Claude Code Cheatsheet

Last updated: May 2025 · Official Docs

A PDF version of this cheat sheet can be found here.

🚀 CLI — Session Management

Commands run from your terminal to start and manage sessions.

Command	Description	Example
claude	Start an interactive session	claude
claude "query"	Start a session with an initial prompt	claude "explain this project"
claude -p "query"	Print response and exit (non-interactive)	claude -p "explain this function"
cat file | claude -p "query"	Process piped content	cat logs.txt | claude -p "explain"
claude -c	Continue the most recent conversation	claude -c
claude -r "<session>" "query"	Resume a session by ID or name	claude -r "auth-refactor" "Finish this PR"
claude -n "name"	Name the current session	claude -n "my-feature-work"
claude update	Update to the latest version	claude update
claude install [version]	Install a specific version (stable, latest, or 2.x.x)	claude install stable

🔐 CLI — Authentication

Command	Description	Example
claude auth login	Sign in to your Anthropic account	claude auth login --console
claude auth logout	Log out from your Anthropic account	claude auth logout
claude auth status	Show authentication status	claude auth status --text
claude setup-token	Generate a long-lived OAuth token for CI/scripts	claude setup-token
claude auth login --sso	Force SSO authentication	claude auth login --sso

⚙️ CLI — Configuration Flags

Pass these flags when launching Claude Code to customize behavior.

Flag	Description	Example
--model <model>	Set model for the session (sonnet, opus, or full name)	claude --model claude-opus-4-6
--effort <level>	Set effort level: low, medium, high, xhigh, max	claude --effort high
--permission-mode <mode>	Start in a permission mode: default, acceptEdits, plan, auto, bypassPermissions	claude --permission-mode plan
--add-dir <path>	Add additional working directories	claude --add-dir ../apps ../lib
--append-system-prompt "..."	Append custom text to the system prompt	claude --append-system-prompt "Always use TypeScript"
--system-prompt "..."	Replace the entire system prompt	claude --system-prompt "You are a Python expert"
--system-prompt-file <file>	Load system prompt from a file	claude --system-prompt-file ./custom-prompt.txt
--tools "..."	Restrict which tools Claude can use	claude --tools "Bash,Edit,Read"
--allowedTools "..."	Tools that run without a permission prompt	"Bash(git log *)" "Read"
--disallowedTools "..."	Tools that are denied or removed	"Bash(rm *)" "Edit"
--mcp-config <file>	Load MCP servers from a JSON file	claude --mcp-config ./mcp.json
--plugin-dir <path>	Load a plugin for this session	claude --plugin-dir ./my-plugin
--bare	Minimal mode — skips hooks, skills, MCP, CLAUDE.md	claude --bare -p "query"
--debug	Enable debug mode (optional category filter)	claude --debug "api,mcp"
--verbose	Show full turn-by-turn output	claude --verbose
--max-turns <n>	Limit agentic turns (print mode only)	claude -p --max-turns 3 "query"
--max-budget-usd <n>	Budget cap in dollars (print mode only)	claude -p --max-budget-usd 5.00 "query"
--bg	Start as a background agent and return immediately	claude --bg "investigate the flaky test"
--output-format <fmt>	Output format: text, json, stream-json	claude -p "query" --output-format json
--worktree, -w	Start in an isolated git worktree	claude -w feature-auth
--chrome	Enable Chrome browser integration	claude --chrome
--version, -v	Output the version number	claude -v

🔧 Configuration (Inside Session)

Use these slash commands to configure Claude from within an active session.

Command	Description
/config	Open Settings UI (theme, model, output style, etc.). Alias: /settings
/model [model]	Select or change the AI model interactively
/effort [level|auto]	Set effort level (low, medium, high, xhigh, max) or use the interactive slider
/theme	Change the color theme (light, dark, daltonized, ANSI, custom)
/permissions	Manage allow/ask/deny rules for tools. Alias: /allowed-tools
/hooks	View hook configurations for tool events
/mcp	Manage MCP server connections and OAuth
/agents	Manage subagent configurations
/terminal-setup	Configure Shift+Enter and other terminal keybindings
/keybindings	Open or create your keybindings configuration file
/statusline	Configure the Claude Code status line

📄 Context Management

Commands to understand and manage the context window.

Command	Description
/context [all]	Visualize context usage as a colored grid with optimization tips
/compact [instructions]	Summarize the conversation to free up context. Pass focus instructions: /compact keep only the plan and diff
/clear [name]	Start a fresh conversation (previous stays in /resume). Aliases: /reset, /new
/btw <question>	Ask a side question without adding it to conversation history

ℹ️ Information & Status

Command	Description
/help	Show help and all available commands
/doctor	Diagnose installation, settings, MCP servers, and context usage
/status	Show version, model, account, and connectivity (works while Claude is responding)
/usage	Show session cost, plan limits, and activity stats. Aliases: /cost, /stats
/skills	List available skills (press t to sort by token count)
/memory	Edit CLAUDE.md memory files and manage auto-memory
/context	Show current context window usage
/release-notes	View the changelog in an interactive version picker
/insights	Analyze your Claude Code sessions for patterns and friction points

🗄️ Memory & Project Setup

Command	Description
/init	Initialize project with a CLAUDE.md guide
/memory	Edit CLAUDE.md files, enable/disable auto-memory, view entries
/team-onboarding	Generate a team onboarding guide from your usage history

CLAUDE.md locations:

Scope	Path
Global (personal)	~/.claude/CLAUDE.md
Project	.claude/CLAUDE.md or CLAUDE.md in repo root

🔀 Session Navigation

Command	Description
/resume [session]	Resume a conversation by ID or name. Alias: /continue
/branch [name]	Fork the current conversation at this point. Alias: /fork
/rename [name]	Rename the current session
/clear	Start a new conversation with empty context
/export [filename]	Export the current conversation as plain text
/rewind	Roll back code and conversation to a checkpoint. Aliases: /checkpoint, /undo

📋 Workflow Commands (Slash)

Commands to manage and accelerate your development workflow.

Command	Description
/plan [description]	Enter plan mode. Pass a description to start immediately
/diff	Open an interactive diff viewer for uncommitted changes and per-turn diffs
/copy [N]	Copy the last assistant response to clipboard. Pass N to copy the Nth-latest
/goal [condition]	Set a persistent goal; Claude works across turns until it's met
/background [prompt]	Detach session to run as a background agent. Alias: /bg

⚡ Parallel Work & Agents

Command	Description
/agents	Open the agent manager to monitor and dispatch parallel sessions
/tasks	List and manage background tasks. Alias: /bashes
/batch <instruction>	[Skill] Break a large change into parallel units, each running in its own worktree
/background	Detach the current session to keep running in the background. Alias: /bg
claude agents	Open agent view from the terminal (CLI command)
claude attach <id>	Attach to a background session in this terminal
claude stop <id>	Stop a background session
claude logs <id>	Print recent output from a background session
claude respawn <id>	Restart a background session with its conversation intact
claude rm <id>	Remove a background session from the list

🔍 Code Quality (Bundled Skills)

These are built-in skills invoked with / commands.

Command	Description
/review [PR]	Review a pull request locally in the current session
/security-review	Analyze pending changes for security vulnerabilities
/simplify [focus]	[Skill] Review recently changed files for reuse, quality, and efficiency issues, then fix them
/ultrareview [PR]	Deep multi-agent cloud-based code review
/debug [description]	[Skill] Enable debug logging and troubleshoot session issues
/fewer-permission-prompts	[Skill] Scan transcripts and add allowlist rules to reduce permission prompts

🛠️ Debugging & Diagnostics

Command	Description
/doctor	Run an automated check of installation, settings, MCP, and context
claude doctor	Same as /doctor but from the shell (use when claude won't start)
/debug [description]	Enable debug logging mid-session
/heapdump	Write a JS heap snapshot to ~/Desktop for memory diagnostics
/feedback [report]	Submit feedback or bug reports to Anthropic. Alias: /bug
claude --debug "api,mcp"	Start with debug output (category-filtered)
claude --debug-file <path>	Write debug logs to a specific file

🧰 MCP & Plugins

Command	Description
/mcp	Manage MCP server connections and OAuth
claude mcp	Configure MCP servers from the CLI
/plugin	Manage Claude Code plugins
claude plugin install <name>	Install a plugin
/reload-plugins	Reload all active plugins without restarting
--mcp-config <file>	Load MCP servers from a JSON file at launch
--strict-mcp-config	Use only MCP servers from --mcp-config

🌐 Remote & Web

Command	Description
/remote-control	Make this session available for remote control from claude.ai. Alias: /rc
claude remote-control	Start a Remote Control server (CLI)
claude --remote "task"	Create a new web session on claude.ai
/teleport	Pull a Claude Code web session into this terminal. Alias: /tp
/autofix-pr [prompt]	Spawn a web session that watches the current PR and pushes fixes
/desktop	Continue this session in the Claude Code Desktop app

⌨️ Keyboard Shortcuts (Interactive Mode)

Shortcut	Action
Ctrl+C	Cancel current operation
Shift+Tab	Cycle through permission modes (default → plan → auto → bypassPermissions)
Ctrl+R	Search prompt history
↑ / ↓	Navigate prompt history
Tab	Autocomplete skill/command name

📁 Key File Locations

File / Directory	Purpose
CLAUDE.md (project root)	Project-level memory and instructions
~/.claude/CLAUDE.md	Personal global memory
.claude/settings.json	Project settings
.claude/settings.local.json	Local (gitignored) project settings
~/.claude/settings.json	User-level settings
~/.claude/skills/<name>/SKILL.md	Personal custom skills
.claude/skills/<name>/SKILL.md	Project-specific custom skills
~/.claude/themes/	Custom themes

💡 Quick Tips

• Type / in a session to see all available commands with live filtering.
• Use Shift+Tab to cycle permission modes without restarting.
• /compact with instructions (e.g., /compact keep only the plan) lets you surgically preserve what matters.
• CLAUDE.md is read at every session start — keep it focused and under 500 lines.
• Use disable-model-invocation: true in a skill's frontmatter to require manual /skill-name invocation.
• claude -p (print mode) is ideal for scripting and CI pipelines.
• Background sessions (--bg) let you run long tasks without tying up a terminal.

Sources: CLI Reference · Commands · Skills · Settings