Verified
New since your last visit
Unverified

Keys & Shortcuts

General Controls

ItemDescription
#Ctrl C / DCancel input / Exit session
#Ctrl LClear screen + full redraw; in fullscreen mode, clears the transcript view like a terminal clear — scroll up to see earlier messages
#Ctrl OToggle between normal and verbose transcript; in a collapsed thinking group, shows the full thinking; focus view toggled with /focus
#Ctrl RHistory picker — searches all projects by default; press Ctrl S to narrow to current project or session
#Ctrl GOpen prompt in external editor
#Ctrl BBackground running task
#Ctrl TToggle task list; in claude agents: pin/unpin background session (pinned sessions stay alive when idle, restart in place for CC updates)
#Ctrl VPaste image from clipboard
#Ctrl Enter / Ctrl X Ctrl SSend-now — interrupt the current turn and send all queued messages at once
#Ctrl BackspaceDelete previous word (Windows)
#Ctrl UClear entire input buffer (Ctrl+Y to restore)
#Ctrl YRestore input buffer cleared by Ctrl+U
#Ctrl X KKill background agents (chord, x2)
#Ctrl XIn claude agents: permanently removes a completed session; deleted background jobs stay deleted
#Esc EscRewind / undo state
#EscClose background-tasks, diff, and workflow detail views (left arrow no longer closes these)
#selection:clearKeybinding action — bind to a key in /keybindings to clear an in-app text selection; works in agents view

Mode & Input

ItemDescription
#TabCycle permission modes
#\ PSwitch model
#\ TToggle thinking on/off
#\ EnterNewline (quick)
#Ctrl JNewline (control seq)
#Ctrl A / Ctrl EMove to start/end of current logical line (multiline input)

Session Picker Nav

ItemDescription
#P / RPreview / Rename session
#/Search sessions
#A / BFilter All projects / Current branch
#Ctrl AShow all projects in /resume picker (picker defaults to current dir)

Transcript View

ItemDescription
#[Dump transcript to scrollback
#vOpen transcript in editor
#?Show keyboard shortcuts in transcript nav view
#{ / }Jump between user prompts in transcript nav view
#v (nav view)Toggle shortcut panel in transcript nav view

Vim Mode

ItemDescription
#v / VVisual / visual-line selection mode (NORMAL mode)
#/ (NORMAL mode)Open reverse history search (like Ctrl+R), matching bash/zsh vi-mode
#s / SSubstitute char / substitute line in NORMAL mode, matching vim behavior

Slash Commands (Core)

Session Management

ItemDescription
#/clearClear conversation context; resets the per-session subagent spawn budget
#/compact [focus]Compact context to free space
#/resumeOpen the session picker; background sessions are listed alongside interactive ones
#/rename [name]Name current session
#/branch [name]Branch conversation
#/forkCopy the conversation into a new background session with its own git worktree
#/subtaskLaunch an in-session subagent for the current task; formerly what /fork did before v2.1.212
#/usageUnified usage view — cost, stats, limits and loops in one place
#/usage-creditsRequest extra usage allowance
#/contextVisualize context window (grid)
#/diffInteractive diff viewer
#/copyCopy last response
#/exportExport conversation
#/recapSummarize away context; auto-triggered on return or invoke manually
#/rewind / /undoRewind conversation to previous state; /undo is alias; supports resuming a conversation from before /clear was run
#/goal [condition]Set a completion condition; Claude keeps working across turns until it is met
#/cdMove session to a new working directory without breaking the prompt cache mid-session

Configuration

ItemDescription
#/configOpen settings interface; search matches option values
#/config key=valueSet any setting from the prompt (e.g. /config thinking=false) — works in interactive, -p, and Remote Control
#/config --helpList all available shorthand keys for /config key=value
#/model [model]Switch model; the choice becomes the default for new sessions
#/effort [level]Set effort (low/med/high/xhigh/auto)
#/fast [on/off]Toggle Fast mode
#/permissionsView/update tool permissions
#/color [color]Set prompt-bar color; bare /color with no args picks a random session color
#/keybindingsCustomize keyboard shortcuts
#/terminal-setupConfigure terminal integration
#/themeChange color theme; create/switch named custom themes, hand-edit ~/.claude/themes/; plugins ship via themes/ dir; includes "Auto (match terminal)"
#/tui [fullscreen]Switch to flicker-free fullscreen TUI rendering in same conversation
#/focusToggle focus view (fullscreen/TUI mode)
#/release-notesInteractive version picker for release notes
#/scroll-speedTune mouse wheel scroll speed with a live preview
#/output-style [name] List and switch output styles; works over Remote Control and in cloud and headless sessions

Slash Commands (Tools)

Development Tools

ItemDescription
#/initCreate CLAUDE.md memory
#/memoryEdit CLAUDE.md files directly
#/mcpManage MCP servers; see per-server tool counts and connection errors
#/hooksManage event lifecycle hooks
#/install-github-appInstall the GitHub App; GitHub Actions workflow and repo secret setup is optional — install just the GitHub App and skip the workflow/secret steps
#/commit-push-prCommit, push, and open a PR in one step
#/skillsList available active skills; includes type-to-filter search box for long lists
#/chromeChrome browser integration; pick which connected browser to use via /chrome → "Select browser…", or in-chat when a browser action runs with multiple connected
#/add-dir <path>Add working directory to context; usable while Claude is working
#/reload-pluginsHot-reload local plugins; available in headless sessions (Claude Code Desktop and SDK); auto-installs missing dependencies from added marketplaces
#/reload-skillsRe-scan skill directories without restarting the session
#/plugin listList installed plugins; --enabled/--disabled filters; search bar available when browsing marketplace plugins; surfaces recently-unused plugins for cleanup
#/plugin install <name>Install a plugin from the marketplace

Special & Utility

ItemDescription
#/btw <query>Side question (does not alter context); press c to copy raw markdown answer to clipboard; Shift+←/Shift+→ (or [/]) to step through earlier answers
#/plan [desc]Switch to Plan Mode directly
#/loop [interval]Schedule recurring background task; self-paced dynamic mode and no-prompt autonomous default are now always available, including on Bedrock, Vertex, and Foundry
#/proactiveAlias for /loop — schedule recurring background task
#/scheduleCloud scheduled tasks
#/deep-researchStructured deep-research session; must be explicitly invoked — Claude no longer auto-starts it
#/workflowsView and manage dynamic workflow runs
#ultracodePrompt trigger keyword — type in the prompt to start a dynamic workflow; highlighted in violet in the prompt input
#/voicePush-to-talk voice prompting
#/doctorFull setup checkup — diagnose and fix installation or environment issues
#/skill-doctorShow which loaded skills go unused and what they cost in context — helps prune skills that aren't contributing
#/statusShow environment status, warnings and what kind of session this is
#/pr-comments [PR]Fetch GitHub PR comments
#/security-reviewSecurity analysis of uncommitted code
#/powerupInteractive lessons teaching CC features with animated demos
#/stats / /insightsShortcut to /usage stats tab / HTML sessions report
#/desktopContinue session in Desktop app
#/remote-controlBridge session to claude.ai/code
#/team-onboardingGenerate teammate ramp-up guide from local CC usage
#/advisorAI advisor mode

CLI & Subcommands

Core Execution

ItemDescription
#claudeStart interactive REPL
#claude -p "q"Headless execution (scripting)
#claude -c / -r "n"Continue last / Resume specific
#claude --from-prResume session linked to a PR; accepts GitHub, GitLab MR, Bitbucket PR, and GitHub Enterprise URLs
#claude --fork-sessionFork existing session on resume
#claude --teleport <session-id>Continue a cloud session locally; session ID shown in the /teleport hint that appears in cloud sessions
#claude ultrareview [target]Run /ultrareview non-interactively from CI or scripts; prints findings to stdout; use --json for raw output
#claude project purge [path]Delete all Claude Code state for a project (transcripts, tasks, file history, config entry); supports --dry-run, -y/--yes, -i/--interactive, and --all
#claude auto-mode resetRestore default auto-mode configuration; prompts for confirmation (pass --yes to skip)
#claude agentsAgent view (Research Preview) — every CC session in one list: running, blocked on you, or done
#claude --bg --exec '<cmd>'Run a shell command as a background session you can attach to and detach from; equivalent to typing ! <cmd> inside claude agents
#claude self-hosted-runnerTurn your own machine or container into a session host for Claude Code web, mobile, and desktop sessions (Team and Enterprise plans)
#claude self-hosted-runner --defer-shutdown-max-min <minutes>On SIGTERM, keep serving attached sessions for up to the given number of minutes before parking remaining sessions and exiting
#claude self-hosted-runner --remove-session-state Delete per-session directories under <base-dir>/_sessions/ when sessions end; off by default
#claude self-hosted-runner --proxy-authorization-command / --proxy-authorization-fileSupply a freshly issued Proxy-Authorization header on every connection for egress proxies that require per-request credentials
#claude remote-control --continueResume the most recent Remote Control session
#claude self-hosted-runner --client-label <label>Override the label the runner registers with (default: hostname); paired with SELF_HOSTED_RUNNER_CLIENT_LABEL env var
#claude self-hosted-runner --drain-marker-file <path>If this file exists at a SIGTERM drain, the runner reports its exit to the server as a host drain (telemetry only)
#claude self-hosted-runner --host-config-snapshot disk|memoryChoose where host config (settings, skills, plugins, MCP servers) is cached; avoids silently losing it once the config directory exceeds 64 MiB

Headless & CI Flags

ItemDescription
#--restrictedRemove built-in tools that run commands or code and WebFetch (unless named in --tools)
#--bareMinimal headless (no hooks/LSP)
#--permission-prompts noneDeny any action that would prompt for unattended headless hosts; the active permission mode (including auto mode) still decides normally
#--append-subagent-system-prompt-fileRead the subagent system prompt from a file; for prompts too large to pass on the command line
#--channelsPermission relay / MCP push msg; now works with console (API key) auth — console orgs with managed settings must set channelsEnabled: true
#--output-format jsonStructured JSON output
#--json-schemaStrict JSON schema validation
#--dangerously-skip-permissionsSkip all approval prompts
#--exclude-dynamic-system-prompt-sectionsExclude dynamic prompt sections in print mode for cross-user prompt caching
#--forward-subagent-textInclude subagent text and thinking in --output-format stream-json output; paired with CLAUDE_CODE_FORWARD_SUBAGENT_TEXT env var
#--system-prompt-snapshot off Renders the system prompt fresh on every request instead of reusing the conversation's recorded prompt; for iterating on prompt text

State & Behavior

ItemDescription
#-w / --worktreeStart in isolated git worktree
#--effortSet effort level (low/med/high/xhigh); xhigh is Opus 4.8 only
#--modelOverride specific model at launch
#--fallback-modelFallback model for when the primary model is overloaded or unavailable
#--thinking disabledDisable thinking on models that think by default; equivalent to MAX_THINKING_TOKENS=0 or toggling the per-model thinking switch
#--permission-mode planStart in read-only plan mode
#--permission-mode manualStart in manual (default) permission mode; renamed from default — both values accepted
#--max-budget-usdSet cost cap for session
#--remoteStart web-accessible session
#--remote-controlEnable Remote Control mode; bridges CLI session to claude.ai/code
#--remote-control-session-name-prefixCustom hostname prefix for Remote Control session names
#--plugin-dir <path>Load plugins from a directory, folder of plugins, or .zip archive at launch
#--plugin-url <url>Fetch a plugin .zip archive from a URL for the current session
#--safe-modeStart with all customizations (CLAUDE.md, plugins, skills, hooks, MCP servers) disabled for troubleshooting; paired with CLAUDE_CODE_SAFE_MODE env var
#--ax-screen-readerEnable screen reader mode — opt-in plain-text rendering for screen reader users

Settings (JSON)

Files & Priority

ItemDescription
#~/.claude/settings.jsonUser global defaults
#.claude/settings.jsonProject shared (committed)
#.claude/settings.local.jsonProject local only (gitignored)
#~/.claude.jsonOAuth, MCP, state configuration
#.mcp.jsonProject-specific MCP servers
#managed-settings.d/Drop-in policy fragments (alphabetical merge)

Key JSON Settings

ItemDescription
#modelOverridesMap model picker to custom IDs
#modelPickerCurate the /model picker with an ordered, labeled list of models
#modelPricingManaged setting — set an organization's contracted per-model rates and discount multiplier
#promptCacheTtl / subagentPromptCacheTtlSet the prompt cache TTL for the conversation and for subagents separately
#permissions.defaultModeSet the permission mode new sessions start in
#autoMemoryDirectoryCustom auto-loaded memory dir
#worktree.sparsePathsSparse checkout dirs for agents
#worktree.baseReffresh (default) branches worktrees from origin/<default>; head branches from local HEAD — applies to --worktree, EnterWorktree, and agent-isolation
#worktree.bgIsolationSet to "none" to disable worktree isolation for background sessions
#agentDefault agent name for dispatched sessions in claude agents; override per-session with --agent <name>
#fallbackModelConfigure up to three fallback models tried in order when the primary model is overloaded or unavailable; paired with --fallback-model CLI flag
#showThinkingSummariesShow thinking summaries in interactive sessions (default: false)
#languagePin session-title language; titles auto-detect the conversation language by default — set to a BCP 47 tag (e.g. en, ja) to override
#disableSkillShellExecutionDisable inline shell execution in skills and custom slash commands
#skillOverridesoff hides skill from model and /; user-invocable-only hides from model only; name-only collapses description
#statusLine.refreshIntervalRe-run status line command every N seconds
#workspace.git_worktreeStatus line JSON input field — set when current directory is inside a linked git worktree
#effort.level / thinking.enabledHooks and status line stdin JSON input fields — current effort level and thinking state
#prompt_cacheStatus line script JSON input field — per-session prompt-cache stats (hit ratio, misses, tokens re-cached, warm/cold)
#autoScrollEnabledDisable conversation auto-scroll in fullscreen mode
#wheelScrollAccelerationEnabledSet to false to disable mouse-wheel scroll acceleration in fullscreen mode
#tuiDefault TUI mode; set to "fullscreen" for flicker-free alt-screen rendering
#axScreenReaderSet to true for opt-in plain-text rendering for screen reader users; paired with --ax-screen-reader flag and CLAUDE_AX_SCREEN_READER=1 env var
#spellcheckUnderline misspelled words in the prompt input as you type; requires aspell, hunspell, or ispell to be installed
#keybindingFlavorSet to "readline" to make Ctrl+W delete back to the previous whitespace (Bash-style)
#outputStyleBuilt-in output style
#timeFormatTurn-end clock and transcript timestamp format: "12h", "24h", "24h-utc", or a strftime pattern; paired with timeZone
#timeZoneTime zone for the turn-end clock and transcript timestamps; paired with timeFormat
#vimInsertModeRemapsMap two-key insert-mode sequences to keys in vim mode; e.g. map jj to Escape
#permissions.blockReadsOutsideWorkingDirectoriesBlock file reads outside configured working directories; in auto mode, a one-time prompt appears before the first such read
#sandbox.network.deniedDomainsBlock specific domains even when a broader allowedDomains wildcard would otherwise permit them
#sandbox.network.strictAllowlistDeny non-allowlisted hosts for sandboxed commands without prompting
#sandbox.allowAppleEventsOpt-in macOS setting to allow sandboxed commands to send Apple Events
#sandbox.credentialsBlock sandboxed commands from reading credential files and secret environment variables
#sandbox.filesystem.disabledSkip filesystem isolation while keeping network egress control
#"$defaults" (autoMode)Add custom auto-mode rules alongside the built-in list
#autoMode.hard_denyAuto mode classifier rules that block unconditionally regardless of user intent — unlike autoMode.soft_deny, cannot be overridden by allow exceptions
#autoMode.classifyAllShellRoute all Bash/PowerShell commands through the auto-mode classifier instead of only arbitrary-code-execution patterns
#allowed_domainsPer-command field on Bash/PowerShell/Monitor in sandboxed auto mode; scopes that command's sandbox network access to just the domains it needs
#disableAutoModeSet to true to disable auto mode on Bedrock, Vertex AI, and Foundry — auto mode is now available there by default without opt-in
#maxEffortLevel Caps the maximum effort level on every provider; users can still pick a lower level
#"*" (deny rule glob)Glob patterns supported in deny rule tool-name position
#Tool(param:value)Permission rule syntax to match a tool's input parameters (supports * wildcard); e.g. Agent(model:opus) to block Opus subagents
#Write(path) / NotebookEdit(path) / Glob(path)Deprecated permission-rule tool names — startup warns to use Edit(path) or Read(path) instead
#attribution.sessionUrlSet to false to omit the claude.ai session link from commits and PRs in web and Remote Control sessions
#prUrlTemplatePoint footer PR badge at a custom code-review URL instead of github.com
#footerLinksRegexesArray of regex patterns for regex-matched link badges in the footer row; configurable via user or managed settings
#extraKnownMarketplacesRegister plugin marketplaces for a repository or an organization
#disableBundledSkillsHide bundled skills, workflows, and built-in slash commands from the model; paired with CLAUDE_CODE_DISABLE_BUNDLED_SKILLS env var
#syncClaudeAiSkills / syncClaudeAiPluginsSet to false to stop syncing the skills or plugins enabled on your claude.ai account into signed-in terminal sessions
#teammateMode: "iterm2"Use iTerm2 as the teammate backend; shows a warning when auto mode cannot find the it2 CLI
#awsAuthRefreshConfigures the "Claude Platform on AWS — refresh credentials" option in /login
#respondToBashCommandsSet to false to keep ! bash command output as context-only; default behavior is for Claude to respond automatically
#workflowSizeGuidelineSet the agent count Claude aims for in dynamic workflows
#emojiCompletionEnabledSet to false to disable emoji shortcode autocomplete in the prompt input (e.g. :heart: → ❤️)
#crossSessionInboundCross-session messages sent to a session running with bypassed permissions are held for your approval before delivery
#dialogExpiryTimeout for inbound cross-session message approval dialogs; messages to sessions without bypassed permissions auto-deliver
#feedbackDraftsSet to false to stop Claude auto-drafting a feedback report
#bashOutputMaxChars / taskOutputMaxCharsRaise how much Bash command output Claude receives inline before it is saved to a file
#bashEditDiffEnabled Show a diff of files a Bash command changed in the tool result when the Bash tool handles file edits

Managed & Enterprise

ItemDescription
#forceRemoteSettingsRefreshBlock startup until remote managed settings are freshly fetched; fail-closed on fetch failure
#sandbox.bwrapPath / sandbox.socatPathCustom bubblewrap and socat binary paths (Linux/WSL managed settings)
#wslInheritsWindowsSettingsWSL sessions inherit Windows-side managed settings (policy key)
#parentSettingsBehaviorAdmin-tier key: first-wins (default) or merge — opts SDK managedSettings parent tier into policy merge
#channelsEnabledSet to true in managed settings to allow --channels with console (API key) auth
#allowAllClaudeAiMcpsEnterprise managed setting — load claude.ai cloud MCP connectors alongside managed-mcp.json
#managedMcpServersManaged setting — provide HTTP/SSE MCP servers to every user in the organization (same entry shape as .mcp.json); entries naming a command to run are skipped
#deniedMcpServersManaged setting — blocklist of managed MCP servers to deny
#pluginSuggestionMarketplacesManaged setting — allowlist org marketplaces whose plugins may be suggested via context-aware tips
#strictKnownMarketplacesManaged setting — allowlist of marketplace repos allowed as plugin sources
#blockedMarketplacesManaged setting — blocklist of marketplace repos denied as plugin sources; supports "owner/*" wildcard to block all repos under a GitHub org
#requiredMinimumVersion / requiredMaximumVersionManaged settings — CC refuses to start if its version is outside the allowed range; directs user to an approved version
#enforceAvailableModelsManaged setting — the availableModels allowlist also constrains the Default model
#forceLoginGatewayUrlManaged setting — gateway URL; machines with this set start as gateway sessions from startup
#gatewayInternalNetworks Managed setting — allow /login to a Claude apps gateway on the organization's own public IPv4 block

Environment Variables

All Variables

ItemDescription
#ANTHROPIC_API_KEYAuth override
#ANTHROPIC_MODELDefault model override
#CLAUDE_CODE_EFFORT_LEVELDefault effort (low/med/high)
#MAX_THINKING_TOKENSCap extended thinking limits; set to 0 to disable thinking on models that think by default via the Claude API
#CLAUDE_CODE_PLUGIN_SEED_DIRMultiple plugin seed directories
#IS_DEMODemo mode (hides email/org)
#CLAUDECODEDetects if running inside CC shell; also injected into stdio MCP server subprocess environments
#CLAUDE_CODE_SUBPROCESS_ENV_SCRUBStrip credentials from subprocesses; enables PID namespace isolation on Linux (sandboxing)
#CLAUDE_CODE_MCP_SERVER_NAME / CLAUDE_CODE_MCP_SERVER_URLInject server identity into MCP headerHelper scripts
#CLAUDE_CODE_NO_FLICKER=1Opt into flicker-free alt-screen rendering with virtualized scrollback
#MCP_CONNECTION_NONBLOCKING=trueSkip MCP connection wait in headless (-p) mode
#CLAUDE_CODE_PLUGIN_KEEP_MARKETPLACE_ON_FAILUREKeep marketplace cache when git pull fails (offline envs)
#CLAUDE_CODE_USE_MANTLE=1Enable Amazon Bedrock powered by Mantle
#CLAUDE_CODE_PERFORCE_MODEWhen set, Edit/Write/NotebookEdit fail on read-only files with a p4 edit hint
#CLAUDE_CODE_SCRIPT_CAPSLimit per-session script invocations (subprocess sandboxing)
#CLAUDE_CODE_CERT_STORE=bundledUse only bundled CAs; by default OS CA cert store is trusted (enterprise TLS proxy support)
#OTEL_LOG_USER_PROMPTS / OTEL_LOG_TOOL_DETAILS / OTEL_LOG_TOOL_CONTENTOTEL beta tracing: opt in to emit sensitive span attributes (prompts/tool details/tool content)
#OTEL_LOG_ASSISTANT_RESPONSESOpt in to emitting the model's response text as an OTEL log event
#ENABLE_PROMPT_CACHING_1HOpt into 1-hour prompt cache TTL (API key, Bedrock, Vertex, Foundry)
#FORCE_PROMPT_CACHING_5MForce 5-minute prompt cache TTL
#CLAUDE_CODE_ENABLE_AWAY_SUMMARYForce-enable recap/away summary (useful when telemetry disabled)
#CLAUDE_CODE_USE_POWERSHELL_TOOLEnable PowerShell tool; on Windows opt in/out, on Linux/macOS requires pwsh on PATH
#OTEL_LOG_RAW_API_BODIESEmit full API request/response bodies as OTel log events for debugging
#CLAUDE_CODE_OTEL_CONTENT_MAX_LENGTHConfigure the 60 KB truncation limit on OpenTelemetry content attributes
#TRACEPARENT / TRACESTATEDistributed trace context injected from environment for SDK/headless sessions
#CLAUDE_CODE_FORK_SUBAGENT=1Enable forked subagents on external (non-Anthropic) builds; also works in non-interactive sessions
#DISABLE_UPDATESBlock all update paths including manual claude update; stricter than DISABLE_AUTOUPDATER
#CLAUDE_CODE_HIDE_CWDHide working directory in the startup logo
#ANTHROPIC_BEDROCK_SERVICE_TIERSelect a Bedrock service tier (default, flex, or priority); sent as the X-Amzn-Bedrock-Service-Tier header
#ANTHROPIC_BEDROCK_REGION_PREFIXPrefer a specific Bedrock cross-region inference profile over the AWS_REGION-derived one
#CLAUDE_CODE_FORCE_SYNC_OUTPUT=1Force-enable synchronized output on terminals that auto-detection misses (e.g. Emacs eat)
#CLAUDE_CODE_PACKAGE_MANAGER_AUTO_UPDATEOn Homebrew or WinGet installs, runs the upgrade command in the background and prompts to restart when set
#CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY=1Opt in to listing gateway /v1/models in the /model picker when using a custom ANTHROPIC_BASE_URL gateway (was automatic in v2.1.126–v2.1.128)
#CLAUDE_CODE_GATEWAY_MODEL_DISCOVERY_TIMEOUT_MS Extend the LLM gateway /v1/models discovery timeout (default 3 s)
#CLAUDE_CODE_GATEWAY_HINT_HEADERS=1Opt in to sending request-class, agent-type and compaction hint headers to LLM gateways
#CLAUDE_CODE_DISABLE_1M_CONTEXTHold every Claude model with a native 1M context window to 200K via auto-compaction
#CLAUDE_CODE_DISABLE_UNKNOWN_MODEL_WINDOW_ENFORCEMENT=1Restore the old behaviour for sessions on unrecognized model IDs
#CLAUDE_CODE_SESSION_IDSession ID injected into Bash tool subprocess environment and stdio MCP server subprocess environments, matching the session_id passed to hooks
#CLAUDE_CODE_DISABLE_ALTERNATE_SCREEN=1Opt out of the fullscreen alternate-screen renderer; keeps conversation in the terminal's native scrollback
#CLAUDE_CODE_ENABLE_FEEDBACK_SURVEY_FOR_OTELRe-enable the session quality survey for enterprises capturing responses through OpenTelemetry
#CLAUDE_PROJECT_DIRInjected into MCP stdio server environments; referenceable as ${CLAUDE_PROJECT_DIR} in plugin configs
#CLAUDE_CODE_PLUGIN_PREFER_HTTPSClone GitHub plugin sources over HTTPS instead of SSH; for environments without a GitHub SSH key
#ANTHROPIC_WORKSPACE_IDWorkload identity federation — scopes the minted token to a specific workspace when the federation rule covers more than one
#OTEL_METRICS_INCLUDE_ENTRYPOINT=trueOpt in to including session entrypoint as an OpenTelemetry metric attribute (app.entrypoint)
#OTEL_METRICS_INCLUDE_REPOSITORY Tag OpenTelemetry metrics and events with vcs.* repository attributes
#COLUMNS / LINESInjected into status line command subprocesses so scripts can size output to the terminal width
#CLAUDE_CODE_ENABLE_AUTO_MODE=1No longer needed — auto mode is available on Bedrock, Vertex and Foundry by default
#OTEL_RESOURCE_ATTRIBUTESStandard OTEL env var — values are now used as labels on metric datapoints, so you can slice usage metrics by custom dimensions (e.g. team, repo)
#CLAUDE_CODE_SAFE_MODEStart with all customizations (CLAUDE.md, plugins, skills, hooks, MCP servers) disabled; paired with --safe-mode CLI flag
#CLAUDE_CODE_DISABLE_BUNDLED_SKILLSHide bundled skills, workflows, and built-in slash commands from the model; paired with disableBundledSkills JSON setting
#API_FORCE_IDLE_TIMEOUT=0Opt out of the default 5-minute idle timeout on Vertex/Foundry; stalled streams will hang indefinitely instead of aborting
#CLAUDE_CLIENT_PRESENCE_FILEPath to a marker file — while the file exists, mobile push notifications are suppressed (signals you're at the machine)
#CLAUDE_CODE_MAX_RETRIESRetry attempt limit; cap of 15 is lifted when CLAUDE_CODE_RETRY_WATCHDOG is set — for unattended sessions use the watchdog instead
#CLAUDE_CODE_RETRY_WATCHDOGRetry watchdog for unattended sessions
#CLAUDE_CODE_MCP_TOOL_IDLE_TIMEOUTOverride the default 5-minute idle timeout for remote MCP tool calls; stalled calls abort with an error instead of blocking indefinitely
#CLAUDE_CODE_DISABLE_BG_SHELL_PRESSURE_REAP=1Disable automatic memory-pressure reaping of idle background shell commands
#CLAUDE_CODE_DISABLE_MOUSE_CLICKSDisable mouse click/drag/hover in fullscreen mode while keeping wheel scroll active
#CLAUDE_AX_SCREEN_READER=1Enable screen reader mode — opt-in plain-text rendering; paired with --ax-screen-reader flag and axScreenReader JSON setting
#CLAUDE_CODE_PROCESS_WRAPPERCorporate launcher wrapper — agent view and background service run every Claude Code self-spawn through the specified executable
#CLAUDE_CODE_FORWARD_SUBAGENT_TEXTInclude subagent text and thinking in stream-json output; paired with --forward-subagent-text flag
#CLAUDE_CODE_MAX_WEB_SEARCHES_PER_SESSIONSession-wide limit on WebSearch tool calls (default 200); prevents runaway search loops
#CLAUDE_CODE_SUBAGENT_MODEL_FORCEApply CLAUDE_CODE_SUBAGENT_MODEL (or the main model when unset) to every subagent, ignoring per-spawn and agent-definition model overrides
#CLAUDE_CODE_MAX_SUBAGENTS_PER_SESSIONSet a per-session subagent spawn cap
#CLAUDE_CODE_MAX_CONCURRENT_SUBAGENTSCap on concurrently-running subagents (default 20); prevents a single message from fanning out unbounded background agents
#CLAUDE_CODE_MAX_SUBAGENT_SPAWN_DEPTHMax nesting depth for subagent spawning (default 3); set to 1 to disable nesting
#CLAUDE_CODE_MCP_AUTO_BACKGROUND_MSThreshold in ms after which long-running MCP tool calls move to background automatically (default 120000 / 2 min); set to 0 to disable
#CLAUDE_CODE_WORKFLOW_PREFIX_STAGGER_MS=0Set to 0 to disable automatic staggering of same-prefix sibling workflow agents; staggering is on by default so later agents reuse the cached prompt prefix
#CLAUDE_CODE_WORKFLOW_MAX_CONCURRENT_AGENTS Raise the Workflow tool's per-run concurrent agent limit (1–256) for inference-bound fan-outs
#CLAUDE_CODE_TOOL_MEMORY_LIMITOpt-in memory cgroup limit for Bash tool commands on Linux; prevents a runaway build from stalling the session
#CLAUDE_CODE_WEBFETCH_CACHE_TTL_MSConfigure the WebFetch session URL cache TTL in milliseconds (default: 15 minutes)
#CLAUDE_CODE_WEBFETCH_DEADLINE_MS Override the 300-second WebFetch response deadline; set to 0 to disable the deadline entirely
#CLAUDE_CODE_ENABLE_TODO_TOOLS=1Re-enable the task and todo tools on models where they are off by default
#CLAUDE_CODE_PROJECT_DIR_NAMECustom short name for the per-project transcript directory; for hosts that give each session its own config directory
#CLAUDE_CODE_GOAL_CHECKIN_MINUTESSet to 0 to disable automatic check-ins when a /goal is waiting on background tasks for 30+ min
#ANTHROPIC_DEFAULT_MODELSets the model new sessions start on; a /model pick still overrides it and persists across restarts — unlike ANTHROPIC_MODEL
#CLAUDE_CODE_RESTRICTED=1Lock the session down: no command execution, no WebFetch, no settings files
#SELF_HOSTED_RUNNER_CLIENT_LABELOverride the label the self-hosted runner registers with (default: hostname); paired with claude self-hosted-runner --client-label <label> CLI flag
#CLAUDE_CODE_AUTO_MODE_SERVER=0Opt out of the server-side auto mode classifier on Bedrock, Vertex, Foundry and gateways
#CLAUDE_CODE_MCP_STARTUP_WAIT_MSBound how long the first non-interactive turn waits for connecting MCP servers
#OTEL_LOG_MANAGED_SETTINGS=1Opt in to redacted settings and digests in the managed-settings OTel event
#MCP_SDK_GENERATION=v1 / MCP_PROTOCOL_NEGOTIATION=legacyOpt out of the new default MCP client and negotiation for direct HTTP MCP servers
#CLAUDE_GATEWAY_PROXY_IS_EGRESS_BOUNDARY=1For Claude apps gateways whose only egress is a forward proxy

Skills & Agent Frontmatter

Built-in Skills & Locations

ItemDescription
#/code-review [level] [pr#]Code review as a background subagent; /review is an alias
#/verifyVerify a code change works end-to-end; must be explicitly invoked — Claude no longer auto-runs it
#/simplifyCleanup-only review (reuse, simplification, efficiency, altitude) applied to working tree; does not run the full /code-review --fix bug hunt
#/batchLarge parallel changes (5-30 worktrees)
#/claude-apiLoad the API + SDK reference into context
#/less-permission-promptsScan transcripts for read-only Bash/MCP calls; propose settings.json allowlist
#/ultrareview [PR#]Cloud code review via parallel agents; no args for current branch, PR# for specific PR
#/datavizChart and dashboard design guidance with a runnable color-palette validator
#/skill-a /skill-b [prompt]Stacked slash-skill invocations — up to 5 leading skills all load simultaneously (e.g. /skill-a /skill-b do XYZ)
#.claude/skills/<name>/Project skill directory
#~/.claude/skills/<name>/Global personal skill directory
#plugin: bin/Plugin executables in bin/ invocable as bare commands from Bash tool
#command (plugin marketplace source)Specify a plugin directory via a local command's output (e.g. from an IDE); re-resolved each session without restart; mode: "link" uses it in place
#headersHelper (plugin marketplace / catalog entry)Command that mints HTTP headers for a url marketplace's fetches

Skill YAML Frontmatter

ItemDescription
#descriptionAuto-invocation trigger string
#effortOverride effort level specifically for skill
#allowed-toolsSkip permission prompts for skill
#disallowed-toolsRemove tools from the model while the skill or slash command is active
#disable-model-invocationPrevent Claude from auto-invoking the skill; Claude is told to ask you to run it manually instead
#context: forkRun skill in an isolated subagent; runs in the background by default — opt out per skill with background: false
#$ARGUMENTSUser input placeholder replacement
#\$Escape to include a literal $ before a digit in skill command bodies
#$FILE_PATHFile path resolution (hook strings only)
#${CLAUDE_SKILL_DIR}Reference the skill's own directory
#${CLAUDE_EFFORT}Current effort level — available in hooks, Bash tool commands, and skill content
#! 'cmd'Dynamic bash execution context injection
#keep-coding-instructionsControl plugin output style for coding tasks
#monitors (plugin manifest)Background monitor support for plugins; auto-arms at session start or on skill invoke
#defaultEnabled: false (plugin manifest)Plugin starts disabled by default; enable with /plugin or claude plugin enable; dependency plugins of enabled plugins are still enabled automatically
#true/false/yes/no/on/off/1/0Accepted values for boolean frontmatter fields in skills and plugins (case-insensitive); all resolve to true or false

Agent YAML Frontmatter

ItemDescription
#experimental.cacheTtlPer-agent prompt cache TTL; set to "5m" or "1h"; used when no subagent TTL setting is configured in the session
#permissionModemanual / acceptEdits / plan / bypass; manual replaces default — both accepted
#omitClaudeMdSkip user, project and local CLAUDE.md files for this agent; managed policy files still load. Also settable via --agents JSON
#isolation: worktreeRun agent in isolated git worktree
#memory: user|projectPersistent memory storage scope
#background: trueRun entirely as background task
#initialPromptAuto-submit first turn on spawn
#SendMessageMessage another agent or session; auto-resumes stopped agents
#hooks:Agent lifecycle hooks; fire when running as main-thread agent via --agent
#mcpServersMCP servers for the agent; now loaded for main-thread agent sessions via --agent

Plugin CLI Commands

ItemDescription
#claude plugin enable/disable <name>Enable or disable an installed plugin
#claude plugin init <name>Scaffold a new plugin in .claude/skills
#claude plugin tagCreate release git tags for plugins with version validation
#claude plugin pruneRemove orphaned auto-installed plugin dependencies; plugin uninstall --prune cascades
#claude plugin details <name>Show a plugin's component inventory and projected per-session token cost
#claude plugin validateValidate plugin manifest
#claude plugin eval Run a plugin's eval suite against Claude Code; outputs scored, reproducible results as JSON + HTML report

Memory, Hooks & MCP

Rules & Auto-Memory

ItemDescription
#.claude/rules/*.mdProject specific structural rules
#~/.claude/rules/*.mdUser specific global rules
#paths: (frontmatter)Path-specific rule triggers
#@path/to/fileImport file directly in CLAUDE.md
#.../memory/MEMORY.mdTopic files, automatically loaded
#AGENTS.mdRead in place of CLAUDE.md in a project that has no CLAUDE.md file

Event Hooks

ItemDescription
#PreToolUse / PostToolUseIntercept before/after tool execution
#PostToolUseFailureFires when a tool call fails; hook input includes duration_ms
#SessionStartFires on session start/resume
#PreModelSwitch / PostModelSwitchBlock, confirm, or annotate a model switch before or after it happens
#SubagentStart / Stop / SubagentStopLifecycle triggers for agent spawning
#SessionEndRun cleanup script when session closes
#CwdChanged / FileChangedReact to directory or file changes
#TaskCreatedFires when task created via TaskCreate
#DirectoryAddedFires after /add-dir or the SDK register_repo_root registers a new working directory mid-session
#if: (hook field)Conditional execution (permission rule syntax)
#"Bash,PowerShell" (hook matcher)Comma-separated tool type names in a hook's matcher field
#mcp__name__.* (hook matcher pattern)Hyphenated hook matcher identifiers exact-match — use mcp__name__.* glob to match all tools from a hyphenated MCP server (e.g. mcp__brave-search__.*)
#PermissionDeniedFires after auto mode classifier denials; return {retry:true} to allow retry
#"defer" (PreToolUse)Headless sessions pause at tool call; resume with -p --resume to re-evaluate
#UserPromptSubmitFires on user prompt submit; supports hookSpecificOutput.sessionTitle to set session title
#PreCompactFires before compaction; block by exiting with code 2 or returning {"decision":"block"}
#type: "mcp_tool" (hook)Hook directly invokes an MCP tool
#args: string[] (hook field)Exec form — spawns command directly without a shell; path placeholders never need quoting
#$CLAUDE_PLUGIN_OPTION_<KEY>Hook env var for accessing plugin config option values in exec-form hook commands
#continueOnBlock (PostToolUse)Set to true to feed the hook's rejection reason back to Claude and continue the turn
#terminalSequence (hook output)Hook JSON output field — emit desktop notifications, window titles, and bells without a controlling terminal
#MessageDisplayFires as assistant message text is displayed; hooks can transform or hide the output
#WorktreeCreate / WorktreeRemoveStandalone async events fired when a git worktree is created or removed
#NotificationFires when a background agent session needs input or finishes; events: agent_needs_input / agent_completed

MCP Servers

ItemDescription
#claude mcp list/add/removeStandard MCP server management
#claude mcp login/logout <name>Authenticate or de-authenticate an MCP server from the CLI without opening /mcp; --no-browser redirects auth code to stdin for SSH completion
#claude mcp serveRun Claude Code as an MCP server
#--transport http/stdio/sseProtocol connection methods
#ElicitationServers can request input mid-task
#_meta["anthropic/maxResultSizeChars"]Override max MCP result size per call (up to 500K chars)
#alwaysLoad (MCP server config)When true, all tools from that server skip tool-search deferral and are always available
#roots/listSession additional working directories (from /add-dir) are included in MCP roots/list; notifications/roots/list_changed sent to servers when the set changes

Auth Commands

ItemDescription
#claude auth login/logoutSession authentication control; auth login accepts OAuth code pasted in terminal when browser callback can't reach localhost (WSL2, SSH, containers)
#claude auth statusCheck current authentication validity
No matches — clear search to see everything.