plugins
bash
/plugin💡 Install and manage Claude Code plugins
⚡ Extend functionality with community tools
🔍 Plugins provide custom slash commands
🎯 Browse marketplace or install from GitHub
pluginextensionsmarketplace
/hooks
Configure event-triggered automation
bash
/hooks💡 Trigger shell commands on tool events
⚡ Automate linting, testing, commits
🔍 Configure in settings.json under the hooks key
🎯 Great for workflow automation
hooksautomationevents
System & Diagnostics
Health checks and system administration
/doctor
Check Claude Code installation health
bash
/doctor💡 Comprehensive installation health check
⚡ Diagnoses configuration and connectivity issues
🔍 Provides fix suggestions
🎯 Run when experiencing problems
doctordiagnosticshealth
/permissions
Manage tool and file access
bash
/permissions💡 Control what Claude can access and execute
⚡ Configure per-tool permissions
🔍 Set file access restrictions
🎯 Important for security and control
permissionssecurityaccess
/sandbox
Enable isolated execution environment
bash
/sandbox💡 Enable isolated, secure execution environment
⚡ Filesystem and network isolation
🔍 Safer for running untrusted code
🎯 Trade-off: security vs functionality
sandboxsecurityisolation
Custom Commands
Create your own slash commands
Project Custom Command
Create team-shared custom command
markdown
# .claude/commands/review-pr.md
---
description: Review pull request changes
---
Review the current PR and provide feedback💡 Create in .claude/commands/ for team sharing
⚡ Shared via git, labeled "(project)"
🔍 Use $1, $2 for arguments
🎯 Great for team workflows and standards
customprojectcommands
User Custom Command
Create personal custom command
markdown
# ~/.claude/commands/daily-standup.md
---
description: Generate daily standup summary
argument-hint: <date>
---
Summarize work completed on $1💡 Create in ~/.claude/commands/ for personal use
⚡ Available across all projects
🔍 Not shared via git, labeled "(user)"
🎯 Build your personal toolkit
customuserpersonal
Command with File Reference
Use @file syntax in custom commands
markdown
# Command that reads files
---
description: Analyze file for security issues
---
Review @$1 for security vulnerabilities💡 Use @ prefix to include file contents
⚡ Works with $1, $ARGUMENTS for dynamic files
🔍 Claude reads files automatically
🎯 Simplifies file-based commands
customfilesreference
Command with Bash Execution
Run bash before command prompt
markdown
# Command with bash prefix
---
description: Test and review results
allowed-tools: [Bash]
---
!npm test
Review test results above💡 Use ! prefix to run bash before prompt
⚡ Requires allowed-tools: [Bash] in frontmatter
🔍 Command output included in context
🎯 Perfect for test/build automation
custombashautomation