Settings keys, all 1,548.

mayBeAuthoredByNonOwner ◇ A new mayBeAuthoredByNonOwner setting is now passed through the settings-management code, though not yet visibly usednever mined ·never mined mayLeadToEvalDir ◇ No description from anybody.v2.1.301 ○gone mcp.connect ◇ Three new names appear for connecting to MCP, appending to a session and a plain MCP server call, next to existing onesnever mined ·never mined mcpAutoDiscovered ◇ The 'Claude Test' builtin feature now also respects the mcpAutoDiscovered managed-policy flagnever mined ·never mined mcpServerNameHash ◇ MCP tool telemetry now sends hashed server and tool names even when full-name consent is withheldnever mined ·never mined mcpServerPolicy ◇ A new permission-rule source called the cloud-session credential guard joins the existing rule sources.never mined ·never mined mcpServers ◇ An MCP server literally named __proto__ now warns you instead of silently loading nothing.never mined ·never mined mcpToolNameHash ◇ MCP tool telemetry now sends hashed server and tool names even when full-name consent is withheldnever mined ·never mined mcpToolTimeoutSec ◇ Admins can now cap how long MCP tool calls are allowed to run.never mined ·never mined medium ◇ Workflow size now defaults to medium, and an admin policy setting overrides whatever you chose.never mined ·never mined memorySyncIntervalMs ◇ Shared memory stores now mount on your own sandbox workers, with tunable sync interval and deletion handling.never mined ·never mined menuDescription ◇ A built-in "prototype" skill for one-page HTML mockups ships but is hard-disabled.never mined ·never mined merge ◇ Admins can set managedSourcesBehavior to merge managed settings sources instead of only the top one.never mined ·never mined mermaid ◇ Code blocks tagged mermaid now always render as diagrams, no capability probe involved.never mined ·never mined message ◇ MCP error messages now hide credentials and strip weird characters before showing them to you.v2.1.149 → v2.1.306 ○gone messageIdleNotifThresholdMs ◇ The idle nudge now checks for open dialogs and pending resumes before pinging you.never mined ·never mined messagingBeyondOwnAgentsDisabled ◇ A managed setting, messagingBeyondOwnAgentsDisabled, now turns this feature off and overrides CLAUDE_CODE_HARBOR_KITEnever mined ·never mined metadata ◇ You can add a free-form metadata block to command, skill and plugin files that Claude Code keeps but ignores.never mined ·never mined metadata.pluginRoot ◇ Marketplace validation errors now name the bad source and tell you how to fix it.never mined ·never mined microsoftAuthBroker ◇ Managed settings now check that forceLoginOrgUUID is a UUID or a list of UUIDs, and add disableMultiAccount and a required value for microsoftAuthBrokernever mined ·never mined midTurnShip ◇ No description from anybody.never mined ·never mined mine ◇ The artifact list action takes a scope of mine, shared or all, plus new asset upload, delete and open modes.never mined ·never mined minimumVersion Keep auto-updates from installing anything below a versionv2.1.138 or earlier → v2.1.283 ○gone mockCalls ◇ No description from anybody.v2.1.301 ○gone mockRecordings ◇ No description from anybody.v2.1.301 ○gone mocks ◇ Plugin evals can swap a plugin's MCP servers for fakes defined in a mocks folder.v2.1.301 ○gone mockSetupFailure ◇ Eval results now list which stand-in servers answered and count calls nothing covered.v2.1.301 ○gone mockTally ◇ No description from anybody.v2.1.301 ○gone mode ◇ Programs driving Claude Code through the SDK now see more detail about background work and turn context.v2.1.138 or earlier → v2.1.283 ○gone model Change the model Claude Code starts withv2.1.138 or earlier → v2.1.283 ○gone model.complete ◇ Plugins can install code that intercepts tool calls and repaints parts of the terminal interface.never mined ·never mined model.fork ◇ Plugins can call model.fork to ask a question against the live transcript, with no tools and a two-turn cap.never mined ·never mined modelOverride ◇ No description from anybody.v2.1.301 ○gone modelOverrides Map model IDs to your provider's IDs, such as Bedrock ARNsv2.1.138 or earlier → v2.1.283 ○gone modelPicker Choose which models the /model picker lists, in your own order and with your own labelsv2.1.242 → v2.1.283 ○gone modelPrefer1mContext ◇ A new setting can make the model picker default to the 1M-context model variant.never mined ·never mined modelPricing Report spend at your organization's contracted rates instead of list pricev2.1.242 → v2.1.283 ○gone modelPricing.multiplier ◇ modelPricing.multiplier now accepts any value greater than 0 up to 10, not just 0 to 1never mined ·never mined modelProposedGoals ◇ A "Claude-proposed goals" setting in the config panel controls whether Claude suggests session goals.v2.1.227 → v2.1.283 ○gone models ◇ Hybrid managed config now supports per-group overrides of settings and token limits.v2.1.287 → v2.1.306 ○gone modelSettings Keep a saved effort level or auto-compact window per model, or cap one model's effortv2.1.239 → v2.1.283 ○gone modelSettings.autoCompactWindow ◇ A new usage report records the auto-compact window in force and where it came from, such as an environment variable, a setting or the defaultnever mined ·never mined modelUsage ◇ SDK results now document exactly what token usage and dollar cost figures include, and that /clear resets them.never mined ·never mined modified ◇ You can add a modified date in skill frontmatter to influence which skills get pinned.never mined ·never mined modules ◇ Plugin hooks.json gains a modules key for code-registered hooks, off unless enabled remotely.never mined ·never mined monitors ◇ Most of the one-shot probes checking whether old code paths still run were removed.never mined ·never mined monitors.json ◇ Plugin loading now reports unreadable SKILL.md or monitors.json paths instead of treating them as missingnever mined ·never mined monochrome ◇ Text rendering now supports a monochrome mode that strips out colornever mined ·never mined multiFileOn ◇ Small text files saved by Artifact tools can now return their contents inline, skipping a separate Readnever mined ·never mined multiplier ◇ Gateway config takes a pricing block with a multiplier and per-model rate overrides.never mined ·never mined muted ◇ Anthropic can centrally switch off remote tool serving, and messages explain that plugins and hooks aren't sent to cloud sessions while it's off.never mined ·never mined n ◇ 'y' and 'n' keyboard shortcuts no longer confirm or cancel dialogs by defaultnever mined ·never mined name ◇ The SDK now rejects malformed skill names with a clear reason instead of creating a rule that never matches.v2.1.284 → v2.1.308 ◆in the build nameHint ◇ Subagents spawned as forks can now be given a custom display name and description instead of one derived from the task textnever mined ·never mined nameRepositoryTrustAnswers ◇ New nameRepositoryTrustAnswers capability flag added, currently offnever mined ·never mined nameToolOnPermissionAllow ◇ The remote session manager can include a tool name on allow responses.never mined ·never mined native ◇ Plugins gain a new "native" execution environment alongside the existing same-thread onenever mined ·never mined navigate ◇ Published artifacts can now declare Claude's own browser tools as a host MCP server.never mined ·never mined nawk ◇ gawk, mawk and nawk now share awk's argument parsing, so their file references are detected in permission checks.never mined ·never mined negotiatedLegacySse ◇ MCP servers can fall back to a legacy HTTP+SSE connection method if the normal connection failsnever mined ·never mined network.allowLocalBinding ◇ On Windows, Claude Code has a new sandbox error for when network.allowLocalBinding disagrees with the loopback mode set up at the startnever mined ·never mined network.parentProxy ◇ Sandbox can now terminate HTTPS itself and route traffic through an upstream proxy, both experimentalnever mined ·never mined network.tlsTerminate ◇ On Windows, TLS interception is skipped unless you supply both a CA cert and key path.never mined ·never mined networkAllowHosts ◇ Permission prompts can now show the list of network-allowed hosts for a pending commandnever mined ·never mined networkPath ◇ Reads outside your working directories can be made to fail closed, with a prompt option that persists the setting.never mined ·never mined neverPublished ◇ Artifact publish takes a syncLive option and can report a distinct never-published error instead of a generic one.never mined ·never mined next.to ◇ Hooks/plugin runtime gains a 'to' messaging function, a 'surface' concept, and table-resolution requestsnever mined ·never mined nextFireAt ◇ CronCreate and CronList now return the next time each scheduled job will run, and CronCreate's description says jobs can run up to 50% latenever mined ·never mined noBearerFor ◇ Inference config now has a per-provider list of contexts that should not carry bearer tokens.never mined ·never mined nonce ◇ No description from anybody.v2.1.301 ○gone none ◇ Clone errors no longer mangle text when the repository URL contains dollar signs.never mined ·never mined noSubscriberUploadHoldMs ◇ Event uploads can be held while a session has no subscribers, but the hold duration comes from a remote flag defaulting to zero.never mined ·never mined not_installed ◇ You can run claude plugin update with just the plugin name, no marketplace needed.v2.1.301 ○gone not_installed_at_scope ◇ Plugin enable/disable failures now return a machine-readable failureCode alongside the human-readable error messagev2.1.301 ○gone note ◇ Watching an artifact can register a wake trigger instead of a live stream, and says which you got.never mined ·never mined notices ◇ Cloud session streams gain a 'notices' field for MCP consent, hook status, and tool mute/unmute messagesnever mined ·never mined nudge ◇ Tools driving Claude Code over IPC can now tell a queued upgrade from an active restart.never mined ·never mined numTurns ◇ No description from anybody.v2.1.301 ○gone oauth.clientId ◇ A stale saved OAuth client for an MCP server is repaired in place instead of re-registered.never mined ·never mined off ◇ The feedback tool now says it covers Claude's own behaviour, not just product bugs.never mined ·never mined offCriticalPath ◇ MCP tool calls now carry account memory data and flag whether they're off the critical pathnever mined ·never mined offersBlockOutsideReads ◇ In Auto mode, a first read outside your working directories now opens a prompt and asks rather than proceeding.never mined ·never mined oidc ◇ Your gateway can accept Enterprise callers with Anthropic-minted tokens and run without any identity provider.v2.1.287 → v2.1.306 ○gone omitClaudeMd ◇ A new omitClaudeMd option lets a delegated subagent run without CLAUDE.md instructionsnever mined ·never mined omitGitStatus ◇ Sessions with a custom system prompt now automatically omit git status from prompt contextnever mined ·never mined onAssistantRound ◇ A new per-session onAssistantRound hook fires after assistant and tool rounds for the main agent, with unclear scope.never mined ·never mined onBackgroundTasks ◇ The bridge transport config now carries internal-event uploader limits and a background-tasks callback.never mined ·never mined onBlock ◇ Auto mode classifier can now flag and allow an action instead of blocking it outrightnever mined ·never mined onCancelAsyncMessage ◇ Remote Control can now cancel a single queued message instead of only interrupting the whole turnnever mined ·never mined onError ◇ A crash in the interface now reports itself instead of leaving a blank screen.never mined ·never mined onExtractNoMatch ◇ You can mask secrets in credential files so sandboxed commands see placeholders while real values are swapped in on egress.never mined ·never mined onFailure ◇ Claude Code now flags policy helpers that require success but have no fallback settingsnever mined ·never mined onForkConversation ◇ Remote-control bridge session gains an onForkConversation control op, plus a reworked reconnect flownever mined ·never mined onGetUsage ◇ SDK gains get_usage and reload_output_styles requests for pulling usage data and refreshing styles mid-session.never mined ·never mined onLinkPress ◇ Plugin and skill terminal UIs can now render a Markdown element with optional clickable linksnever mined ·never mined onMaxLifetime ◇ Runner sessions gain lifecycle hooks for max-lifetime and termination events.never mined ·never mined onPairingState ◇ Claude Code adds an encrypted local-pairing handshake for connecting to the Claude-in-Chrome extensionnever mined ·never mined onRequestUsage ◇ Forked agent runs can now report token usage as it streams in, not just at the endnever mined ·never mined onRoute ◇ No description from anybody.v2.1.301 ○gone onStreamingRewrite ◇ MessageDisplay hooks can now rewrite text already streamed to the screen, not just append to itnever mined ·never mined