Skip to content

Claude Code settings reference

Last updated:

Complete reference for settings.json configuration and environment variables. Covers confirmed settings through Claude Code v2.1.310 (last reviewed October 27, 2026). The official settings reference remains authoritative.

Sources: Official settings docs · Official env-vars docs · JSON Schema

Legend:

  • No badge = confirmed in official documentation
  • 📋 Schema only = present in JSON schema but not on official settings page (still valid)
  • ⚠️ Unverified = not confirmed in official sources

Claude Code uses four settings scopes, applied from highest to lowest priority:

PriorityScopeLocationShared?Purpose
1ManagedServer, MDM profile, registry, or system managed-settings.jsonYes (IT-deployed)Org-wide policies, cannot be overridden
2Command line-- flags at startupNoTemporary session overrides
3Local.claude/settings.local.jsonNo (gitignored)Personal project-specific
4Project.claude/settings.jsonYes (committed)Team-shared settings
5User~/.claude/settings.jsonNoGlobal personal defaults

Array merging: Settings like permissions.allow, sandbox.filesystem.allowWrite, and allowedHttpHookUrls are concatenated and deduplicated across scopes, not replaced.

Deny precedence: permissions.deny rules always take effect regardless of allow/ask rules at any scope.

Managed settings delivery methods:

  • Server-managed (Claude.ai admin console)
  • macOS MDM: com.anthropic.claudecode plist
  • Windows registry: HKLM\SOFTWARE\Policies\ClaudeCode
  • File: managed-settings.json at /Library/Application Support/ClaudeCode/ (macOS), /etc/claude-code/ (Linux/WSL), C:\Program Files\ClaudeCode\ (Windows)
  • Drop-in directory: managed-settings.d/*.json alongside managed-settings.json, merged alphabetically

Other config: ~/.claude.json stores OAuth session, MCP server configs, per-project trust state, and preferences like editorMode. Do not put ~/.claude.json keys into settings.json: it will trigger schema validation errors.


Type: string Scope: all Default: none

JSON Schema URL for IDE validation and autocomplete. Add "https://json.schemastore.org/claude-code-settings.json" to enable inline validation in VS Code, Cursor, and other editors.

{
"$schema": "https://json.schemastore.org/claude-code-settings.json"
}

Type: string Scope: all Default: "default"

Override the default model for all sessions. Accepts aliases ("sonnet", "opus", "haiku", "opusplan") or full model IDs like "claude-sonnet-4-6". The ANTHROPIC_MODEL environment variable takes precedence.

{ "model": "opus" }

Type: string Scope: all Default: none

Run the main thread as a named subagent. Applies that agent’s system prompt, tool restrictions, and model. Value must match an agent defined in .claude/agents/. Also available via --agent CLI flag.

{ "agent": "code-reviewer" }

Type: string Scope: all Default: "english"

Claude’s preferred response language. Also sets the voice dictation language. Examples: "japanese", "spanish", "french".

Type: number Scope: all Default: 30

Sessions inactive longer than this number of days are deleted at startup. Setting to 0 deletes all existing transcripts at startup and disables session persistence entirely: no .jsonl files are written, /resume shows no conversations, and hooks receive an empty transcript_path.

Type: string Scope: all Default: "latest" Values: "latest" | "stable"

Release channel to follow. "stable" is typically about one week behind "latest" and skips versions with major regressions.

Type: boolean Scope: all Default: false

Enable extended thinking by default for all sessions. Usually configured via /config rather than editing directly.

Type: boolean Scope: all Default: true

Include built-in commit and PR workflow instructions and a git status snapshot in the system prompt. Set to false when using a custom git workflow skill. The CLAUDE_CODE_DISABLE_GIT_INSTRUCTIONS env var takes precedence.

Type: boolean Scope: user Default: none

Enable push-to-talk voice dictation. Written automatically when you run /voice. Requires a Claude.ai account.

Type: array of strings Scope: all Default: none

Announcements displayed to users at startup. Multiple announcements are cycled through at random.

{
"companyAnnouncements": [
"Welcome to Acme Corp! Review code guidelines at docs.acme.com",
"All PRs require code review before merge"
]
}

Type: array of strings Scope: all Default: none

Restrict which models users can select via /model, --model, Config tool, or ANTHROPIC_MODEL. Does not affect the Default option unless enforceAvailableModels is enabled.

{ "availableModels": ["sonnet", "haiku"] }

Type: boolean Scope: all Default: false

When true, fast mode does not persist across sessions. Each session starts with fast mode off, requiring users to enable it with /fast. The user’s preference is still saved.

Type: string Scope: all Default: "auto" Values: "auto" | "in-process" | "tmux"

How agent team teammates display. "auto" uses split panes in tmux or iTerm2, in-process otherwise.

Type: boolean Scope: all Default: false

Show the “clear context” option on the plan accept screen. Set to true to restore the option, which was hidden by default starting in v2.1.81.

Type: number Scope: all Default: none

Probability (0–1) that the session quality survey appears when eligible. Set to 0 to suppress entirely. Useful when using Bedrock, Vertex, or Foundry.

Type: string Scope: all Default: none Values: "disable"

Set to "disable" to prevent auto mode from being activated. Removes auto from the Shift+Tab cycle and rejects --permission-mode auto at startup.

Type: boolean Scope: user / local Default: true

Whether plan mode uses auto mode semantics when auto mode is available. Not read from shared project settings.

Type: object Scope: user / local Default: none

Customize the auto mode classifier. Contains environment, allow, and soft_deny arrays of prose rules. Not read from shared project settings.

Type: string Scope: all Default: "bash" Values: "bash" | "powershell"

Default shell for input-box ! commands. "powershell" requires CLAUDE_CODE_USE_POWERSHELL_TOOL=1.

Type: boolean Scope: all Default: false

Skip the WebFetch blocklist check before fetching URLs.

Type: object Scope: all Default: none

Environment variables applied to every session. Use this instead of wrapper scripts to set variables. See the Environment Variables section for all supported keys.

{
"env": {
"NODE_ENV": "development",
"CLAUDE_CODE_EFFORT_LEVEL": "medium"
}
}

Type: string Scope: all Default: "~/.claude/plans"

Directory where /plan outputs are stored. Path is relative to the project root.

Type: boolean Scope: all Default: true

Enable or disable the auto-memory feature that automatically saves context across sessions.

Type: string Scope: user / local / managed Default: none

Custom directory for auto-memory storage. Accepts ~/-expanded paths. Not accepted in project settings (.claude/settings.json) to prevent shared repos from redirecting memory writes to sensitive locations.


Control what tools and operations Claude can perform.

Type: array of strings Scope: all Default: []

Permission rules that allow tool use without prompting. Arrays are concatenated across scopes. See Permission Rule Syntax below.

Type: array of strings Scope: all Default: []

Permission rules requiring user confirmation before tool use.

Type: array of strings Scope: all Default: []

Permission rules blocking tool use. Highest safety precedence, cannot be overridden by allow/ask rules at any scope.

Type: array of strings Scope: all Default: []

Additional working directories that Claude has access to, beyond the current project root.

{ "permissions": { "additionalDirectories": ["../shared-libs/"] } }

Type: string Scope: all Default: unset (built-in starting mode for the surface: auto in interactive terminal and VS Code sessions since v2.1.283) Values: "default" (alias "manual") | "acceptEdits" | "plan" | "auto" | "dontAsk" | "bypassPermissions"

Default permission mode when opening Claude Code. auto and bypassPermissions take effect only from user or managed settings (or --permission-mode), not from project or local settings. In Remote environments, only "acceptEdits" and "plan" are honored.

Type: string Scope: all Default: none Values: "disable"

Set to "disable" to prevent bypassPermissions mode from being activated. Disables the --dangerously-skip-permissions flag. Most useful in managed settings.

Type: boolean Scope: managed only Default: false

When true, user and project allow, ask, and deny rules are ignored. Only managed permission rules apply.

Rules follow the format Tool or Tool(specifier). Evaluation order: deny first, then ask, then allow. The first matching rule wins.

ToolPatternExample
BashCommand pattern with wildcardsBash(npm run *), Bash(git *)
ReadFile path patternRead(.env), Read(./secrets/**)
EditFile path patternEdit(src/**), Edit(*.ts)
WriteFile path patternWrite(*.md)
WebFetchdomain:hostnameWebFetch(domain:example.com)
WebSearchNo specifierWebSearch
TaskAgent nameTask(Explore)
AgentAgent nameAgent(researcher)
MCPmcp__server__tool or MCP(server:tool)mcp__memory__*

Path prefixes for Read/Edit rules:

PrefixMeaning
//Absolute path from filesystem root
~/Relative to home directory
/Relative to project root
./ or noneRelative path (current directory)

Bash wildcard notes: * matches at any position. Bash(ls *) (space before *) matches ls -la but NOT lsof. Bash(*) is equivalent to Bash (matches all commands). The legacy :* suffix (e.g., Bash(npm:*)) is deprecated.

{
"permissions": {
"allow": ["Edit(*)", "Bash(npm run *)", "Bash(git *)"],
"ask": ["Bash(git push *)"],
"deny": ["Read(.env)", "Read(./secrets/**)"],
"defaultMode": "acceptEdits"
}
}

Type: object Scope: all Default: none

Configure custom commands to run at lifecycle events. See the hooks documentation and §7.1 of the guide for the full set of 33 hook events, exit codes, and environment variables.

Type: boolean Scope: all Default: false

Disable all hooks and any custom status line.

Type: boolean Scope: managed only Default: false

When true, only managed hooks and SDK hooks are loaded. User, project, and plugin hooks are blocked.

Type: array of strings Scope: all Default: none (no restriction)

Allowlist of URL patterns that HTTP hooks may target. Supports * as a wildcard. When defined, hooks with non-matching URLs are silently blocked. Empty array blocks all HTTP hooks. Arrays merge across settings sources.

{ "allowedHttpHookUrls": ["https://hooks.example.com/*"] }

Type: array of strings Scope: all Default: none (no restriction)

Allowlist of environment variable names that HTTP hooks can interpolate into header values. Each hook’s effective allowedEnvVars is the intersection with this list. Arrays merge across settings sources.


Type: boolean Scope: all Default: false

Automatically approve all MCP servers defined in project .mcp.json files. Avoids per-server confirmation prompts.

Type: array of strings Scope: all Default: none

Allowlist of specific server names from .mcp.json files to approve.

Type: array of strings Scope: all Default: none

Blocklist of specific server names from .mcp.json files to reject.

Type: array Scope: managed only Default: none (no restrictions)

Allowlist of MCP servers users can configure. Each entry matches by serverName, serverCommand, or serverUrl. Undefined = no restrictions, empty array = lockdown.

{
"allowedMcpServers": [
{ "serverName": "github" },
{ "serverCommand": "npx @modelcontextprotocol/*" },
{ "serverUrl": "https://mcp.company.com/*" }
]
}

Type: array Scope: managed only Default: none

Blocklist of MCP servers that are explicitly blocked. Applies to all scopes including managed servers. Takes precedence over allowedMcpServers.

Type: boolean Scope: managed only Default: false

When true, only allowedMcpServers from managed settings are respected. Users can still add MCP servers, but only admin-defined servers are usable. deniedMcpServers still merges from all sources.

Type: boolean Scope: managed only Default: false

Allow channels for claude.ai Team and Enterprise users. When unset or false, channel message delivery is blocked regardless of what users pass to --channels. Console API-key deployments use Console defaults and managed controls instead; do not infer their policy from this Team/Enterprise-only setting.

Type: array Scope: managed only Default: none (uses default Anthropic allowlist)

Allowlist of channel plugins that may push messages. Replaces the default Anthropic allowlist when set. Requires channelsEnabled: true. Empty array blocks all channel plugins.

Security boundary: this allowlist controls which Channel plugins may deliver messages. It does not grant a delivered message, its sender, or the plugin any Bash, filesystem, GitHub, or other tool permission. Gate senders and treat Channel content as untrusted input, including when a plugin can relay a permission prompt.


Configure bash command sandboxing for security. Available on macOS, Linux, and WSL2.

Type: boolean Scope: all Default: false

Enable bash sandboxing. Isolates bash commands from your filesystem and network.

Type: boolean Scope: all Default: false

Exit with an error at startup if sandbox.enabled is true but the sandbox cannot start (missing dependencies, unsupported platform). When false, a warning is shown and commands run unsandboxed. Useful in managed deployments that require sandboxing as a hard gate.

Type: boolean Scope: all Default: true

Auto-approve bash commands when sandboxed. When the sandbox is active, bash commands that would normally require confirmation are automatically approved.

Type: array of strings Scope: all Default: []

Commands that bypass the sandbox and run directly in your environment.

Use the glob form, not the bare name. A bare entry such as "git" matches only the zero-argument string git, so it never fires on a real invocation and the command stays sandboxed. Write "git *" instead. The bare form is what the published JSON schema suggests, which is why this is easy to get wrong: you configure something that does nothing, notice the command is still sandboxed, and only then find the glob form (anthropics/claude-code#10524). Verified on 2.1.220: "git" had no effect, "git *" worked immediately.

Every command in a call must match (v2.1.277+). Claude Code checks entries against the whole command line it sends. npm ci && docker compose build stays sandboxed unless another entry covers npm ci, and a redirect to a file, a cd, a loop or a $(...) substitution keeps the whole call sandboxed. Before v2.1.277, one matching command unsandboxed the entire call, including commands that ran before it (anthropics/claude-code#81157). Entries match the text of the call: git -C <path> push, /usr/local/bin/docker and a script that calls the tool internally do not match.

An excluded command runs with no filesystem limit and no proxy, so it can reach any host, not only allowedDomains. Scope entries to the subcommands that genuinely need to leave the sandbox rather than the whole binary. Git over SSH on macOS is the common case (on Linux and WSL2 it works sandboxed once the host is allowed): only the network operations need an exception, so list those and leave local git confined.

When the sandbox is admin-required (allowUnsandboxedCommands: false from managed settings or --settings, or managed allowManagedDomainsOnly), entries in a repository’s .claude/settings.json and .claude/settings.local.json are ignored (v2.1.282+).

"excludedCommands": [
"git push *", "git pull *", "git fetch *",
"git clone *", "git ls-remote *", "git remote *", "git submodule *"
]

This keeps git status, git diff, git log, git add, and git commit inside the sandbox.

Type: boolean Scope: all Default: true

Allow commands to opt out of the sandbox via the dangerouslyDisableSandbox parameter. Set to false for strict sandboxing where all commands must run inside the sandbox or be in excludedCommands. The /sandbox Overrides tab shows false as Strict sandbox mode.

Since v2.1.285, a false in user settings, --settings or managed settings holds even when a project sets true (before, the project’s true won). A false from managed settings or --settings also makes the sandbox admin-required: Claude Code then ignores the repository settings that would loosen it, excludedCommands included. See admin-required sandbox.

Type: boolean Scope: all Default: false

Enable a weaker sandbox for unprivileged Docker environments (Linux and WSL2 only). Reduces security.

Type: boolean Scope: all Default: false

(macOS only) Allow access to the system TLS trust service (com.apple.trustd.agent). Required for Go-based tools like gh, gcloud, and terraform when using httpProxyPort with a MITM proxy and custom CA. Reduces security.

Type: array of strings Scope: all Default: []

Specific Unix socket paths accessible in the sandbox on macOS (for SSH agents, local databases, etc.). Ignored on Linux and WSL2, where the seccomp filter cannot inspect socket paths; use allowAllUnixSockets only after accepting its broader boundary. The cross-session inbox case does not require this exception for normal SendMessage calls.

Type: boolean Scope: all Default: false

Allow all Unix socket connections in the sandbox. Overrides allowUnixSockets. On Linux and WSL2, this is the only way to permit Unix sockets because it skips the seccomp filter that otherwise blocks socket(AF_UNIX, ...) calls.

Type: boolean Scope: all Default: false

Allow binding to localhost ports (macOS only). Sandboxed commands can then listen on network ports and connect to any port on localhost, which includes every other service listening there; an unauthenticated local service such as a debugger can then act for the command outside the sandbox. On Linux and WSL2 a sandboxed command’s localhost is private to it and this key has no effect.

Type: array of strings Scope: all Default: []

Domains allowed for outbound network traffic. Supports wildcards like *.example.com.

Type: boolean Scope: managed only Default: false

When true, only allowedDomains and WebFetch(domain:...) allow rules from managed settings are respected. Non-allowed domains are blocked without prompting. Denied domains are still respected from all sources.

Type: number Scope: all Default: none

HTTP proxy port for a custom proxy (1–65535). If not specified, Claude runs its own proxy.

Type: number Scope: all Default: none

SOCKS5 proxy port for a custom proxy (1–65535).

Type: array of strings Scope: all Default: []

Domains sandboxed commands can never reach, even when a broader allowedDomains wildcard would permit them. A denied host is refused in every permission mode, and a per-command allowed_domains list in auto mode cannot open it. Set in managed settings or --settings, it also stops a repository from setting proxy ports (v2.1.285+). IPv6 addresses use the bracketed form, "[::1]" or "[::1]:443" (v2.1.229+).

Type: array of strings Scope: all Default: []

Additional paths where sandboxed commands can write. Merged across all settings scopes. Also merged with paths from Edit(...) allow permission rules.

Path prefix conventions: / = absolute, ~/ = home-relative, ./ or no prefix = project-relative in project settings / ~/.claude-relative in user settings.

Type: array of strings Scope: all Default: []

Paths where sandboxed commands cannot write. Merged with Edit(...) deny rules.

Type: array of strings Scope: all Default: []

Paths where sandboxed commands cannot read. Merged with Read(...) deny rules.

Type: array of strings Scope: all Default: []

Paths to re-allow read access within denyRead regions. Takes precedence over denyRead. Arrays merge across all settings scopes.

When read rules overlap, the more specific path wins in both directions: denyRead: ["~/"] with allowRead: ["~/projects"] opens only that subtree, and allowRead: ["~/"] with denyRead: ["~/.env"] keeps that one file blocked. A broad allow cannot silently re-expose a secret an exact deny covers.

Type: boolean Scope: user, managed, --settings (project settings ignored) Default: false Since: v2.1.216

Skip filesystem isolation while keeping network isolation. Sandboxed commands get unrestricted read and write access to the host, and their egress stays confined to allowedDomains. Use when the point of sandboxing is controlling where commands connect rather than what they write.

Turning the layer off also disables filesystem.denyRead, credentials.files, and the protection on Claude Code’s own settings files, and stops the $TMPDIR override. credentials.envVars still applies, since environment scrubbing is independent of the filesystem layer.

Project settings cannot set it, so a checked-out repository cannot switch filesystem isolation off. When managed settings configure sandbox.filesystem at all, or list any credentials.files entry, only managed settings can set it. CLAUDE_CODE_SUBPROCESS_ENV_SCRUB makes Claude Code ignore it from every source.

With the filesystem layer off and commands auto-allowed, a sandboxed command can write shell startup files, executables on $PATH, or ~/.claude/settings.json, and use them to widen its own access on the next run. Set it only for workloads you trust not to escalate.

Type: array of { path, mode } Scope: all (deny narrows only, so any scope may add one) Default: [] Since: v2.1.187

Credential files to hide from sandboxed commands. mode accepts "deny", which blocks reads the same way filesystem.denyRead does, and "mask" (v2.1.221+). With mask, on Linux and WSL2 sandboxed commands read a sentinel copy and the proxy substitutes the real value on egress; on macOS the file is simply unreadable, the same effect as deny, so a tool that authenticates with it stops working inside the sandbox. Partial masking uses extract (a regex marking the secret), decode: "jwt" with maskClaims, onExtractNoMatch and maskDuplicates (v2.1.224+). The default onExtractNoMatch: "warn" skips the entry when the pattern matches nothing, leaving the real file readable; use deny when the secret may be present but the pattern may miss it. A mask entry on a directory, a glob, a file over 8 MiB or a non-UTF-8 file falls back to deny. mask is honored only from user, managed and --settings settings.

The sandbox’s default read policy covers the entire machine, and there is no built-in credential denylist, so ~/.ssh and ~/.aws/credentials are readable by every sandboxed command until you list them. Paths follow the same prefix rules as sandbox.filesystem.*, and deny entries merge across every scope: any scope can add one, no scope can remove one another scope added.

{
"sandbox": {
"credentials": {
"files": [
{ "path": "~/.ssh", "mode": "deny" },
{ "path": "~/.aws", "mode": "deny" },
{ "path": "~/.gnupg", "mode": "deny" },
{ "path": "~/.config/gh", "mode": "deny" }
]
}
}
}

Verified on 2.1.220: with the entry in place, ls ~/.ssh from a sandboxed command returns Operation not permitted while the directory still appears in a listing of the home directory.

Type: array of { name, mode, injectHosts? } Scope: deny from all scopes; mask from user, managed, --settings only Default: [] Since: v2.1.187 (deny), v2.1.199 (mask)

Environment variables to withhold from sandboxed commands. Sandboxed commands otherwise inherit the parent environment as-is, credentials included, which is the gap a credentials.files entry alone leaves open.

deny unsets the variable before each sandboxed command runs. It affects sandboxed Bash only, so MCP servers keep their credentials: they run as separate processes, not as sandboxed commands. Wildcards such as AWS_* are not supported; list each variable. A shell startup file can undo it: zsh reads ~/.zshenv at every start, so a variable exported there is set again inside the sandboxed command, as measured on v2.1.302 (see Default behavior for the probe and the CLAUDECODE guard).

For structured values such as a DATABASE_URL or a JWT, extract, decode, maskClaims and onExtractNoMatch keep parsing tools working (v2.1.224+). AWS SigV4 requests need AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY masked together so the proxy can re-sign them (v2.1.221+); group differently named variables with credentials.awsPairs, and use credentials.sigv4 to forward streaming uploads, presigned URLs or SigV4A requests the proxy cannot re-sign (both v2.1.224+). claude doctor warns when TLS termination is unavailable or an injectHosts entry can never match.

mask protects the credential while keeping the tool that authenticates with it working. The command sees a per-session sentinel; when a request leaves the sandbox for one of the credential’s injectHosts, the proxy substitutes the real value. The command and anything it logs never hold the real credential. Each injectHosts entry must itself be covered by allowedDomains, and network.tlsTerminate is required because the proxy has to see request contents to substitute. Without it, masking fails closed: the sentinel reaches the server unchanged and authentication fails, and Claude Code reports the misconfiguration at startup.

{
"sandbox": {
"network": { "tlsTerminate": {}, "allowedDomains": ["*.github.com"] },
"credentials": {
"envVars": [
{ "name": "ANTHROPIC_API_KEY", "mode": "deny" },
{ "name": "GH_TOKEN", "mode": "mask", "injectHosts": ["api.github.com"] }
]
}
}
}

deny wins when the same variable appears with both modes in any scope. Start with deny and move a variable to mask only when a CLI actually breaks without it, since mask requires TLS termination and authorizes the proxy to send the real credential to the listed hosts.

To strip Anthropic and cloud provider credentials from all subprocesses regardless of sandboxing, set CLAUDE_CODE_SUBPROCESS_ENV_SCRUB instead.

sandbox.network.tlsTerminate ⚠️ Experimental

Section titled “sandbox.network.tlsTerminate ⚠️ Experimental”

Type: object Scope: user, managed, --settings (project settings ignored) Since: v2.1.199

Make the built-in proxy terminate TLS itself. Required by mask entries in credentials.envVars and credentials.files. It enables credential substitution; it does not add content filtering.

Type: boolean Scope: user, managed, --settings (project settings ignored) Default: false Since: v2.1.219

Deny sandboxed commands any host outside the allowlist instead of prompting. The allowlist is allowedDomains plus domains from WebFetch(domain:...) allow rules, or only the managed entries when allowManagedDomainsOnly is set. Applies to sandboxed commands only: in-process tools such as WebFetch still follow their permission rules.

Enable it last, once the domain list has survived a week of real work. Before that it converts every missing domain from a one-time prompt into a hard failure.

Type: boolean Scope: user, managed, --settings (project settings ignored) Default: false

Allow Apple Events on macOS, which the sandbox blocks by default. Fixes open, osascript, and browser-based auth flows failing with error -600.

Enabling it removes code-execution isolation: sandboxed commands can launch other applications unsandboxed with no prompt, and send AppleScript to running applications, subject to the macOS automation-consent prompt. Prefer adding the specific command to excludedCommands.

sandbox.filesystem.allowManagedReadPathsOnly

Section titled “sandbox.filesystem.allowManagedReadPathsOnly”

Type: boolean Scope: managed only Default: false

When true, only allowRead paths from managed settings are respected. allowRead entries from user, project, and local settings are ignored.

Sandbox example:

{
"sandbox": {
"enabled": true,
"autoAllowBashIfSandboxed": true,
"excludedCommands": ["git push *", "git fetch *", "docker *"],
"filesystem": {
"allowWrite": ["/tmp/build", "~/.kube"],
"denyRead": ["~/.aws/credentials"]
},
"network": {
"allowedDomains": ["github.com", "*.npmjs.org"],
"allowLocalBinding": true
}
}
}

Unix-socket exceptions are intentionally absent from this baseline. Granting /var/run/docker.sock would give sandboxed code control of the Docker daemon and an effective path to the host.


For marketplace publication and the <claude-code-hint /> recommendation boundary, see Plugin Distribution and Recommendation Hints. The settings below govern local and managed marketplace controls.

Type: object Scope: all Default: none

Enable or disable specific plugins by key (format: plugin-name@marketplace-name).

{
"enabledPlugins": {
"formatter@acme-tools": true,
"experimental@acme-tools": false
}
}

Type: object Scope: project Default: none

Add custom plugin marketplaces. Use source: "settings" to declare plugins inline without hosting a repository.

Type: array Scope: managed only Default: none (no restrictions)

Allowlist of permitted plugin marketplaces. When set, users can only add plugins from listed marketplaces. Empty array blocks all additions.

Type: array Scope: managed only Default: none

Block specific plugin marketplace sources. Blocked sources are checked before downloading, so they never touch the filesystem.

Type: string Scope: managed only Default: none

Custom message appended to the plugin trust warning shown before installation. Use for org-specific context like confirming plugins from an internal marketplace are vetted.

Type: array Scope: all Default: none

Marketplaces the user declined to install (stored automatically).

Type: array Scope: all Default: none

Plugins the user declined to install (stored automatically).

Type: object Scope: all Default: none

Per-plugin MCP server configurations, keyed by plugin@marketplace.


Type: string Scope: all Values: "low" | "medium" | "high" | "xhigh"

Fallback effort for a model without a saved choice. Defaults depend on the model: Opus 5.5, Sonnet 5.5 and Haiku 5.5 use medium; Fable 5.1 uses high. The legacy Haiku 4.5 has no effort parameter. Opus 5.5 ignores this legacy key in user settings, but explicit project, local, managed, and --settings values still apply.

Type: object Scope: all

Since v2.1.251, /effort saves the active model’s level here. Confirm with Enter to save or s for this session only. max is session-only and cannot be stored as effortLevel.

{
"modelSettings": {
"claude-opus-5-5": { "effortLevel": "high" }
}
}

Within one settings file, a model-specific level wins over the fallback. Across files, normal settings precedence applies. CLAUDE_CODE_EFFORT_LEVEL overrides the chosen level, subject to organization caps. /effort auto clears the saved choice for the active model.

Type: boolean Scope: all

Pair with availableModels to keep the /model Default choice inside that allowlist. Without this setting, the allowlist alone does not constrain Default. See the official model settings reference.

Type: object Scope: all Default: none

Map Anthropic model IDs to provider-specific model IDs (e.g., Bedrock inference profile ARNs). Each key is a model picker entry name; each value is the provider model ID.

{
"modelOverrides": {
"claude-opus-4-6": "arn:aws:bedrock:us-east-1:123456789:inference-profile/anthropic.claude-opus-4-6-v1:0"
}
}

Model aliases reference:

AliasDescription
"default"Recommended model for your account type
"sonnet"Sonnet 5.5 on the Anthropic API; cloud-provider aliases may resolve to 4.6 or 4.5
"opus"Opus 5.5 on most providers; Opus 4.6 on Foundry
"haiku"Haiku 5.5 on the Anthropic API (Claude Code v2.1.293+); Haiku 4.5 on cloud providers
"fable"Fable 5.1 where available; Fable 5 through the Claude apps gateway
"best"Fable where available, otherwise Opus
"sonnet[1m]"Sonnet with 1M token context
"opusplan"Opus for planning, Sonnet for execution

Type: object Scope: all Default: none

Configure a custom status line. The command receives a JSON object on stdin with fields like context_window.used_percentage, rate_limits.five_hour.used_percentage, etc.

{
"statusLine": {
"type": "command",
"command": "~/.claude/statusline.sh",
"padding": 0
}
}

Type: object Scope: all Default: none

Configure a custom script for @ file path autocomplete. The command receives JSON on stdin with a query field and outputs newline-separated file paths (max 15).

{
"fileSuggestion": {
"type": "command",
"command": "~/.claude/file-suggestion.sh"
}
}

Type: string Scope: all Default: "Default"

Controls how Claude communicates throughout the session. Select with /output-style or /config. The setting value is case-sensitive; the command accepts case-insensitive names.

Built-in values:

  • "Default": no style instructions, Claude Code’s standard system prompt
  • "Proactive": starts implementing right away and makes reasonable assumptions on routine decisions; permission mode unchanged
  • "Concise": leads with the result, no preamble, narration, or closing recap
  • "Explanatory": adds Insight blocks explaining the choices behind the code
  • "Learning": adds Insight blocks and leaves design-heavy pieces for you at TODO(human) markers

A value that does not match a style name exactly (for example explanatory) falls back to Default. /output-style and the /config menu save the choice to .claude/settings.local.json.

Custom styles: reference the style name (the file name without .md, unless the frontmatter sets name) from .claude/output-styles/ for a project or ~/.claude/output-styles/ for a user-wide style. Custom styles omit the built-in Claude Code software engineering instructions unless their YAML frontmatter sets keep-coding-instructions: true. Style selection applies to the next message. Restart to discover newly created style files.

{ "outputStyle": "Explanatory" }
{ "outputStyle": "strict-reviewer" }

Setting persists across sessions. Explanatory and Learning increase output tokens; prompt caching can reduce repeated input cost but does not discount generated output. See Section 9.7 for full documentation and custom style examples.

Type: boolean Scope: all Default: true

Show tips in the spinner while Claude is working.

Type: object Scope: all Default: none

Customize the action verbs shown in the spinner and turn duration messages. Set mode to "replace" to use only your verbs, or "append" to add to defaults.

{
"spinnerVerbs": {
"mode": "replace",
"verbs": ["Cooking", "Brewing", "Crafting", "Conjuring"]
}
}

Type: object Scope: all Default: none

Override spinner tips with custom strings. tips: array of strings. excludeDefault: when true, only show custom tips.

{
"spinnerTipsOverride": {
"tips": ["Use /compact at 50% context", "Plan mode helps for complex tasks"],
"excludeDefault": true
}
}

Type: boolean Scope: all Default: true

Control whether the @ file picker respects .gitignore patterns.

Type: boolean Scope: all Default: false

Reduce or disable UI animations (spinners, shimmer, flash effects) for accessibility.


Type: string Scope: all Default: none

Shell script path (executed in /bin/sh) that outputs an auth token sent as X-Api-Key and Authorization: Bearer headers for model requests. Useful for short-lived credentials.

{ "apiKeyHelper": "/bin/generate_temp_api_key.sh" }

Type: string Scope: all Default: none Values: "claudeai" | "console"

Restrict login to Claude.ai accounts ("claudeai") or Claude Console API-billing accounts ("console").

Type: string Scope: all Default: none

UUID of an organization to automatically select during login, bypassing the org selection step. Requires forceLoginMethod to be set.


Type: string Scope: all Default: Git trailer with co-authored-by line

Attribution text added to git commits. Supports git trailers. Set to empty string to disable commit attribution entirely.

Type: string Scope: all Default: Generated message with Claude Code link

Attribution text added to pull request descriptions. Set to empty string to disable.

Type: boolean Scope: all Default: true Status: DEPRECATED, use attribution instead

Whether to include the Co-Authored-By byline. Superseded by the attribution object.

{
"attribution": {
"commit": "Generated with AI\n\nCo-Authored-By: AI <ai@example.com>",
"pr": ""
}
}

Type: array of strings Scope: all Default: []

Directories to symlink from the main repository into each worktree, avoiding large duplicated directories on disk (e.g., node_modules).

Type: array of strings Scope: all Default: []

Directories to check out in each worktree via git sparse-checkout (cone mode). Only listed paths are written to disk, useful for large monorepos.

{
"worktree": {
"symlinkDirectories": ["node_modules", ".cache"],
"sparsePaths": ["packages/my-app", "shared/utils"]
}
}

Type: string Scope: all Default: none

Custom script that modifies the .aws directory. Runs to refresh AWS credentials before API calls.

{ "awsAuthRefresh": "aws sso login --profile myprofile" }

Type: string Scope: all Default: none

Custom script that outputs JSON with AWS credentials. Used for non-standard credential sources.

Type: string Scope: all Default: none

Script to generate dynamic OpenTelemetry headers. Runs at startup and periodically. See the monitoring docs for the expected output format.


These settings are stored in ~/.claude.json, not settings.json. Adding them to settings.json triggers a schema validation error.

KeyTypeDefaultDescription
autoConnectIdebooleanfalseAuto-connect to a running IDE when Claude Code starts from an external terminal
autoInstallIdeExtensionbooleantrueAuto-install the Claude Code IDE extension when running from a VS Code terminal
editorModestring"normal"Key binding mode: "normal" or "vim". Written automatically by /vim
showTurnDurationbooleantrueShow turn duration messages after responses (e.g., “Cooked for 1m 6s”)
terminalProgressBarEnabledbooleantrueShow terminal progress bar in ConEmu, Ghostty 1.2.0+, and iTerm2 3.6.6+

Keys confirmed in the JSON schema not covered in the sections above:

KeyTypeDescription
claudeMdExcludes 📋 Schema onlyarrayGlob patterns for CLAUDE.md files to exclude from loading
allowManagedMcpServersOnlyboolean(Managed) Only managed MCP servers are usable
allowManagedHooksOnlyboolean(Managed) Only managed and SDK hooks are loaded
allowManagedModsOnlyboolean(Managed) Stop user-installed mods from loading; their plugins stay installed and their skills, commands, agents, and MCP servers still load. Built-in mods keep running
autoMemoryEnabledbooleanEnable/disable auto-memory feature
feedbackSurveyRatenumberSurvey appearance probability (0–1)

Set in your shell before launching claude, or configure under the env key in settings.json to apply to every session. When an env var and an equivalent settings field both apply, the env var takes precedence (e.g. ANTHROPIC_MODEL overrides the model setting). Changes take effect on the next claude launch.

VariableDescription
ANTHROPIC_API_KEYAPI key for Anthropic API access. When set, used instead of your Claude subscription even if logged in
ANTHROPIC_AUTH_TOKENCustom Authorization header value (prefixed with Bearer )
ANTHROPIC_BASE_URLCustom API endpoint for proxies or LLM gateways
ANTHROPIC_CUSTOM_HEADERSCustom headers added to API requests. Format: Name: Value, newline-separated for multiple
ANTHROPIC_BETASComma-separated extra anthropic-beta header values. Works with all auth methods including Claude.ai subscription
ANTHROPIC_WORKSPACE_IDWorkspace ID for workload identity federation when a federation rule covers more than one workspace
CLAUDE_CODE_OAUTH_TOKENOAuth access token for Claude.ai authentication. Alternative to /login for SDK and automated environments
CLAUDE_CODE_OAUTH_REFRESH_TOKENOAuth refresh token for headless auth. claude auth login exchanges this directly instead of opening a browser. Requires CLAUDE_CODE_OAUTH_SCOPES
CLAUDE_CODE_OAUTH_SCOPESSpace-separated OAuth scopes the refresh token was issued with. Required when CLAUDE_CODE_OAUTH_REFRESH_TOKEN is set
CLAUDE_CONFIG_DIROverride the configuration directory (default: ~/.claude)
VariableDescription
ANTHROPIC_MODELModel to use. Accepts aliases (sonnet, opus, haiku) or full model IDs. Overrides the model setting
ANTHROPIC_DEFAULT_HAIKU_MODELOverride the Haiku model alias with a custom model ID
ANTHROPIC_DEFAULT_HAIKU_MODEL_NAMEDisplay name for the Haiku model override
ANTHROPIC_DEFAULT_HAIKU_MODEL_DESCRIPTIONDescription for the Haiku model override
ANTHROPIC_DEFAULT_HAIKU_MODEL_SUPPORTED_CAPABILITIESCapabilities for the Haiku model override
ANTHROPIC_DEFAULT_SONNET_MODELOverride the Sonnet model alias
ANTHROPIC_DEFAULT_SONNET_MODEL_NAMEDisplay name for the Sonnet model override
ANTHROPIC_DEFAULT_SONNET_MODEL_DESCRIPTIONDescription for the Sonnet model override
ANTHROPIC_DEFAULT_SONNET_MODEL_SUPPORTED_CAPABILITIESCapabilities for the Sonnet model override
ANTHROPIC_DEFAULT_OPUS_MODELOverride the Opus model alias (e.g. claude-opus-4-6[1m])
ANTHROPIC_DEFAULT_OPUS_MODEL_NAMEDisplay name for the Opus model override
ANTHROPIC_DEFAULT_OPUS_MODEL_DESCRIPTIONDescription for the Opus model override
ANTHROPIC_DEFAULT_OPUS_MODEL_SUPPORTED_CAPABILITIESCapabilities for the Opus model override
ANTHROPIC_CUSTOM_MODEL_OPTIONModel ID to add as a custom entry in the /model picker
ANTHROPIC_CUSTOM_MODEL_OPTION_NAMEDisplay name for the custom model entry
ANTHROPIC_CUSTOM_MODEL_OPTION_DESCRIPTIONDisplay description for the custom model entry
ANTHROPIC_CUSTOM_MODEL_OPTION_SUPPORTED_CAPABILITIESCapabilities for the custom model entry
ANTHROPIC_SMALL_FAST_MODELDEPRECATED. Use ANTHROPIC_DEFAULT_HAIKU_MODEL instead
ANTHROPIC_SMALL_FAST_MODEL_AWS_REGIONAWS region for the Haiku-class model on Bedrock when ANTHROPIC_DEFAULT_HAIKU_MODEL is also set
CLAUDE_CODE_SUBAGENT_MODELOverride model for subagents (e.g. haiku)
CLAUDE_CODE_SUBAGENT_MODEL_FORCESet to 1 (v2.1.257+) to force every subagent, teammate and workflow agent onto CLAUDE_CODE_SUBAGENT_MODEL, or onto the lead’s model when that variable is unset or inherit, ignoring the spawn prompt and the agent definition’s model (Agent Teams)
CLAUDE_CODE_EFFORT_LEVELEffort level: low, medium, high, xhigh, max, or auto. Takes precedence over /effort and the effortLevel setting
FALLBACK_FOR_ALL_PRIMARY_MODELS ⚠️ Not in official list (checked 2026-10-27)Any non-empty value triggers fallback to --fallback-model after repeated overload errors on any primary model, not just Opus
CLAUDE_CODE_DISABLE_LEGACY_MODEL_REMAPSet to 1 to prevent automatic remapping of Opus 4.0 and 4.1 to the current Opus version on the Anthropic API
VariableDescription
CLAUDE_CODE_USE_BEDROCKUse Amazon Bedrock (1 to enable)
AWS_BEARER_TOKEN_BEDROCKBedrock API key for authentication
ANTHROPIC_BEDROCK_BASE_URLOverride the Bedrock endpoint URL. Use for custom regions or when routing through an LLM gateway
ANTHROPIC_BEDROCK_SERVICE_TIERBedrock service tier: default, flex, or priority. Sent as X-Amzn-Bedrock-Service-Tier
CLAUDE_CODE_SKIP_BEDROCK_AUTHSkip AWS auth for Bedrock (e.g. when using an LLM gateway)
CLAUDE_ENABLE_BYTE_WATCHDOG_BEDROCKSet to 1 to enable byte-level streaming idle watchdog on Bedrock
VariableDescription
CLAUDE_CODE_USE_MANTLEUse the Bedrock Mantle endpoint
ANTHROPIC_BEDROCK_MANTLE_BASE_URLOverride the Bedrock Mantle endpoint URL
CLAUDE_CODE_SKIP_MANTLE_AUTHSkip AWS auth for Bedrock Mantle
VariableDescription
CLAUDE_CODE_USE_VERTEXUse Google Vertex AI (1 to enable)
ANTHROPIC_VERTEX_BASE_URLOverride the Vertex AI endpoint URL
ANTHROPIC_VERTEX_PROJECT_IDGCP project ID for Vertex AI requests (overridden by GCLOUD_PROJECT or GOOGLE_CLOUD_PROJECT)
CLAUDE_CODE_SKIP_VERTEX_AUTHSkip Google auth for Vertex
VERTEX_REGION_CLAUDE_3_5_HAIKURegion override for Claude 3.5 Haiku on Vertex AI
VERTEX_REGION_CLAUDE_3_5_SONNETRegion override for Claude 3.5 Sonnet on Vertex AI
VERTEX_REGION_CLAUDE_3_7_SONNETRegion override for Claude 3.7 Sonnet on Vertex AI
VERTEX_REGION_CLAUDE_4_0_OPUSRegion override for Claude 4.0 Opus on Vertex AI
VERTEX_REGION_CLAUDE_4_0_SONNETRegion override for Claude 4.0 Sonnet on Vertex AI
VERTEX_REGION_CLAUDE_4_1_OPUSRegion override for Claude 4.1 Opus on Vertex AI
VERTEX_REGION_CLAUDE_4_5_OPUSRegion override for Claude Opus 4.5 on Vertex AI
VERTEX_REGION_CLAUDE_4_5_SONNETRegion override for Claude Sonnet 4.5 on Vertex AI
VERTEX_REGION_CLAUDE_4_6_OPUSRegion override for Claude Opus 4.6 on Vertex AI
VERTEX_REGION_CLAUDE_4_6_SONNETRegion override for Claude Sonnet 4.6 on Vertex AI
VERTEX_REGION_CLAUDE_4_7_OPUSRegion override for Claude Opus 4.7 on Vertex AI. Added in v2.1.111
VERTEX_REGION_CLAUDE_4_8_OPUSRegion override for Claude Opus 4.8 on Vertex AI
VERTEX_REGION_CLAUDE_HAIKU_4_5Region override for Claude Haiku 4.5 on Vertex AI
VariableDescription
CLAUDE_CODE_USE_FOUNDRYUse Microsoft Foundry (1 to enable)
ANTHROPIC_FOUNDRY_API_KEYAPI key for Microsoft Foundry authentication
ANTHROPIC_FOUNDRY_BASE_URLFull base URL for the Foundry resource (alternative to ANTHROPIC_FOUNDRY_RESOURCE)
ANTHROPIC_FOUNDRY_RESOURCEFoundry resource name. Required if ANTHROPIC_FOUNDRY_BASE_URL is not set
CLAUDE_CODE_SKIP_FOUNDRY_AUTHSkip Azure auth for Foundry
VariableDescription
CLAUDE_CODE_USE_ANTHROPIC_AWSUse Claude Platform on AWS
ANTHROPIC_AWS_API_KEYWorkspace API key for Claude Platform on AWS, generated in the AWS Console
ANTHROPIC_AWS_BASE_URLOverride the Claude Platform on AWS endpoint URL
ANTHROPIC_AWS_WORKSPACE_IDRequired workspace ID. Sent as the anthropic-workspace-id header on every request
CLAUDE_CODE_SKIP_ANTHROPIC_AWS_AUTHSkip client-side authentication for Claude Platform on AWS
VariableDescription
CLAUDE_CODE_ENABLE_AUTO_MODESet to 1 to make auto mode available on Bedrock, Vertex AI, and Foundry. Added in v2.1.158. No effect on Anthropic API
CLAUDE_CODE_PROVIDER_MANAGED_BY_HOSTSet by host platforms managing model provider routing. When set, provider-selection and auth variables in settings files are ignored
VariableDescription
API_TIMEOUT_MSAPI request timeout in milliseconds (default: 600000). Maximum: 2147483647
BASH_DEFAULT_TIMEOUT_MSDefault bash command timeout in milliseconds (default: 120000)
BASH_MAX_TIMEOUT_MSMaximum bash command timeout in milliseconds (default: 600000)
BASH_MAX_OUTPUT_LENGTHMaximum characters in bash output before saving to a file and sending the path
MAX_THINKING_TOKENSFixed budget on older extended-thinking models. 0 disables thinking where supported, but not on Opus 5.5, Sonnet 5.5, Haiku 5.5 or Fable. The fixed-budget compatibility switch applies only to Opus 4.6 and Sonnet 4.6
CLAUDE_CODE_MAX_OUTPUT_TOKENSOutput limit; default and cap depend on the model. Unrecognized gateway IDs default to 32,000. Values above the known model cap are reduced to that cap
CLAUDE_CODE_FILE_READ_MAX_OUTPUT_TOKENSOverride default file read token limit
CLAUDE_CODE_MAX_CONTEXT_TOKENSOverride context window size Claude Code assumes for the active model. Only takes effect when DISABLE_COMPACT is also set
CLAUDE_CODE_MAX_TURNSCap the number of agentic turns per session. Equivalent to --max-turns (the flag takes precedence when both are set)
CLAUDE_CODE_MAX_RETRIESOverride the number of retries for failed API requests (default: 10)
CLAUDE_CODE_MAX_TOOL_USE_CONCURRENCY ⚠️ Not in official list (checked 2026-10-27)Maximum read-only tools and subagents executing in parallel (default: 10)
CLAUDE_ASYNC_AGENT_STALL_TIMEOUT_MSStall timeout in milliseconds for background subagents (default: 600000)
TASK_MAX_OUTPUT_LENGTHNo effect since TaskOutput was removed in v2.1.278
MAX_STRUCTURED_OUTPUT_RETRIESRetries when model response fails --json-schema validation in non-interactive mode (default: 5)
CLAUDE_CODE_STOP_HOOK_BLOCK_CAPMaximum consecutive times a Stop hook may block the turn from ending before Claude Code overrides it (default: 8)
CLAUDE_CODE_SESSIONEND_HOOKS_TIMEOUT_MSSessionEnd hook time budget in ms. Default: 1.5s, raised to the highest configured per-hook timeout up to 60s
CLAUDE_CODE_API_KEY_HELPER_TTL_MSCredential refresh interval in ms for apiKeyHelper
MCP_TIMEOUTMCP server startup timeout in ms (default: 30000)
MCP_TOOL_TIMEOUTMCP tool execution timeout in ms (default: 100000000)
VariableDescription
CLAUDECODESet to 1 in subprocesses Claude Code spawns (Bash/PowerShell tools, tmux sessions, hook commands, status line commands, stdio MCP subprocesses). Use to detect when a script runs inside Claude Code
DEBUGSet to 1 (or true/yes/on) to enable debug mode. Logs written to ~/.claude/debug/<session-id>.txt. Namespace patterns like DEBUG=express:* do not trigger it
CLAUDE_CODE_DEBUG_LOGS_DIROverride the debug log file path (file path, not a directory). Requires debug mode enabled separately
CLAUDE_CODE_DEBUG_LOG_LEVELMinimum log level for the debug file: verbose, debug (default), info, warn, error
CLAUDE_CODE_SHELLOverride automatic shell detection
CLAUDE_CODE_SHELL_PREFIXCommand prefix prepended to all shell commands Claude Code spawns
CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIRReturn to the original working directory after each Bash command in the main session (1 to enable)
CLAUDE_CODE_NEW_INITSet to 1 to make /init run an interactive setup flow asking which files to generate
CLAUDE_CODE_SIMPLERun with minimal system prompt and only Bash, file read, and file edit tools. Equivalent to --bare
CLAUDE_CODE_SIMPLE_SYSTEM_PROMPTSet to 1 to use a shorter system prompt and abbreviated tool descriptions. Set to 0/false/no/off to opt out
CLAUDE_CODE_SUBPROCESS_ENV_SCRUBSet to 1 to strip Anthropic and cloud provider credentials from subprocess environments (Bash, hooks, MCP stdio servers)
CLAUDE_CODE_ADDITIONAL_DIRECTORIES_CLAUDE_MDSet to 1 to load CLAUDE.md files from directories specified with --add-dir
CLAUDE_CODE_TMPDIROverride the temp directory used for internal temp files
CLAUDE_CODE_GIT_BASH_PATHWindows only: path to Git Bash executable when not in PATH
CLAUDE_CODE_GLOB_HIDDENSet to false to exclude dotfiles from Glob tool results (dotfiles included by default)
CLAUDE_CODE_GLOB_NO_IGNORESet to false to make the Glob tool respect .gitignore patterns
CLAUDE_CODE_GLOB_TIMEOUT_SECONDSGlob tool file discovery timeout in seconds (default: 20, 60 on WSL)
CLAUDE_CODE_USE_NATIVE_FILE_SEARCH ⚠️ Not in official list (checked 2026-10-27)Set to 1 to discover commands/subagents using Node.js file APIs instead of ripgrep
CLAUDE_CODE_PERFORCE_MODESet to 1 to enable Perforce-aware write protection (edits fail on files lacking owner-write bit)
CLAUDE_CODE_POWERSHELL_RESPECT_EXECUTION_POLICYSet to 1 to stop Claude Code from passing -ExecutionPolicy Bypass when spawning PowerShell
CLAUDE_CODE_FORK_SUBAGENTControls fork mode, which is on by default: Claude can start forks that inherit the full conversation context. See the official variable list for the off value and per-surface behavior
CLAUDE_AUTO_BACKGROUND_TASKSSet to 1 to force-enable automatic backgrounding of long-running agent tasks: a subagent moves to the background after about two minutes. CLAUDE_CODE_WORKER_CHECKIN_SCHEDULE sets the check-in reminders
CLAUDE_CODE_REMOTESet automatically to true when running as a cloud session
CLAUDE_CODE_REMOTE_SESSION_IDSet automatically in cloud sessions to the current session ID
CLAUDE_CODE_SESSION_IDSet automatically to the current session ID in Bash/PowerShell subprocesses, hook commands, and MCP stdio subprocesses
CLAUDE_EFFORTSet automatically in Bash/hook subprocesses to the active effort level: low, medium, high, xhigh, or max
CLAUDE_ENV_FILEPath to a shell script run before each Bash command. Also populated dynamically by SessionStart, Setup, CwdChanged, and FileChanged hooks
USE_BUILTIN_RIPGREPSet to 0 to use system-installed rg instead of the one bundled with Claude Code
VariableDescription
CLAUDE_AUTOCOMPACT_PCT_OVERRIDEAuto-compact trigger threshold as a percentage (1-100). Default ~95%. Lower values trigger compaction earlier
CLAUDE_CODE_AUTO_COMPACT_WINDOWContext capacity in tokens used for compaction calculations. Defaults to model’s context window (200K standard, 1M for extended). Lower value on a 1M model treats it as smaller for compaction purposes
CLAUDE_CODE_DISABLE_1M_CONTEXTSet to 1 to disable 1M token context window support
DISABLE_AUTO_COMPACTSet to 1 to disable automatic compaction when approaching the context limit. Manual /compact remains available
DISABLE_COMPACTSet to 1 to disable all compaction, including manual /compact
CLAUDE_CODE_DISABLE_THINKINGSet to 1 to force-disable extended thinking regardless of model support
DISABLE_INTERLEAVED_THINKINGSet to 1 to prevent sending the interleaved-thinking beta header. Use when your gateway does not support it
VariableDescription
CLAUDE_CODE_ENABLE_TELEMETRYEnable OpenTelemetry data collection (1 to enable). Required before configuring OTel exporters
DISABLE_TELEMETRYDisable telemetry (1 to disable)
DO_NOT_TRACKSet to 1 to opt out of telemetry. Equivalent to DISABLE_TELEMETRY
DISABLE_ERROR_REPORTINGDisable Sentry error reporting (1 to disable)
DISABLE_GROWTHBOOKSet to 1 to disable GrowthBook feature-flag fetching and use code defaults for every flag
CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFICEquivalent to setting DISABLE_AUTOUPDATER, DISABLE_FEEDBACK_COMMAND, DISABLE_ERROR_REPORTING, and DISABLE_TELEMETRY

Standard OTEL exporter variables (OTEL_METRICS_EXPORTER, OTEL_LOGS_EXPORTER, OTEL_EXPORTER_OTLP_ENDPOINT, OTEL_EXPORTER_OTLP_PROTOCOL, OTEL_EXPORTER_OTLP_HEADERS, OTEL_METRIC_EXPORT_INTERVAL, OTEL_RESOURCE_ATTRIBUTES, and signal-specific variants) are supported. CLAUDE_CODE_ENABLE_TELEMETRY=1 is required to activate OTel collection.

VariableDescription
OTEL_LOG_RAW_API_BODIESEmit API request/response JSON as log events. Set to 1 for inline bodies (truncated at 60 KB), or file:<dir> to write full bodies to disk
OTEL_LOG_TOOL_CONTENTSet to 1 to include tool input/output content in OTel span events. Disabled by default
OTEL_LOG_TOOL_DETAILSSet to 1 to include tool input arguments, MCP server names, and raw error strings in OTel traces. Disabled by default
OTEL_LOG_USER_PROMPTSSet to 1 to include user prompt text in OTel traces. Disabled by default
OTEL_METRICS_INCLUDE_ACCOUNT_UUIDSet to false to exclude account UUID from metrics attributes (default: included)
OTEL_METRICS_INCLUDE_ENTRYPOINTSet to true to include session entrypoint in metrics attributes (default: excluded). Added in v2.1.152
OTEL_METRICS_INCLUDE_RESOURCE_ATTRIBUTESSet to false to exclude OTEL_RESOURCE_ATTRIBUTES keys from metric datapoint labels (default: included). Added in v2.1.161
OTEL_METRICS_INCLUDE_SESSION_IDSet to false to exclude session ID from metrics attributes (default: included)
OTEL_METRICS_INCLUDE_VERSIONSet to true to include Claude Code version in metrics attributes (default: excluded)
CLAUDE_CODE_OTEL_FLUSH_TIMEOUT_MSTimeout in ms for flushing pending OTel spans on shutdown (default: 5000)
CLAUDE_CODE_OTEL_SHUTDOWN_TIMEOUT_MSTimeout in ms for the OTel exporter to finish on exit (default: 2000)
CLAUDE_CODE_OTEL_HEADERS_HELPER_DEBOUNCE_MSRefresh interval in ms for dynamic OTel headers (default: 1740000, 29 minutes)
CLAUDE_CODE_ENABLE_FEEDBACK_SURVEY_FOR_OTELSet to 1 to route session quality survey ratings to your OTel collector instead of Anthropic
CLAUDE_CODE_PROPAGATE_TRACEPARENTSet to 1 to propagate W3C trace context when ANTHROPIC_BASE_URL points at a custom proxy. Added in v2.1.152
VariableDescription
CLAUDE_CODE_DISABLE_ADAPTIVE_THINKINGDisable adaptive reasoning on supported models (1 to disable). No effect on Opus 4.7 and later
CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETASStrip anthropic-beta request headers and beta tool-schema fields. Use when a proxy rejects unknown beta headers
CLAUDE_CODE_DISABLE_BACKGROUND_TASKSDisable all background task functionality including run_in_background, auto-backgrounding, and Ctrl+B
CLAUDE_CODE_DISABLE_AUTO_MEMORYSet to 1 to disable auto memory. Set to 0 to force it on when --bare mode would disable it
CLAUDE_CODE_DISABLE_FAST_MODEDisable fast mode entirely (1 to disable)
CLAUDE_CODE_DISABLE_CRONDisable scheduled/cron tasks (1 to disable)
CLAUDE_CODE_DISABLE_GIT_INSTRUCTIONSRemove built-in commit/PR instructions from system prompt. Takes precedence over includeGitInstructions
CLAUDE_CODE_DISABLE_NONSTREAMING_FALLBACKDisable the non-streaming fallback for failed streaming requests
CLAUDE_CODE_DISABLE_FEEDBACK_SURVEYDisable session quality survey prompts (1 to disable)
CLAUDE_CODE_DISABLE_TERMINAL_TITLEDisable automatic terminal title updates and background title-generation requests
CLAUDE_CODE_DISABLE_AGENT_VIEWSet to 1 to disable background agents and agent view (claude agents, --bg, /background)
CLAUDE_CODE_DISABLE_WORKFLOWSSet to 1 to disable workflows
CLAUDE_CODE_DISABLE_FILE_CHECKPOINTINGSet to 1 to disable file checkpointing. /rewind will not restore code changes
CLAUDE_CODE_DISABLE_CLAUDE_MDSSet to 1 to prevent loading any CLAUDE.md memory files, including auto-memory
CLAUDE_CODE_DISABLE_ATTACHMENTSSet to 1 to disable attachment processing. @ file mentions sent as plain text
CLAUDE_CODE_DISABLE_POLICY_SKILLSSet to 1 to skip loading skills from the system-wide managed skills directory
CLAUDE_CODE_DISABLE_OFFICIAL_MARKETPLACE_AUTOINSTALLSet to 1 to skip automatic addition of the official plugin marketplace on first run
CLAUDE_CODE_ENABLE_PROMPT_SUGGESTIONSet to false to disable prompt suggestions in the input area
CLAUDE_CODE_ENABLE_TASKSAs of v2.1.142, Task tools are the default. Set to 0 to revert to TodoWrite
CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMSEnable experimental agent teams feature (1 to enable)
ENABLE_CLAUDEAI_MCP_SERVERSSet to false to disable Claude.ai MCP servers (enabled by default for logged-in users)
CLAUDE_CODE_USE_POWERSHELL_TOOLControl the PowerShell tool. Auto-enabled on Windows without Git Bash. Set 0 to disable, 1 to opt in
CLAUDE_CODE_ENABLE_AWAY_SUMMARYOverride session recap availability: 0 to force off, 1 to force on
CLAUDE_CODE_ENABLE_FINE_GRAINED_TOOL_STREAMINGControl streaming of tool call inputs. Set 0 to opt out, 1 to force on via proxy
CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERYSet to 1 to populate the /model picker from your gateway’s /v1/models endpoint
DISABLE_DOCTOR_COMMANDSet to 1 to hide the /doctor command
DISABLE_LOGIN_COMMANDSet to 1 to hide the /login command
DISABLE_LOGOUT_COMMANDSet to 1 to hide the /logout command
DISABLE_UPGRADE_COMMANDSet to 1 to hide the /upgrade command
DISABLE_EXTRA_USAGE_COMMANDSet to 1 to hide the /usage-credits command
DISABLE_INSTALL_GITHUB_APP_COMMANDSet to 1 to hide the /install-github-app command
DISABLE_UPDATESSet to 1 to block all updates including manual claude update and claude install
IS_DEMOEnable demo mode: hides email/org info, skips onboarding. Useful when screensharing
VariableDescription
DISABLE_PROMPT_CACHINGDisable all prompt caching (1 to disable, takes precedence over per-model settings)
DISABLE_PROMPT_CACHING_HAIKUDisable prompt caching for Haiku model requests
DISABLE_PROMPT_CACHING_SONNETDisable prompt caching for Sonnet model requests
DISABLE_PROMPT_CACHING_OPUSDisable prompt caching for Opus model requests
ENABLE_PROMPT_CACHING_1HSet to 1 to request a 1-hour prompt cache TTL instead of the default 5 minutes
ENABLE_PROMPT_CACHING_1H_BEDROCKDEPRECATED. Use ENABLE_PROMPT_CACHING_1H instead
FORCE_PROMPT_CACHING_5MSet to 1 to force the 5-minute cache TTL even when 1-hour TTL would otherwise apply
CLAUDE_CODE_ATTRIBUTION_HEADERSet to 0 to omit the attribution block from the system prompt. Improves prompt-cache hit rates through LLM gateways
VariableDescription
MAX_MCP_OUTPUT_TOKENSMax MCP output tokens per call (default: 25000). Warning shown when output exceeds 10,000 tokens
ENABLE_TOOL_SEARCHMCP tool search mode: auto:N activates when tool count exceeds N% of context, true always defers, false loads all upfront
MCP_CLIENT_SECRETMCP OAuth client secret. Avoids the interactive prompt when adding a server with --client-secret
MCP_OAUTH_CALLBACK_PORTFixed port for the MCP OAuth redirect callback
MCP_CONNECTION_NONBLOCKINGAs of v2.1.142, MCP startup is non-blocking by default. Set to 0 to restore the blocking 5-second wait
MCP_CONNECT_TIMEOUT_MSHow long blocking MCP startup waits before snapshotting the tool list (default: 5000)
MCP_SERVER_CONNECTION_BATCH_SIZEMaximum stdio MCP servers connected in parallel during startup (default: 3)
MCP_REMOTE_SERVER_CONNECTION_BATCH_SIZEMaximum remote (HTTP/SSE) MCP servers connected in parallel during startup (default: 20)
CLAUDE_CODE_MCP_ALLOWLIST_ENVSet to 1 to spawn stdio MCP servers with only a safe baseline environment plus the server’s configured env
CLAUDE_AGENT_SDK_MCP_NO_PREFIXSet to 1 to skip the mcp__<server>__ prefix on tool names from SDK-created MCP servers
VariableDescription
HTTP_PROXYHTTP proxy URL for network requests
HTTPS_PROXYHTTPS proxy URL for network requests
NO_PROXYComma-separated hosts that bypass the proxy
CLAUDE_CODE_PROXY_RESOLVES_HOSTSAllow the proxy to perform DNS resolution (1 to enable)
CLAUDE_CODE_CERT_STOREComma-separated CA certificate sources: bundled (Mozilla CA set) and/or system (OS trust store). Default: bundled,system
CLAUDE_CODE_CLIENT_CERTClient certificate path for mTLS
CLAUDE_CODE_CLIENT_KEYClient private key path for mTLS
CLAUDE_CODE_CLIENT_KEY_PASSPHRASEPassphrase for an encrypted mTLS key
CLAUDE_CODE_EXTRA_BODYJSON object merged into the top level of every API request body. Useful for provider-specific parameters
CLAUDE_ENABLE_STREAM_WATCHDOGSet to 1 to enable event-level streaming idle watchdog. Off by default
CLAUDE_ENABLE_BYTE_WATCHDOGForce-enable byte-level streaming idle watchdog. Default on for Anthropic API connections. Set 0 to disable
CLAUDE_STREAM_IDLE_TIMEOUT_MSTimeout in ms before the streaming watchdog closes a stalled connection (default minimum: 300000)
VariableDescription
CLAUDE_CODE_NO_FLICKERSet to 1 to enable fullscreen rendering (reduces flicker, keeps memory flat in long conversations). Equivalent to the tui setting
CLAUDE_CODE_DISABLE_ALTERNATE_SCREENSet to 1 to disable fullscreen rendering and use the classic main-screen renderer
CLAUDE_CODE_ALT_SCREEN_FULL_REPAINTSet to 1 to repaint the entire screen on every frame instead of incremental updates. Use if fullscreen mode shows stale text
CLAUDE_CODE_DISABLE_VIRTUAL_SCROLLSet to 1 to disable virtual scrolling in fullscreen mode. Use if scrolling shows blank regions
CLAUDE_CODE_DISABLE_MOUSESet to 1 to disable mouse tracking in fullscreen mode. Keeps terminal native copy-on-select behavior
CLAUDE_CODE_SCROLL_SPEEDMouse wheel scroll multiplier in fullscreen mode (1-20)
CLAUDE_CODE_NATIVE_CURSORSet to 1 to show the terminal’s own cursor at the input caret instead of a drawn block
CLAUDE_CODE_ACCESSIBILITYSet to 1 to keep the native terminal cursor visible and disable the inverted-text cursor indicator
CLAUDE_CODE_HIDE_CWDSet to 1 to hide the working directory in the startup logo
CLAUDE_CODE_TMUX_TRUECOLORSet to 1 to allow 24-bit truecolor output inside tmux
CLAUDE_CODE_FORCE_SYNC_OUTPUTSet to 1 to force-enable synchronized output (DEC private mode 2026) when your terminal supports it but is not auto-detected
CLAUDE_CODE_SYNTAX_HIGHLIGHTSet to false to disable syntax highlighting in diff output
CLAUDE_CODE_AUTO_CONNECT_IDEOverride automatic IDE connection. Set to false to prevent it, true to force a connection attempt
CLAUDE_CODE_IDE_HOST_OVERRIDEOverride the host address used to connect to the IDE extension
CLAUDE_CODE_IDE_SKIP_VALID_CHECKSet to 1 to skip validation of IDE lockfile entries during connection
DISABLE_COST_WARNINGSDisable cost warning messages
DISABLE_INSTALLATION_CHECKSDisable installation warnings
DISABLE_AUTOUPDATERDisable the auto-updater. Manual claude update still works
DISABLE_FEEDBACK_COMMANDDisable the /feedback command. Also accepted as DISABLE_BUG_COMMAND
CLAUDE_CODE_IDE_SKIP_AUTO_INSTALLSkip automatic IDE extension installation (1 to skip)
SLASH_COMMAND_TOOL_CHAR_BUDGETCharacter budget for skill metadata shown to the Skill tool. Scales dynamically at 1% of the context window
VariableDescription
CLAUDE_CODE_PLUGIN_CACHE_DIROverride the plugins root directory (default: ~/.claude/plugins)
CLAUDE_CODE_PLUGIN_SEED_DIRPath(s) to read-only plugin seed directories, separated by : on Unix or ; on Windows. Use to bundle plugins into container images
CLAUDE_CODE_PLUGIN_PREFER_HTTPSSet to 1 to clone GitHub owner/repo sources over HTTPS instead of SSH
CLAUDE_CODE_PLUGIN_KEEP_MARKETPLACE_ON_FAILURESet to 1 to keep existing marketplace cache when a git pull fails. Useful in offline environments
CLAUDE_CODE_PLUGIN_GIT_TIMEOUT_MSTimeout in ms for git operations during plugin install/update (default: 120000)
CLAUDE_CODE_SYNC_PLUGIN_INSTALLSet to 1 in non-interactive mode (-p flag) to wait for plugin installation before the first query
CLAUDE_CODE_SYNC_PLUGIN_INSTALL_TIMEOUT_MSTimeout in ms for synchronous plugin installation. No default (waits until complete without this var)
CLAUDE_CODE_ENABLE_BACKGROUND_PLUGIN_REFRESHSet to 1 to refresh plugin state at turn boundaries in non-interactive mode after a background install
FORCE_AUTOUPDATE_PLUGINSSet to 1 to force plugin auto-updates even when the main auto-updater is disabled
VariableDescription
CLAUDE_AGENT_SDK_DISABLE_BUILTIN_AGENTSSet to 1 to disable all built-in subagent types (Explore, Plan) in non-interactive mode only
CLAUDE_CODE_EXIT_AFTER_STOP_DELAYTime in ms to wait after the query loop becomes idle before automatically exiting. Useful for SDK/scripted sessions
CLAUDE_CODE_RESUME_INTERRUPTED_TURNSet to 1 to automatically resume if the previous session ended mid-turn. For SDK use
CLAUDE_CODE_RESUME_PROMPTOverride the continuation message injected when resuming a mid-turn session (default: Continue from where you left off.)
CLAUDE_CODE_SKIP_PROMPT_HISTORYSet to 1 to skip writing prompt history and session transcripts to disk
CLAUDE_CODE_SYNC_SKILLSSet to 1 to download enabled Claude.ai skills before the first query and resync every 10 minutes. Non-interactive mode only
CLAUDE_CODE_SYNC_SKILLS_WAIT_TIMEOUT_MSTimeout in ms for initial skills sync when CLAUDE_CODE_SYNC_SKILLS is set (default: 5000)
CLAUDE_CODE_SCRIPT_CAPSJSON object limiting invocations of specific scripts per session when CLAUDE_CODE_SUBPROCESS_ENV_SCRUB is set
CCR_FORCE_BUNDLESet to 1 to force claude --remote to bundle and upload the local repo even when GitHub access is available
CLAUDE_REMOTE_CONTROL_SESSION_NAME_PREFIXPrefix for auto-generated Remote Control session names. Defaults to machine hostname
VariableDescription
CLAUDE_CODE_TEAM_NAME ⚠️ Not in official list (checked 2026-10-27)Team name for agent teams. Set automatically on agent team members
CLAUDE_CODE_TASK_LIST_IDShare a task list across sessions. Set the same ID in multiple Claude Code instances to coordinate on a shared task list

These appear in community sources or older documentation but are not confirmed in current official docs.

VariableDescription
CLAUDE_CODE_USER_EMAIL ⚠️ UnverifiedProvide user email synchronously for authentication
CLAUDE_CODE_ORGANIZATION_UUID ⚠️ UnverifiedProvide organization UUID synchronously for authentication
CLAUDE_CODE_ACCOUNT_UUID ⚠️ UnverifiedOverride account UUID for authentication
CLAUDE_CODE_PLAN_MODE_REQUIRED ⚠️ UnverifiedRequire plan mode for sessions
CLAUDE_CODE_SKIP_FAST_MODE_NETWORK_ERRORS ⚠️ UnverifiedAllow fast mode when org status check fails due to network error
CLAUDE_CODE_SKIP_SETTINGS_SETUP ⚠️ UnverifiedSkip first-run settings setup flow
CLAUDE_CODE_DISABLE_TOOLS ⚠️ UnverifiedComma-separated list of tools to disable
CLAUDE_CODE_DISABLE_MCP ⚠️ UnverifiedDisable all MCP servers (1 to disable)
CLAUDE_CODE_HIDE_ACCOUNT_INFO ⚠️ UnverifiedHide email/org info from UI
CLAUDE_CODE_PROMPT_CACHING_ENABLED ⚠️ UnverifiedOverride prompt caching behavior
DISABLE_NON_ESSENTIAL_MODEL_CALLS ⚠️ UnverifiedDisable flavor text and non-essential model calls

Settings added in Claude Code v2.1.275 to v2.1.310

Section titled “Settings added in Claude Code v2.1.275 to v2.1.310”

Checked against the official settings reference and environment variables on October 27, 2026. Release numbers come from the release tracker.

Key or variableScopeEffect
idleCompactionAny filefalse turns off idle compaction, which compacts a 200K+ token conversation left idle about 54 minutes before its one-hour prompt cache expires (v2.1.299). CLAUDE_CODE_IDLE_COMPACT_MIN_TOKENS changes the threshold (minimum 100,000)
syncClaudeAiSkills, syncClaudeAiPluginsUser, local, managed, --settingsfalse stops syncing the skills or plugins enabled on your claude.ai account into terminal sessions; a repository cannot turn it off for you (v2.1.275)
maxProseWidthAny fileCaps response prose at this many columns (minimum 40); tables and code keep the full width (v2.1.282)
deniedModelsManagedBlocks models by family alias or ID, even when availableModels permits them (v2.1.283)
availableModelsMatchManaged"exact" keeps a newer version blocked until it is listed; the default "prefix" lets "claude-opus-5" permit Opus 5.5
allowedProvidersManagedLists the services a machine may reach Claude through (anthropic, bedrock, gateways…); other providers are refused at startup and login (v2.1.285)
sandbox.allowNestedUserNamespacesUser, managedLets sandboxed Linux/WSL2 commands create user namespaces (Chromium, nested bubblewrap). Turns off a security boundary: sandboxed code can end up running outside the sandbox (v2.1.305)
CLAUDE_CODE_AUTO_MODE_SERVEREnvironment0 uses Claude Code’s own classifier requests instead of server-side auto mode review
CLAUDE_CODE_AUTO_MODE_MODELEnvironmentModel for the classifier requests Claude Code sends in auto mode, as a full model ID (v2.1.302)
CLAUDE_CODE_DISABLE_DANGEROUS_RM_TIMEOUTEnvironment1 turns off the time limit and auto-denial on critical-path removals in auto and bypassPermissions (v2.1.281)
CLAUDE_CODE_DISABLE_WEB_FETCHEnvironment1 turns off WebFetch; WebSearch stays available (v2.1.285)
CLAUDE_CODE_DISABLE_FILE_MENTIONSEnvironment1 stops expanding @ file mentions; Claude gets the paths and reads them with its tools (v2.1.299)
CLAUDE_CODE_DISABLE_ULTRAREVIEW_TRUST_CHECKEnvironment1 lets claude ultrareview run in an untrusted folder when it cannot ask, such as CI; or pass --trust-workspace (v2.1.309)
MCP_PROTOCOL_NEGOTIATIONEnvironmentlegacy skips the probe for MCP protocol revision 2026-07-28 on the v2 MCP client runtime

Further settings keys in the official reference

Section titled “Further settings keys in the official reference”

These keys appear in the official settings reference (read October 27, 2026) without a dedicated entry above. Descriptions are the first sentence of the official entry; read it for types, defaults and precedence.

KeyScopePurpose
advisorModelAny filePick which model answers when Claude calls the server-side advisor tool.
agentPushNotifEnabledAny fileAllow Claude to send a push notification to your phone when it decides one is worth sending, for example when a long task finishes.
appendPluginsUser or managedList the managed plugins whose mods run after every mod a user installs, in the listed order.
askUserQuestionTimeoutUser or managedLet an unanswered AskUserQuestion dialog auto-continue after a period of idle time, submitting whatever options you had already selected.
autoContinueAtUsageLimitUser or managedAfter a claude.ai usage limit stops your session, wait in the open session and continue the task automatically after the reset.
awaySummaryEnabledAny fileShow a one-line session recap when you return to the terminal after a few minutes away.
browserExternalPageToolsManagedStop Claude from using its tools to read or act on external pages in the desktop app’s Browser pane.
claudeInChromeDefaultEnabledGlobal configStart every interactive CLI session with Chrome integration on, without passing --chrome each time.
copyFullResponseGlobal configMake /copy copy the full response every time, without showing the picker.
copyOnSelectGlobal configCopy text to your clipboard automatically when you finish selecting it with the mouse in fullscreen rendering or agent view.
defaultToAgentsViewGlobal configOpen agent view instead of a new conversation when you run claude with no arguments.
diffToolGlobal configChoose where Claude Code shows the diff of an Edit or Write change it proposes when a VS Code or JetBrains IDE is connected: "auto" opens it in the IDE’s diff viewer, "terminal" keeps it in the terminal.
disableAgentViewAny fileTurn off background agents and agent view: claude agents, --bg, /background, and the on-demand supervisor.
disableArtifactAny file> [!WARNING]
disableBrowserExternalNavigationManagedTurn off external browsing in the desktop app’s Browser pane for people and Claude alike.
disableCommandPluginSourcesManagedBlock the command plugin source, which installs a plugin by running a marketplace-declared command on the user’s machine.
disableDesktopLocalSessionsManagedTurn off Code sessions that run on the device in the desktop app, for deployments where developers should work on remote machines over SSH.
disableMobileSimulatorToolsManagedBlock Claude’s tools for the desktop app’s iOS Simulator pane.
disableRemoteControlAny fileTurn off Remote Control: Claude Code then refuses claude remote-control, the --remote-control flag, auto-start, and the in-session toggle, and reports that your organization’s policy disabled it.
disableSideloadFlagsManagedReject the --plugin-dir, --plugin-url, --agents, and --mcp-config CLI flags at startup, which users could otherwise pass to bypass strictKnownMarketplaces for a single run.
disableWorkflowsAny fileTurn off dynamic workflows and the bundled workflow commands for everyone your settings reach, such as an organization through managed settings.
enableArtifactAny fileTurn off the Artifact tool, which publishes session output as a private web page on claude.ai.
enableWorkflowsAny fileTurn dynamic workflows on or off for yourself when your plan’s default isn’t what you want.
externalEditorContextGlobal configWhen you press Ctrl+G, Claude Code opens the prompt you’re typing in your external editor.
fileCheckpointingEnabledAny fileHave Claude Code snapshot files before each edit so /rewind can restore them.
gatewayInternalNetworksManagedDeclare the public IPv4 blocks that your organization numbers its internal network from, so /login accepts a cloud gateway there.
gcpAuthRefreshAny fileRun your own command to refresh Google Cloud Application Default Credentials when Claude Code finds they’ve expired or can’t be loaded, so Google Cloud’s Agent Platform requests keep working without you re-authenticating by hand.
inputNeededNotifEnabledAny fileGet a push notification on your phone when a permission prompt or question is waiting for your input.
leftArrowOpensAgentsGlobal configPress ← on an empty prompt to background the session and open agent view.
managedSourcesBehaviorManagedChoose whether Claude Code applies only the highest-priority managed source your organization delivers, or combines every admin source it delivers.
minimumVersionAny fileKeep background auto-updates and claude update from installing any version below this one, so moving to the "stable" channel doesn’t downgrade you from a newer "latest" build.
permissionExplainerEnabledGlobal config> [!WARNING]
policyHelperManagedRun an executable you deploy that computes managed settings at startup, so you can derive policy from device posture, identity, or a remote service instead of a static file.
policyHelper.pathManagedName the helper executable Claude Code runs.
policyHelper.refreshIntervalMsManagedHave Claude Code re-run the helper in the background on an interval so policy changes reach a running session.
policyHelper.timeoutMsManagedSet how long Claude Code waits for the helper before treating the run as failed.
prStatusFooterEnabledGlobal configShow a badge in the prompt footer for the current branch’s open pull request or merge request, with a colored underline that shows its status.
preferredNotifChannelAny fileChoose how Claude Code notifies you when a task completes or a permission prompt is waiting.
prependPluginsUser or managedList the managed plugins whose mods run before every mod a user installs, in the listed order.
processWrapperUser or managedOn macOS and Linux, place a corporate launcher command in front of the background processes Claude Code starts.
promptSuggestionEnabledAny fileShow or hide prompt suggestions, the grayed-out predictions that appear in your prompt input.
remote.defaultEnvironmentIdAny filePick the default cloud environment for cloud sessions you create from the CLI, such as with claude --cloud.
remoteControlAtStartupAny fileConnect Remote Control automatically when each interactive session starts, instead of waiting for /remote-control.
sandbox.credentials.allowPlaintextInjectUser or managedAllow mask substitution on plain HTTP requests as well as TLS-terminated HTTPS.
sandbox.credentials.awsPairsUser or managedGroup masked environment variables that form one AWS credential for SigV4 re-signing when your credential lives in variables with non-standard names.
sandbox.credentials.sigv4User or managedChoose what the sandbox proxy does with AWS request forms it can’t re-sign: streaming for aws-chunked streaming uploads, presigned for presigned URLs, and sigv4a for SigV4A asymmetric signatures.
sandbox.ignoreViolationsAny fileSilence sandbox violation reports for paths you expect a command to probe and be refused, such as a tool that checks /etc/hosts on startup, so those denials don’t show up as violations or in what Claude sees.
skillListingBudgetFractionAny fileEach turn, Claude sees a listing of your skills with their descriptions, and Claude Code caps that listing at a share of the context window.
skillListingMaxDescCharsAny fileEach turn, Claude sees a listing of your skills that shows each skill’s description and when_to_use text.
skipAutoPermissionPromptUser or managedSkip the one-time notice describing auto mode that Claude Code shows when you first enter auto mode yourself, for example through your own settings or the mode selector, rather than when the built-in default starts a session in it.
skipDangerousModePermissionPromptUser, local, or managedSkip the confirmation dialog Claude Code shows before a session enters bypassPermissions mode, whether from --dangerously-skip-permissions or from defaultMode: "bypassPermissions".
sshConfigsUser or managedAdd SSH connections to the Desktop environment selector.
sshHostAllowlistManagedLimit the hosts a Desktop SSH session can connect to.
sshHostCommandsUser or managedAdd SSH connections whose host comes from a command to the Desktop environment selector.
strictPluginOnlyCustomizationManagedBlock skills, agents, hooks, and MCP servers from user and project sources, so they can come only from plugins or managed settings.
strictPluginOnlyCustomization.agentsManagedLock the agents surface.
strictPluginOnlyCustomization.hooksManagedLock the hooks surface:
strictPluginOnlyCustomization.mcpManagedLock the mcp surface.
strictPluginOnlyCustomization.skillsManagedLock the skills surface.
switchModelsOnFlagAny fileChoose what happens when a safety classifier flags a request: switch to the fallback model and continue, or pause so you can choose between switching and editing the prompt.
syntaxHighlightingDisabledAny fileClaude Code colors code by language in the diffs, code blocks, and file previews it shows in the terminal, with its built-in highlighter; no plugin or language server is involved.
teammateDefaultModelGlobal config> [!WARNING]
terminalTitleFromRenameAny fileClaude Code sets your terminal tab’s title.
timeFormatAny fileChoose how Claude Code writes the times it shows in the interface, such as the done 6:05 PM at the end of each turn duration message and the timestamps in the transcript viewer.
timeZoneAny fileShow the times in the interface in a time zone other than your system’s.
viewModeAny fileSet the transcript view Claude Code starts in: "default", "verbose", or "focus".
workflowKeywordTriggerEnabledAny fileChoose whether typing the keyword ultracode in a prompt triggers a dynamic workflow.
worktree.locationUserChoose the folder on a remote machine where Desktop SSH sessions create their worktrees, instead of <project-root>/.claude/worktrees/.

Further environment variables in the official list

Section titled “Further environment variables in the official list”

These variables appear in the official environment variable list (read October 27, 2026) without a row above. Descriptions are the first sentence of the official entry.

VariablePurpose
ANTHROPIC_DEFAULT_FABLE_MODEL_DESCRIPTIONDisplay description for the pinned Fable model in the /model picker.
ANTHROPIC_DEFAULT_FABLE_MODEL_NAMEDisplay name for the pinned Fable model in the /model picker.
ANTHROPIC_DEFAULT_FABLE_MODEL_SUPPORTED_CAPABILITIESComma-separated list of capabilities the pinned Fable model supports, for example effort,thinking.
ANTHROPIC_FEDERATION_RULE_IDFederation rule ID for Workload Identity Federation.
ANTHROPIC_FOUNDRY_AUTH_TOKENBearer token for Microsoft Foundry authentication, such as a Microsoft Entra access token.
ANTHROPIC_ORGANIZATION_IDOrganization ID for Workload Identity Federation.
ANTHROPIC_PROFILEName of the Anthropic profile to authenticate with, such as one created by ant auth login or by signing in to a Console account without an API key.
API_FORCE_IDLE_TIMEOUTOverride the 5-minute body idle timeout that aborts a streaming model response when no bytes arrive.
BETA_TRACING_ENDPOINTOTLP/HTTP endpoint for detailed beta tracing: with ENABLE_BETA_TRACING_DETAILED=1, logs and traces go there instead of to the configured exporters.
CLAUDE_AUTO_BACKGROUND_TASKSSet to 1 to force-enable automatic backgrounding of long-running agent tasks.
CLAUDE_AX_PREPARK_MSIn screen reader mode, how many milliseconds Claude Code waits before it writes a new or changed line.
CLAUDE_AX_STARTUP_QUIET_MSIn screen reader mode, how many milliseconds Claude Code holds the first interface render after the startup confirmation line.
CLAUDE_BYTE_STREAM_IDLE_TIMEOUT_MSTimeout in milliseconds for the byte-level streaming idle watchdog; when set, it takes precedence over CLAUDE_STREAM_IDLE_TIMEOUT_MS for that watchdog and leaves the event-level watchdog unchanged.
CLAUDE_CODE_ARTIFACT_AUTO_OPENSet to 0 to stop Claude Code from opening the browser automatically when a new artifact is published
CLAUDE_CODE_ARTIFACT_COMMENTSSet to 0 to stop Claude reading and replying to comments on an artifact.
CLAUDE_CODE_ARTIFACT_COMMENTS_AUTOREACTSet to 0 to stop Claude replying on its own to comments sent to it.
CLAUDE_CODE_ARTIFACT_PUBLISH_EDITSSet to 0 to stop Claude from applying small changes to an artifact’s file as part of publishing it.
CLAUDE_CODE_AUTO_BACKGROUND_WORKER_CHECKIN_SECONDSRemoved in v2.1.283.
CLAUDE_CODE_AWS_CHAIN_RESOLVE_TIMEOUT_MSTime in milliseconds Claude Code waits for the AWS default credential provider chain to produce credentials before the request fails with AWS default-chain credential resolve timed out (default: 60000).
CLAUDE_CODE_BASH_EDIT_DIFFSet to 0 to turn off the diff of the files that changed while a Bash command ran, or 1 to record it in every permission mode.
CLAUDE_CODE_BG_TASKS_REPORT_RUNNINGSet to 0 to make a non-interactive session report an idle status to its host at every turn end, even while background work is still running.
CLAUDE_CODE_BRIDGE_SESSION_IDSet automatically in Bash tool and hook command subprocesses while the session has an active Remote Control connection, and removed when the connection ends.
CLAUDE_CODE_BS_AS_CTRL_BACKSPACESet to 0 to make Claude Code read the 0x08 byte, also written ^H, as plain Backspace, or 1 to read it as Ctrl+Backspace.
CLAUDE_CODE_CONNECT_TIMEOUT_MSRemoved in v2.1.186 and now a no-op.
CLAUDE_CODE_DISABLE_ADVISOR_TOOLSet to 1 to disable the advisor tool.
CLAUDE_CODE_DISABLE_ARTIFACTSet to 1 to turn off the Artifact tool, which publishes session output as a private web page on claude.ai.
CLAUDE_CODE_DISABLE_AUTH_REFRESH_LOCKSet to 1 to make a Claude Code process run its gcpAuthRefresh or awsAuthRefresh command itself instead of waiting while another process runs it.
CLAUDE_CODE_DISABLE_BEDROCK_CONTENT_TYPE_DEFAULTSet to 1 to stop Claude Code from treating an Amazon Bedrock streaming response with a missing or empty Content-Type header as Amazon Bedrock’s binary event stream.
CLAUDE_CODE_DISABLE_BEDROCK_CONTENT_TYPE_GUARDSet to 1 to skip the check that an Amazon Bedrock streaming response carries the application/vnd.amazon.eventstream content-type.
CLAUDE_CODE_DISABLE_BG_EXIT_HANDOFFSet to 1 to stop a background session’s in-flight work when the supervisor stops, restarts, or updates that session’s process, instead of handing it to the session’s next process.
CLAUDE_CODE_DISABLE_CFC_PROMPTSet to 1 to keep the Claude in Chrome browser tools available while omitting the Chrome section of the system prompt and the /claude-in-chrome bundled skill.
CLAUDE_CODE_DISABLE_EXPLORE_PLAN_AGENTSSet to 1 to disable the built-in Explore and Plan subagents.
CLAUDE_CODE_DISABLE_INLINE_SHELL_RM_PROMPTSet to 1 to stop Claude Code from reading scripts passed to a shell with -c, such as bash -c 'rm -rf ~', for critical-path removals.
CLAUDE_CODE_DISABLE_MCP_TASK_BACKGROUNDSet to 1 so that Claude waits for the result of an MCP task in an interactive session.
CLAUDE_CODE_DISABLE_MODEL_ACCESS_FALLBACKSet to 1 to stop Claude Code on Amazon Bedrock and Google Cloud’s Agent Platform from switching to an older model when your account loses access to a session’s model mid-session; the refused request fails at once instead.
CLAUDE_CODE_DISABLE_MOUNTED_DRIVE_RM_PROMPTSet to 1 to turn off the mounted drive prompt for paths such as /Volumes/USB.
CLAUDE_CODE_DISABLE_MTLS_RELOAD_ON_STALE_CONNECTIONSet to 1 to stop Claude Code from re-reading the mTLS client certificate and key when an API request fails with a connection-level error, such as a connection reset or a TLS handshake error.
CLAUDE_CODE_DISABLE_NOTIFICATION_PRESENCE_CHECKSet to 1 to send the PushNotification tool’s desktop notification even while you are typing in or focused on the terminal.
CLAUDE_CODE_DISABLE_PERMISSION_PROMPT_NOTIFY_HOOKSSet to 1 to stop Claude Code from running your Notification hooks for unanswered permission requests in sessions where Claude Code sends them to the Agent SDK’s canUseTool callback, which is how Claude Desktop and the VS Code extension host Claude Code.
CLAUDE_CODE_DISABLE_POWERSHELL_CMD_RM_DENYSet to 1 to turn off the PowerShell tool check that denies the cmd built-ins rd, rmdir, del, and erase on a system path, such as a drive root or your home directory.
CLAUDE_CODE_DISABLE_REFUSAL_FALLBACKSet to 1 to turn off the automatic model switch when a safety classifier flags a request, the behavior the switchModelsOnFlag setting controls
CLAUDE_CODE_DISABLE_STRUCTURED_OUTPUTSSet to 1 to stop Claude Code from sending the structured-output output_config.format field and the anthropic-beta value that pairs with it, for an LLM gateway whose upstream rejects them.
CLAUDE_CODE_DISABLE_WINDOWS_SHELL_LAUNCHERSet to 1 to start PowerShell tool commands on Windows directly instead of through the cmd.exe launcher.
CLAUDE_CODE_EMIT_SESSION_STATE_EVENTSSet to 1 to add session_state_changed messages, which carry the session’s state, to the message stream.
CLAUDE_CODE_ENABLE_CFCSet to 1 to start a CLI session with Chrome integration on, or 0 to start with it off.
CLAUDE_CODE_ENABLE_OPUS_4_7_FAST_MODERemoved in v2.1.142, when the fast mode default moved from Opus 4.6 to Opus 4.7
CLAUDE_CODE_ENABLE_SHARE_BUNDLESet to 1 to make /share save its report and the whole session transcript to a local archive under ~/.claude/feedback-bundles/ instead of sending the report to Anthropic.
CLAUDE_CODE_FORCE_SESSION_PERSISTENCESet to 1 to force transcript persistence, prompt history, and claude agents registration even when this claude was launched from inside another Claude Code session.
CLAUDE_CODE_FORCE_STRIKETHROUGHSet to 1 to force strikethrough rendering for ~~text~~ in Claude’s responses when your terminal supports it but is not auto-detected, such as over SSH without TERM_PROGRAM forwarded.
CLAUDE_CODE_FORCE_TERMINAL_IMAGESSet to 1 to draw mod Image elements as pictures when your terminal draws kitty graphics protocol images with Unicode placeholders and you still see an Image’s dimmed alt text.
CLAUDE_CODE_GATEWAY_MODEL_DISCOVERY_TIMEOUT_MSTimeout in milliseconds for the gateway model discovery request that CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY turns on (default: 3000).
CLAUDE_CODE_NONBLOCKING_STDOUTSet to 1 to write terminal output through a second non-blocking file descriptor, so a terminal that stops reading, such as a paused tmux control-mode pane or a stalled SSH connection, can’t freeze Claude Code mid-session.
CLAUDE_CODE_OTEL_DIAG_STDERRSet to 1 to write OpenTelemetry exporter diagnostic errors to stderr.
CLAUDE_CODE_PLUGIN_DIRSPlugin directories to load for the session, each loaded the way a --plugin-dir flag loads it.
CLAUDE_CODE_PLUGIN_DIR_WATCHControls whether Claude Code reloads a mod when the mod’s files change.
CLAUDE_CODE_PRINT_BG_WAIT_CEILING_MSCeiling in milliseconds on idle waiting for background work, such as subagents and workflows, after the final turn in non-interactive mode with the -p flag.
CLAUDE_CODE_PROMPT_CACHE_TTLSet 5m or 1h, the only values Claude Code accepts, to choose the prompt cache TTL for the main conversation: your interactive, -p, and SDK turns, plus the helpers that run inline with them.
CLAUDE_CODE_RESUME_INTERRUPTED_TURN_MAX_AGE_MSMaximum age in milliseconds of the last transcript message for a session that ended mid-turn to continue automatically on resume.
CLAUDE_CODE_SEND_FEEDBACKSet to 0 to turn off Claude-drafted feedback for a session.
CLAUDE_CODE_SKIP_AWS_CRED_CACHESet to 1 to turn off the in-process cache of credentials resolved from the AWS default credential provider chain, so Claude Code resolves the chain on every API request.
CLAUDE_CODE_SKIP_FAST_MODE_ORG_CHECKSet to 1 to skip the client-side fast mode availability check, for proxies that intercept the check’s request rather than refuse it.
CLAUDE_CODE_SKIP_MODEL_ACCESS_MEMORYThe startup model checks on Amazon Bedrock and Google Cloud’s Agent Platform remember on this machine which models they found your account can’t invoke, for up to a day.
CLAUDE_CODE_STARTUP_FAILURE_RESULTSSet to 1 to have a session started with --output-format stream-json write a result message naming why Claude Code refused to start for startup failures that otherwise end with stderr alone.
CLAUDE_CODE_SUBAGENT_PROMPT_CACHE_TTLSet 5m or 1h, the only values Claude Code accepts, to choose the prompt cache TTL for requests outside the main conversation, such as subagents, workflows, and background work.
CLAUDE_CODE_SYNC_SKILLS_INSTALL_TIMEOUT_MSTimeout in milliseconds for the skills resync that runs mid-session when an app built on the Agent SDK reloads skills (default: 30000).
CLAUDE_CODE_TOOL_MEMORY_CGROUP_EXCLUDEOn Linux and WSL, set to a comma-separated list of the kinds of processes Claude Code excludes from the tool memory cap, such as mcp or lsp.
CLAUDE_CODE_TRANSCRIPT_LOCAL_GCSet to 1 to limit how large the transcript file of a long -p or Agent SDK session grows.
CLAUDE_CODE_USER_DIALOG_TIMEOUT_MSDeadline in milliseconds before Claude Code cancels a dialog it forwards to a remote client such as a Remote Control or SDK host, or the approval dialog for a held cross-session message; permission prompts and AskUserQuestion questions use their own…
CLAUDE_CODE_WARN_DISK_FREE_GIBSet to a whole number of GiB, such as 10, to have Claude Code warn Claude when free space on the working directory’s filesystem falls below that amount.
CLAUDE_CODE_WARN_MEMORY_FREE_GIBOn Linux and WSL, set to a whole number of GiB such as 4 to have Claude Code warn Claude when available memory falls below that amount.
CLAUDE_CODE_WORKER_CHECKIN_SCHEDULEWhen CLAUDE_AUTO_BACKGROUND_TASKS is set to 1, how long Claude Code waits before each reminder to Claude to check on background subagents that are still running.
CLAUDE_CODE_WORKFLOW_MAX_CONCURRENT_AGENTSHow many agents a single workflow run executes at once, from 1 to 256.
CLAUDE_CODE_YOU_SHOULD_KNOW_DEBUGSet to 1 to show, above the prompt, a line for each of the last three checks the You should know mod made in this session.
CLAUDE_JOB_DIRSet by Claude Code in each background session to that session’s ~/.claude/jobs/<id> directory.
CLAUDE_STREAM_FIRST_BYTE_TIMEOUT_MSDeadline in milliseconds for the first response byte of a streaming request, on the connections where the first-byte deadline runs.
CLAUDE_SUBAGENT_BG_SHELL_MAX_MSRemoved in v2.1.260 and now a no-op.
DISABLE_PROMPT_CACHING_FABLESet to 1 to disable prompt caching for Fable models
ENABLE_BETA_TRACING_DETAILEDSet to 1, and set BETA_TRACING_ENDPOINT to your OTLP/HTTP collector endpoint, to turn on detailed beta tracing, which adds content-bearing span attributes and the claude_code.hook span.
MCP_DISCOVERY_CACHETurns the MCP discovery cache on or off.
MCP_DISCOVERY_CACHE_MAX_STALE_SMaximum age, in seconds, of a discovery-cache entry (default: 14400, or 4 hours).
MCP_DISCOVERY_CACHE_STRIKESAt a start where a discovery-cache entry is older than MCP_DISCOVERY_CACHE_TTL_S, Claude Code refreshes it in the background.
MCP_DISCOVERY_CACHE_TTL_SSeconds for which Claude Code uses a discovery-cache entry without refreshing it (default: 900).
OTEL_ATTRIBUTE_VALUE_LENGTH_LIMITStandard OpenTelemetry SDK limit on attribute value length.
OTEL_METRICS_INCLUDE_REPOSITORYSet to true to tag OpenTelemetry metrics and events with vcs.* attributes identifying the session’s repository (default: excluded).
VERTEX_REGION_CLAUDE_5_5_OPUSOverride region for Claude Opus 5.5 when using Google Cloud’s Agent Platform.
VERTEX_REGION_CLAUDE_5_5_SONNETOverride region for Claude Sonnet 5.5 when using Google Cloud’s Agent Platform.
VERTEX_REGION_CLAUDE_5_OPUSOverride region for Claude Opus 5 when using Google Cloud’s Agent Platform.
VERTEX_REGION_CLAUDE_5_SONNETOverride region for Claude Sonnet 5 when using Google Cloud’s Agent Platform.
VERTEX_REGION_CLAUDE_FABLE_5Override region for Claude Fable 5 when using Google Cloud’s Agent Platform.
VERTEX_REGION_CLAUDE_FABLE_5_1Override region for Claude Fable 5.1 when using Google Cloud’s Agent Platform.
VERTEX_REGION_CLAUDE_HAIKU_5_5Override region for Claude Haiku 5.5 when using Google Cloud’s Agent Platform.
{
"$schema": "https://json.schemastore.org/claude-code-settings.json",
"model": "sonnet",
"language": "english",
"cleanupPeriodDays": 30,
"autoUpdatesChannel": "stable",
"alwaysThinkingEnabled": false,
"includeGitInstructions": true,
"effortLevel": "medium",
"plansDirectory": "./plans",
"worktree": {
"symlinkDirectories": ["node_modules"],
"sparsePaths": ["packages/my-app", "shared/utils"]
},
"permissions": {
"allow": [
"Edit(*)",
"Write(*)",
"Bash(npm run *)",
"Bash(git *)",
"WebFetch(domain:*)",
"mcp__*"
],
"ask": ["Bash(git push *)"],
"deny": [
"Read(.env)",
"Read(./secrets/**)"
],
"additionalDirectories": ["../shared/"],
"defaultMode": "acceptEdits"
},
"enableAllProjectMcpServers": true,
"sandbox": {
"enabled": true,
"excludedCommands": ["git push *", "git fetch *", "docker *"],
"filesystem": {
"allowWrite": ["/tmp/build"],
"denyRead": ["~/.aws/credentials"]
},
"network": {
"allowedDomains": ["github.com", "*.npmjs.org"]
}
},
"attribution": {
"commit": "Generated with Claude Code",
"pr": ""
},
"statusLine": {
"type": "command",
"command": "git branch --show-current"
},
"spinnerTipsEnabled": true,
"prefersReducedMotion": false,
"env": {
"NODE_ENV": "development",
"CLAUDE_CODE_EFFORT_LEVEL": "medium"
}
}

TaskSetting / Variable
Set default modelmodel in settings or ANTHROPIC_MODEL env var
Lock model choicesavailableModels array
Silence telemetryDISABLE_TELEMETRY=1 or CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC=1
Auto-approve bashpermissions.defaultMode: "acceptEdits"
Block sensitive filespermissions.deny: ["Read(.env)"]
Reduce context compactionCLAUDE_AUTOCOMPACT_PCT_OVERRIDE=70
Set response languagelanguage: "japanese"
Custom spinner textspinnerVerbs + spinnerTipsOverride
Remove attributionattribution.commit: "", attribution.pr: ""
Pin to stable releasesautoUpdatesChannel: "stable"
Dynamic auth tokenapiKeyHelper: "/path/to/script.sh"
Large monorepoworktree.symlinkDirectories + worktree.sparsePaths
Enable sandboxingsandbox.enabled: true
Trust all project MCP serversenableAllProjectMcpServers: true