Claude Code Cheat Sheet

Keyboard Shortcuts

General Controls

Shortcut	Action
Ctrl+C	Cancel input/generation
Ctrl+D	Exit session
Ctrl+L	Clear screen
Ctrl+O	Toggle verbose output
Ctrl+R	Reverse search history
Ctrl+G	Open prompt in editor
Ctrl+B	Background running task
Ctrl+T	Toggle task list
Ctrl+S	Stash prompt (run slash command mid-input, auto-restores)
Ctrl+V	Paste image
Ctrl+F	Kill background agents (×2)
Ctrl+_	Undo typing (readline)
Esc Esc	Rewind / undo

Mode Switching

Shortcut	Action
⇧Tab	Cycle permission modes
⌥P	Switch model
⌥T	Toggle thinking

Input

Shortcut	Action
\ + Enter	Newline (quick)
Ctrl+J	Newline (control seq)

Prefixes

Prefix	Action
/	Slash command
!	Direct bash
@	File mention + autocomplete

Session Picker (via /resume)

Key	Action
↑↓	Navigate
←→	Expand/collapse
P	Preview
R	Rename
/	Search
A	All projects
B	Current branch

MCP Servers

Add Servers

Flag	Transport
--transport stdio	Local process
--transport sse	Remote SSE

Scopes

File	Scope
.claude.json	Local (per project)
.mcp.json	Project (shared/VCS)
~/.claude.json	User (global)

Manage

Command	Action
/mcp	Interactive UI
claude mcp list	List all servers
claude mcp serve	CC as MCP server

Elicitation: servers can request input mid-task (NEW)

Slash Commands

Session

Command	Action
/clear	Clear conversation
/compact [focus]	Compact context
/resume	Resume/switch session
/rename [name]	Name current session
/branch [name]	Branch conversation (/fork alias)
/cost	Token usage stats (API key billing only)
/context	Visualize context (grid)
/diff	Interactive diff viewer
/copy	Copy last response
/export	Export conversation

Config

Command	Action
/config	Open settings
/model [model]	Switch model (←→ effort)
/fast [on|off]	Toggle fast mode
/vim	Toggle vim mode
/theme	Change color theme
/permissions	View/update permissions
/effort [level]	Set effort (low/med/high/max/auto) (NEW)
/color [color]	Set prompt-bar color
/keybindings	Customize keyboard shortcuts
/terminal-setup	Configure terminal keybindings

Tools

Command	Action
/init	Create CLAUDE.md
/memory	Edit CLAUDE.md files
/mcp	Manage MCP servers
/hooks	Manage hooks
/skills	List available skills
/agents	Manage agents
/chrome	Chrome integration
/reload-plugins	Hot-reload plugins
/add-dir <path>	Add working directory

Special

Command	Action
/btw <question>	Side question (no context)
/plan [desc]	Plan mode (+ auto-start)
/loop [interval]	Schedule recurring task
/voice	Push-to-talk voice (20 langs)
/doctor	Diagnose installation
/stats	Usage streaks & prefs
/insights	Analyze sessions report (highly recommended)
/desktop	Continue in Desktop app
/remote-control	Bridge to claude.ai/code (/rc) (NEW)
/usage	Plan limits & rate status
/schedule	Cloud scheduled tasks
/security-review	Security analysis of changes
/help	Show help + commands
/feedback	Submit feedback (alias: /bug)
/release-notes	View full changelog
/stickers	Order stickers!

CLAUDE.md

Locations

Path	Scope
./CLAUDE.md	Project (team-shared)
~/.claude/CLAUDE.md	Personal (all projects)
/etc/claude-code/	Managed (org-wide)
.claude/rules/*.md	Project rules
~/.claude/rules/*.md	User rules
paths: frontmatter	Path-specific rules
@path/to/file	Import in CLAUDE.md

Auto Memory

Path	Description
~/.claude/projects/<proj>/memory/	Per-project memory directory
MEMORY.md + topic files	Auto-loaded into context

Plan Mode

Item	Description
⇧Tab	Cycle: Normal → Auto-Accept → Plan
--permission-mode plan	Start in plan mode
⌥T	Toggle thinking on/off
"ultrathink"	Max effort for turn
Ctrl+O	See thinking (verbose)
/effort	○ low · ◐ med · ● high (NEW)

Git Worktrees

Item	Description
--worktree name	Isolated branch per feature
isolation: worktree	Agent in own worktree
sparsePaths	Checkout only needed dirs (NEW)
/batch	Auto-creates worktrees

Voice Mode

Item	Description
/voice	Enable push-to-talk
Space (hold)	Record, release to send
20 languages	EN, ES, FR, DE, CZ, PL…

Context Management

Item	Description
/context	Usage + optimization tips
/compact [focus]	Compress with focus
Auto-compact	Triggers at ~95% capacity
1M context	Opus 4.6 (Max/Team/Ent)
CLAUDE.md	Survives compaction
claude -c	Continue last conversation
claude -r "name"	Resume by name
/btw question	Side question, no context cost
claude -p "query"	Non-interactive
--output-format json	Structured output
--max-budget-usd 5	Cost cap
cat file | claude -p	Pipe input
/loop 5m msg	Recurring task
/rc	Remote control
--remote	Web session on claude.ai

Config Files

File	Purpose
~/.claude/settings.json	User settings
.claude/settings.json	Project (shared)
.claude/settings.local.json	Local only
~/.claude.json	OAuth, MCP, state
.mcp.json	Project MCP servers

Key Settings

Setting	Description
modelOverrides	Map model picker → custom IDs
autoMemoryDirectory	Custom memory dir
worktree.sparsePaths	Sparse checkout dirs (NEW)

Environment Variables

Full reference — ~80+ variables. Most useful ones:

Core / Auth

Variable	Description
ANTHROPIC_API_KEY	API key (overrides subscription when set)
ANTHROPIC_AUTH_TOKEN	Custom Authorization header value
ANTHROPIC_BASE_URL	Proxy/gateway endpoint
ANTHROPIC_MODEL	Default model
ANTHROPIC_CUSTOM_MODEL_OPTION	Custom /model picker entry

Cloud Providers

Variable	Description
CLAUDE_CODE_USE_BEDROCK	Use Amazon Bedrock
CLAUDE_CODE_USE_VERTEX	Use Google Vertex AI
CLAUDE_CODE_USE_FOUNDRY	Use Microsoft Foundry

Model / Thinking

Variable	Description
CLAUDE_CODE_EFFORT_LEVEL	low/med/high/max/auto
MAX_THINKING_TOKENS	0 = off
CLAUDE_CODE_SUBAGENT_MODEL	Override model for subagents
CLAUDE_CODE_DISABLE_ADAPTIVE_THINKING	=1 to fall back to fixed thinking budget

Context / Behavior

Variable	Description
CLAUDE_AUTOCOMPACT_PCT_OVERRIDE	Auto-compact threshold (1–100, default ~95)
CLAUDE_CODE_AUTO_COMPACT_WINDOW	Override context window size for compaction
BASH_DEFAULT_TIMEOUT_MS	Default bash command timeout
CLAUDE_CODE_SHELL	Override shell detection (bash/zsh)
CLAUDE_ENV_FILE	Script sourced before each bash command (virtualenv persistence)
CLAUDE_CODE_PLUGIN_SEED_DIR	Plugin seed dirs (: separated)

Privacy / Enterprise

Variable	Description
DISABLE_TELEMETRY	=1 to opt out of Statsig telemetry
DISABLE_AUTOUPDATER	=1 to disable auto-updates
CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC	Disables telemetry, updates, error reporting, feedback
CLAUDE_CODE_DISABLE_AUTO_MEMORY	=1 to disable auto memory

Runtime Detection

Variable	Description
CLAUDECODE	=1 in CC-spawned shells
IS_DEMO	Demo mode (hide email/org)
IS_SANDBOX	=1 to allow running as root (needed for --dangerously-skip-permissions in VMs)

Skills

Built-in Skills

Command	Description
/simplify	Code review (3 parallel agents)
/batch	Large parallel changes (5–30 worktrees)
/debug [desc]	Troubleshoot from debug log
/loop [interval]	Recurring scheduled task
/claude-api	Load API + SDK reference

Skill Locations

Path	Scope
.claude/skills/<name>/	Project skills
~/.claude/skills/<name>/	Personal skills

Skill Frontmatter

Field	Description
description	Auto-invocation trigger
allowed-tools	Skip permission prompts
model	Override model for skill
effort	Override effort level (NEW)
context: fork	Run in subagent
$ARGUMENTS	User input placeholder
${CLAUDE_SKILL_DIR}	Skill’s own directory
!`cmd`	Dynamic context injection

Agents

Built-in Agents

Agent	Description
Explore	Fast read-only (Haiku)
Plan	Research for plan mode
General	Full tools, complex tasks
Bash	Terminal, separate context

Agent Frontmatter

Field	Description
permissionMode	default/acceptEdits/plan/dontAsk/bypass
isolation: worktree	Run in git worktree
memory: user|project	Persistent memory
background: true	Background task
maxTurns	Limit agentic turns
SendMessage	Resume agents (replaces resume) (NEW)

CLI Reference

Core Commands

Command	Action
claude	Interactive
claude "q"	With prompt
claude -p "q"	Headless
claude -c	Continue last
claude -r "n"	Resume
claude update	Update

Key Flags

Flag	Description
--model	Set model
-w	Git worktree
-n / --name	Session name
--add-dir	Add dir
--agent	Use agent
--allowedTools	Pre-approve
--output-format	json/stream
--json-schema	Structured
--max-turns	Limit turns
--max-budget-usd	Cost cap
--console	Auth via Anthropic Console
--verbose	Verbose
--bare	Minimal headless (no hooks/LSP) (NEW)
--channels	Permission relay / MCP push (NEW)
--remote	Web session
--effort	low/med/high/max
--permission-mode	plan/default/…
--dangerously-skip-permissions	Skip all prompts ⚠️
--chrome	Chrome