Sessions can now start in a sandbox VM by default when an option is set. The new --unsandboxed flag opts out of that and of cloud-by-default too. Setting CLAUDE_CODE_ATTACHMENTS_BY_PATH turns on the send-file tool and stops Claude Code uploading the files it sends you. On self-hosted runners, all session traffic can now go through a proxy you run. Admins get a beta selfHosted section in managed settings for self-hosted session hosts. A new codeAllowedRepositories setting lists the git repositories Claude Code may work in on those hosts.
Claude Code is now listed as a reader of organization-managed settings, but those settings are not served to it yet. Organization config also gains an Allow Quick Write setting that is not switched on. Claude Desktop accepts codeAllowedRepositories but does not apply it to on-device sessions yet. The memory dialog has a new cloud memory option that does nothing yet. A new check for one tool in a permission mode called arbiter stays off unless the server turns it on.
An MCP tool whose parameter description the API rejects no longer breaks the conversation. Claude Code now blanks that description, warns you and retries, so the tool stays callable. Starting with --worktree now loads plugin WorktreeCreate hooks first, so a plugin can handle worktree creation. Command hooks now stop when their configured timeout runs out. The CLAUDE_CODE_FORCE_SANDBOX environment variable is removed. Passing --sandbox or --no-sandbox to claude remote-control is now an error instead of being silently accepted.
Written by our agent from the shipped bundle, not by Anthropic.
Setting CLAUDE_CODE_ATTACHMENTS_BY_PATH enables the send-file tool without its usual checks, skips uploads and counts the session as remote
ElsewhereAdmin config gains a selfHosted block and a codeAllowedRepositories list for self-hosted sessions; the repo list isn't applied on-device yet
Settings & configSession start can now place a session in a sandbox VM by default, and a new --unsandboxed flag opts out of that and of cloud-by-default
Permissions & safetyThe prompt draft can now tag edits with their origin and report changes, and extensions reach it through composer.attribute and attributeDraft
ExtensionsNew ui.view plugin capability switches the main view to an agent's transcript, and plugin calls judged under an outdated policy are refused
Extensions
CLAUDE_CODE_FORCE_SANDBOX no longer existsPermissions & safetyWant the reasoning? Read walks the 47 entries that probably matter to you, each one opening to what changed and why.
Read this release → Every row →Anything you can use today, anything that visibly changes, and anything worth poking at. One line each, open for detail.
Session start can now place a session in a sandbox VM by default, and a new --unsandboxed flag opts out of that and of cloud-by-default
Unclear It is unclear whether --unsandboxed is a flag users can pass or only an internal entry, and what it does.
New ui.view plugin capability switches the main view to an agent's transcript, and plugin calls judged under an outdated policy are refused
Unclear Whether the plugin system that offers ui.view is available to everyone is not stated.
New permission holds pause tool calls until the user writes, and stop plugin allow verdicts from loosening Cowork cloud web calls
Unclear Nothing shown sets the held state, so it is not clear what turns it on or whether it is in use.
Gateway, Bedrock, Vertex, OTLP telemetry, attribution and tool policy settings in the shared managed config now list Claude Code as a reader
Unclear It is not confirmed that Claude Code acts on these settings yet, only that they now list it as a reader.
What would fix it?
Smaller changes and internals, grouped as the pipeline found them. Nothing is dropped, it is only further down.
12 more of these are in What probably matters to you, above.
PreToolUse and PermissionRequest hooks now have input rewrites validated, fail toward refusing or asking, and can't switch mode in some cloud sessions
Unclear Whether a malformed hook change is refused in an interactive session when the remote setting is off, or only recorded, is not settled.
Resuming a workflow run can now be refused when its script is hidden, and resume notices show checked, cleaned names, ids and a manual resume call
Command plugins, marketplaces, hooks modules and /goal are held until organization settings are read; an unreadable policy now ends with a clear message
Unclear Which kinds of sessions wait for organization settings in this way is not clear.
Hook selection for organisation-managed setups now includes policy hooks delivered by plugins, and hooks that continue on failure can be marked contested
Unclear What being marked contested changes about whether a hook that continues on failure actually lets the action go ahead is not shown.
Interrupting Claude Code can now stop a running shell command in some cases, where before an interrupt always left the command running
Unclear Which commands are now stopped by an interrupt, for example foreground or background ones, is not shown.
When creating a cloud session on a self-hosted runner, a mode or effort from your settings is no longer sent, with a notice to use --permission-mode
The built-in web-fetch helper now uses Haiku when certain model and provider checks pass, and is told not to retry sites that refuse it
Unclear Which model and provider setups qualify for Haiku is not stated.
A thinking setting of disabled can now be changed to adaptive in some cases, with summarised display
Unclear It is not known when this conversion happens, so it is unclear which sessions get thinking despite a disabled setting.
Plugin hooks now run only after an extra admission check, and under managed-hooks-only setups they can be held until the organization answers
Unclear It is not clear what conditions make a plugin hook wait or fail the check, so whether a given plugin's hooks are affected is unsettled.
In a hosted session, only an organization's settings can place a mod above the user tier, and Claude Code now says so
Plugin helper entries are now held until your organization answers, and the plugin list hides entries whose plugin fails a check
Unclear It is not clear what check decides which plugin entries are hidden.
If link requests are switched off, an MCP task that asks for input by link is now cancelled instead of being handled
Plugin command sources and marketplace headersHelper commands no longer run in a tool server and wait until organization settings have loaded
A hooks module that was approved under a different policy text than the one now in force is skipped instead of run
Two new env vars cap how many PDF pages Claude reads per message and per turn after the context has filled once
Unclear It is not clear what sets the enabled-and-enforced state, so setting the variables alone may not turn the limit on.
Hook input now leaves transcript_path blank whenever a certain check holds, and agent hooks are no longer told to ignore that field
Unclear Which kinds of session the condition covers is not known.
A set_permission_mode request can now carry from_settings, and a runner child ignores settings-supplied modes and keeps its own
Unclear It does not say what makes a Claude Code process a runner child, so it is unclear which setups now lose a mode from settings.
allowedPluginMcpServers is now marked as read by Claude Code, and network endpoint lists now drop entries meant for other platforms
When a message is only pasted text, Claude treats the paste as your request, but still ignores instructions buried inside pasted documents
claude remote-control now asks up front about serving a folder unattended and fails on --sandbox, --no-sandbox or a sandbox config key
On Linux, Claude Code now refuses to write a file whose path looks like a Mac or Windows user path, and nothing is written
If a URL failed earlier in the session, WebFetch skips a quick second attempt and tells the model to say the page could not be read
Skill shell commands meant for tools other than Bash can now be rejected, and skill descriptions go through a normaliser
Unclear It is not clear which commands the new check rejects or when a skill's allowed tools list ends up empty.
deviceToolsEnabled and the selfHosted credential keys are retired, and gateways now reject settings that run Desktop sessions off the device
Startup hooks and plugin hooks now wait for organisation managed settings, and are skipped with a warning if those never arrive
Unclear Whether this waiting applies to every setup or only to some is not stated.
Signing in now checks the new account against the pinned organisation and rolls the sign-in back if it is refused
In auto mode, publishing a file with edits is now denied if the automatic permission check cannot see those edits
Unclear It does not name the tool or command that does this publishing, so it is unclear where you would run into the denial.
An organisation setting for plugins is now marked as read by both the desktop app and Claude Code, with a stand-in for when it cannot be read
Teleporting a session to the cloud is now refused when the branch has untracked files, with advice to commit and push them or run git stash -u
Before Remote Control starts, Claude Code can now ask once whether to allow tool calls while you are away, unless --allow-unattended-tool-calls is passed
Unclear It is not clear what starts this question or whether it is shown when Remote Control starts in this release.
Claude Code's Bash tool now rejects commands that match a new check with error code 13, unless you typed the command yourself
Unclear Which commands the new check blocks, and the message shown when it does, is not clear.
The model can no longer write into Claude Code's jobs folder, and some other internal folders are protected when a server-side switch allows it
Unclear It is not clear whether the daemon, shell-snapshots and session-env protections are switched on by default.
A PostToolUse hook that ran in another environment can no longer replace a tool's result, and Claude is told it read the original
Claude Desktop managed settings now refuse a config URL whose host, user name or password looks like an API key or token, and list crash dumps as data sent
Transcript entries whose type starts with api-request are no longer copied to the transcript mirror, and empty batches are not sent
If Claude Code hits an internal fault while checking permission for a tool, the action is now denied instead of carrying on
disableBundledSkills and coworkEgressAllowedHosts are now marked in the settings schema as read by both Claude Desktop and Claude Code
Unclear Whether Claude Code now enforces these two settings differently, or only the schema labels changed, is not stated.
Claude Code now enforces the JSON Schema rules for unique items, required contents and property counts, which it used to ignore
Signing in now checks the new token against the organisation required by managed settings before saving it, and fails if it does not match
Some errors raised while Claude Code works out which hooks apply now stop the operation instead of being logged and skipped
Claude Code will not restart a session onto a new version if a file passed with --settings or --mcp-config can no longer be read
A command containing the invisible characters U+FEFF or U+180E is now refused, with a note on how to write them safely
Unclear It is not clear which command tool runs this check, the Bash tool or the PowerShell tool, or whether the check is always on.
If a web address contains something that looks like an API key, Claude Code now refuses to read it and reports no hostname
Command hooks run through this path now stop when their configured timeout in seconds runs out
Unclear It is unclear whether this applies to every command hook or only to one particular way of running them.
CLAUDE_CODE_FORCE_SANDBOX no longer exists#The CLAUDE_CODE_FORCE_SANDBOX environment variable has been removed, and remote bridge setup no longer passes a sandbox option
Unclear It is not shown whether the sandbox can now be forced some other way.
Remote control no longer passes or reports a sandbox option and now records where its sign-in token came from when a refresh fails
Published verbatim by Anthropic for v2.1.300. Text is unmodified from the upstream changelog. Everything else on this page came out of the bundle instead, which is why the two lists don't match.
Of these 88 bullets, 18 name something an entry on this page also names, 28 name something no entry here does, and 42 name nothing specific enough to line up either way. The pairings are made on names both sides wrote down, a flag or a setting or a slash command, so read one as probably the same thing rather than as a fact, and read the middle number as candidates rather than as a miss count.
virtualText to a mod's prompt decorations: text drawn in the prompt box that never enters the draft or the submitted prompt
No entry names this cache_creation_5m_tokens and cache_creation_1h_tokens to the OpenTelemetry api_request event, splitting cache writes by lifetime
Probably cache-creation-5m1h-token-split-in-api-telemetry $.ui.view({ agentId }) for mods: shows an agent's transcript in the main view when you ask, and $.ui.view({}) goes back
No entry names this dir prop on a mod's Text for right-to-left text: on Windows and in VS Code's terminal its rows read in that direction and sit flush right
No entry names this /advisor on Claude Platform on AWS; no advisor runs until you pick one with /advisor, advisorModel or --advisor, and CLAUDE_CODE_DISABLE_ADVISOR_TOOL=1 turns it off
No entry names this gen_ai.provider.name, gen_ai.conversation.id, gen_ai.tool.name and error.type
Probably tool-error-span-attribute-for-failed-non-interrupted-tool-c, otel-spans-get-gen-ai-attributes-prompt-redaction-handle, tool-and-llm-telemetry-spans-gain-gen-ai-attributes prompt and model in the agent.spawn mod hook, which could only refuse workflow agents before
No entry names this prompt.change mod hook, raised when the prompt box's text changes without a typed edit: a recalled history entry, a clear, an undo, or a plugin's fill
Probably new-telemetry-and-event-names, mod-event-table-gains-promptchange-policy-decision-gets-se, prompt-change-hook-dispatcher-with-loop-stall-guard /autocompact warning about high token usage when you choose a smaller auto-compact window
Probably autocompact-telemetry-expanded-and-the-high-token-warning-m claude --help naming a removed tool in its description of --restricted
Probably restricted-help-text-no-longer-names-repl claude agents not refreshing any pull request's status when the data for one of them came back incomplete
No entry names this /tasks shortly after finishing, leaving no way back to the conversation
No entry names this ENABLE_TOOL_SEARCH=auto) turning off part-way through a conversation, for example while MCP servers reconnect after a resume
No entry names this forceLoginOrgUUID leaving the refused organization's token stored, or live after /login
Probably remote-control-credential-renewal-explains-sign-in-failures, cloud-agents-not-logged-in-error-can-carry-a-custom-message, oauth-login-paths-gain-an-org-pin-verification-step-before-c, forceloginorguuid-check-refactored-with-a-reusable-verifier, cloud-session-creation-has-a-separate-console-profile-path Artifact("(unprintable path)") while a call's input arrives; its row now shows "Waiting…" and "Running…" as Bash's does
No entry names this /branch dropping messages, and claude purge skipping subfolder projects, when they contain a Unicode line or paragraph separator (U+2028, U+2029)
No entry names this WorktreeCreate and WorktreeRemove hooks not running when Claude Code starts with --worktree (not yet with --tmux)
Probably worktree-pr-creation-supports-a-forge-and-uses-head-commit, worktree-creation-plugin-worktreecreate-hooks-registered-fi, pr-worktree-fetch-verifies-the-fetched-ref-tengu-melodic-oc, plugin-worktreecreate-hooks-are-now-registered-before-work #100905[BUG] Plugin `WorktreeCreate` hook still ignored for `claude --worktree` on 2.1.296 (#46664 closed as stale, never fixed) Open
tool_result events missing git_commit_id after git commit -q in repositories using reftable ref storage
No entry names this prompt.submit hook rewrote or dropped still being recorded as typed in the OpenTelemetry user_prompt event and interaction span, and in session title requests
Probably user-prompt-telemetry-helper-and-orphan-answer-skip-flag-att, userpromptsubmit-hook-detection-restricted-by-a-new-module-c, mod-event-table-gains-promptchange-policy-decision-gets-se model: aliases like haiku now resolve, a subagent's model no longer overrides it, and an unavailable model falls back to the caller's instead of allowing the action
Nothing to match on --worktree sessions sometimes starting from the wrong commit when a tag or branch was named like a commit ID, and their starting git status naming a branch other than the new worktree's
Probably worktree-pr-creation-supports-a-forge-and-uses-head-commit, worktree-creation-plugin-worktreecreate-hooks-registered-fi, pr-worktree-fetch-verifies-the-fetched-ref-tengu-melodic-oc, plugin-worktreecreate-hooks-are-now-registered-before-work --worktree failing to reopen a worktree in repositories using git's reftable ref storage
Probably worktree-pr-creation-supports-a-forge-and-uses-head-commit, worktree-creation-plugin-worktreecreate-hooks-registered-fi, pr-worktree-fetch-verifies-the-fetched-ref-tengu-melodic-oc, plugin-worktreecreate-hooks-are-now-registered-before-work #96060[BUG] Reftable repos: CLI reads .git/HEAD directly, branch resolves to ".invalid" and PR status is dropped from the statusLine payload Open
claude remote-control servers giving up on renewing their connection after a brief sign-in error, such as a locked keychain; they now retry, and warn when you need to run claude /login
Probably remote-control-credential-renewal-explains-sign-in-failures, cloud-agents-not-logged-in-error-can-carry-a-custom-message, remote-control-asks-about-unattended-tool-calls-up-front-in, remote-control-rejects-sandbox-no-sandbox-instead-of-acc, cloud-session-creation-has-a-separate-console-profile-path #95262remote-control session permanently 401s after token refresh fails; /login cannot reach the process Open
--worktree failing in a worktree of a bare repository when origin's default branch can't be fetched; the fallback base is now the current checkout's HEAD, not the main repository's
Probably worktree-pr-creation-supports-a-forge-and-uses-head-commit, worktree-creation-plugin-worktreecreate-hooks-registered-fi, pr-worktree-fetch-verifies-the-fetched-ref-tengu-melodic-oc, plugin-worktreecreate-hooks-are-now-registered-before-work prompt.edit hook not being raised for a newline entered with the chat:newline keybinding (Ctrl+J, or a rebound Enter)
No entry names this claude daemon stop --any printing "stopped" while a stuck background service kept running: it now waits, force-kills the service, and says so with exit 1 if it is still there
No entry names this claude daemon stop exiting with no output, and leaving sessions listed as running, when telemetry is turned off
Probably daemon-stop-now-escalates-sigterm-to-sigkill-and-explains-st /healthz or SIGTERM shutdown) while nobody reads their output; logs keep the latest 256 KiB and count drops (not on Windows)
No entry names this plugin-authoring skill: Claude scans a mod's files for keys before opening them, and asks before scanning a folder outside the project
Probably secret-scan-tool-for-the-plugin-authoring-mods-built-in-pl CLAUDE_CODE_PROVIDER_MANAGED_BY_HOST sessions to ignore CA-bundle, tool-proxy, git TLS and git config variables like SSL_CERT_FILE in settings; under Claude Desktop, only in project settings
No entry names this forceLoginGatewayUrl names an Anthropic-operated gateway that serves the session's hostname; that gateway must be reachable at start
No entry names this claude gateway to stop at start, naming the key, on a desktop: policy key that Claude Desktop cannot use from a gateway, where it earlier started with a warning
No entry names this CLAUDE_CODE_GZIP_REQUEST_BODIES=0 to turn it off
No entry names this claude-api skill to leave Managed Agents sessions in place unless you ask to archive or delete them
No entry names this /sandbox mode options to read "Sandbox commands" instead of "Sandbox BashTool"
Probably sandbox-option-label-renamed, sandbox-menu-labels-renamed PermissionRequest hooks in Cowork cloud sessions: a setMode update in a hook's allow is now ignored; the rest of the approval still applies
Probably hook-updatedinput-rewrites-are-now-validated-against-the-too, managed-settings-startup-message-distinguishes-unloaded-hook, permissionrequest-hook-setmode-dropped-in-cowork-cloud-cod, hook-load-failures-a-managed-settings-hook-that-cant-be-lo updatedInput from PreToolUse and PermissionRequest hooks is now validated like Claude's own input: unknown keys are dropped, and input the tool rejects denies the call
Probably hook-updatedinput-rewrites-are-now-validated-against-the-too, updatedinput-refusal-message-varies-by-settings-source --sandbox and --no-sandbox flags from claude remote-control, which had no effect; passing either now exits with an error. Set sandbox.enabled in settings to sandbox sessions
Probably remote-control-asks-about-unattended-tool-calls-up-front-in, remote-control-rejects-sandbox-no-sandbox-instead-of-acc, gateway-inference-settings-are-now-read-by-claude-code-cod claudeCode.attachSelection setting to stop text selected in the editor from being added to your messages
No entry names this /c:/… path instead of a mention relative to the working directory
No entry names this /bin/ps directly instead of searching PATH for ps on every check
No entry names this @claude review comments from people with write access whose GitHub organization membership is private; their comments now start a review
No entry names this A model matched these bullets to the GitHub issues they fix, so a link can be wrong.
1 of 27 tool descriptions changed.
Claude Code, interactive mode
One prompt change in this release could not be quoted from the build, so no entry on this page describes it.
668 documentation changes were recorded within 24 hours either side of this release, nearest first. The closest 12 are below. They're here because they happened near this release in time. That's not a claim that this release caused the edit, or that the page documents anything in it.
The 57 literal strings found in the bundle, with the number of entries that name each one. Picking one searches for it. A name is here because this build's code mentions it, which is not the same as it working or being finished.
What's wrong with this entry?