MCP tools that take file inputs can now accept local file paths. Claude Code uploads the file and passes the tool a reference instead. The Read tool's new allow_large option lets Claude read an oversized text file when there is room left in the conversation. Hook commands now take an on_failure setting of continue or block. You can set CLAUDE_CODE_WORKFLOW_SUBAGENT_MODEL to choose the model subagents use during a workflow run. Administrators can switch off messaging beyond a user's own agents with the managed setting messagingBeyondOwnAgentsDisabled.
This build holds 11 features that are not switched on yet. Claude can ask a subagent for lower or higher effort than its own, but the option is off by default. A session moved to the cloud from auto mode is being prepared to stay in auto mode. Remote sessions gain a check that probes a stream after about 8 seconds of quiet. That check is controlled remotely and is not on yet. A new restriction could limit an agent to messaging only its own agents.
Messages you typed ahead are no longer lost when a turn is cancelled. They go back into the input box or to the front of the queue. A PreToolUse hook that blocks a tool now shows its full reason in the error. Workspace diffs now follow renamed files and show deleted files as removed. On Windows, MCP servers now get a chance to shut down cleanly before being force-closed. Resumed sessions no longer restore saved MCP tasks from the 2025-11-25 tasks protocol.
Written by our agent from the shipped bundle, not by Anthropic.
allowedProviders can now set ANTHROPIC_BASE_URL automaticallySettings & configCLAUDE_CODE_AGENT_PROGRESS_SUMMARIES, CLAUDE_CODE_SLEEP_COMPACT_AFTER_MS and CLAUDE_CODE_RESUME_SESSIONSTART_HOOKS_WAIT_MS are new and are read by Claude Code
SDK set_model requests can swap the system prompt only if a named model is running, failing with model_not_current otherwise; Remote Control refuses it
Sessions & agentsA managed setting, messagingBeyondOwnAgentsDisabled, now turns this feature off and overrides CLAUDE_CODE_HARBOR_KITE
Gateway policies can use a code block for Claude Desktop's Code tab, cannot mix it with cli or settings, and get clearer warnings about serve_to_desktop
Platform & internalsModel catalog entries can set min_claude_code_version, and older clients see the model as unavailable with a message naming the version to update to
Extensions.mcp.json files can now be skipped when loading MCP serversExtensionsWant the reasoning? Read walks the 38 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.
The Agent tool can take effort "lower" or "higher" to step a subagent's reasoning effort one level, behind a helper check and tengu_snappy_pelican
Unclear It is not settled whether CLAUDE_CODE_SNAPPY_PELICAN is read as true or false or whether any value at all, including 0, switches it on.
CLAUDE_CODE_AGENT_PROGRESS_SUMMARIES, CLAUDE_CODE_SLEEP_COMPACT_AFTER_MS and CLAUDE_CODE_RESUME_SESSIONSTART_HOOKS_WAIT_MS are new and are read by Claude Code
Unclear It is not clear whether the CLAUDE_CODE_RESUME_SESSIONSTART_HOOKS_WAIT_MS value limits how long resume waits or only sets when the still-running notice appears.
SDK set_model requests can swap the system prompt only if a named model is running, failing with model_not_current otherwise; Remote Control refuses it
A backgroundLaunchRule can now veto moving a foreground agent to the background, and new policy getters cover agent moves and cross-agent messaging
Unclear How these two values are set, and whether a user or administrator can set them, is not stated.
MCP tools that declare file inputs can now take local file paths, which Claude Code uploads and replaces with references
Unclear Which MCP servers pass the policy check that allows file attachments is not stated.
Model catalog entries can set min_claude_code_version, and older clients see the model as unavailable with a message naming the version to update to
Unclear It is not clear which manifest this field belongs to or whether Claude Code acts on it.
When a resumed session adopts a parked permission prompt, it can now use a saved matching answer instead of asking again, with inheritedAnswerOff to disable it
Unclear It is not clear which setting inheritedAnswerOff reads.
Skills can supply a formatForegroundForkResult hook to reshape foreground forked results, and failed background agents now record a cause
Hook commands accept an on_failure value of continue or block that cloud hook comparison now includes, and cloud launches report built-in tool count
Unclear It is not clear that setting on_failure changes anything yet, since nothing was found that acts on it.
A new restriction lets messages go only as plain text to agents in the same session or to the main agent
Unclear When this restriction is switched on is not known.
Background-task rules can now allow a running agent to be moved into the background, still respecting the block on background agents
Unclear It is not clear when moving a running agent into the background is available or how a user triggers it.
Claude Code gains a restriction that appears to stop an agent from messaging anything beyond its own agents
Unclear Which setting or mode turns this restriction on is not known.
CLAUDE_CODE_ADOPT_UNDERIVABLE_PARKED_PERMISSION is read as an on/off setting#CLAUDE_CODE_ADOPT_UNDERIVABLE_PARKED_PERMISSION is an on/off variable for handling permission requests left waiting in resumed sessions
CLAUDE_CODE_WEBSEARCH_CITATIONS_DECLARED setting for web search citations#A new environment variable, CLAUDE_CODE_WEBSEARCH_CITATIONS_DECLARED, joins the existing web search citations switch
Unclear What setting this variable changes in web search answers, and which conditions it needs, is not established.
CLAUDE_CODE_WORKFLOW_SUBAGENT_MODEL picks the model for workflow subagents#Setting CLAUDE_CODE_WORKFLOW_SUBAGENT_MODEL chooses which model subagents use during a workflow run, separately from other subagents
The Read tool accepts allow_large, letting Claude read an oversized text file in one go when there is still room in the conversation
The tether live check and its kept-reminder mode now also turn on when a further condition is met, besides the environment variable or server setting
Unclear What the tether reminder does for a user, and what the new condition checks, is not stated.
A reference document bundled with Claude Code now covers managed agent outcomes in place of scheduled deployments
After a session restarts, Claude is now told when a command meant for another computer may not have been sent
Claude Code now counts turns since a person last wrote and can act once a limit is reached, but this is off unless switched on remotely
Unclear What happens when the limit is reached is not stated.
When you move a session to the cloud from a terminal in auto mode, Claude Code now asks the cloud session to stay in auto mode
Agent and plugin agent definitions accept autoCompactWindow, so a subagent can compact its own conversation at its own token threshold
A new task type for MCP monitors can be stopped, which also stops the monitor streaming updates for it
Unclear Whether this kind of task can be started at all yet, and what switches it on, is not known.
The bundled managed-agents quickstarts gain bug-hunter, explainer-video-maker, performance-tuner and watchlist-scanner, for 13 in total
Session and preview tunnels number their liveness pings, reconnect after a run of unanswered ones, and report dropped streams and oversized frames
A new x-claude-code-wiggly-dove request header can carry client version, permission mode, sandbox and git details, and is off by default
show_plan_tip and read_research_report were added to a list of tool names alongside current_time and others
Starting a cloud session with --tools now checks the list and gives a clear error for lists, environments or existing sessions it cannot honour
Unclear Which --tools lists are now accepted for cloud sessions is not stated.
The context given to plugin hooks gains isDispatching, and the special handling of ui.render and ui.focus events now goes through a shared helper
Unclear What isDispatching reports and what the change means for a plugin in practice are not stated.
A PostToolUse hook can now replace an MCP tool's result using updatedMCPToolOutput when it does not set updatedToolOutput
The plugin install hint now uses --replace and --scope, and disabled MCP servers stay disconnected when Claude Code reconciles its server list
skillOverrides now applies to skills that come from MCP servers#Settings in skillOverrides now apply to skills supplied by MCP servers, matched by the skill's name or its alias names
CLAUDE_AUTO_BACKGROUND_TASKS is now read from Claude Code's own configuration#Claude Code now reads CLAUDE_AUTO_BACKGROUND_TASKS from its own configuration values instead of straight from the shell environment
Unclear Whether the variable can now be set through Claude Code settings, and whether values like 0 or false are read the same way, is not shown.
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.
When a reply hits the output limit during thinking, Claude Code nudges differently and, if recovery fails, suggests lowering the effort level
Claude Code now limits how many tools one MCP server can list, and MCP tools marked for file grants get a wrapper and skip XML-tag stripping
Unclear The maximum number of tools Claude Code keeps from one server is not stated.
A flagged mode limits SendMessage to agents in the current session, with matching description wording, a shorter error and peer-message handling
Session info and initialize replies, including Remote Control's, now carry the Claude Code version that runs the session
Unclear Whether this version of Claude Code actually fills in claude_code_version is not known.
Claude Code's formatted text display gains a windowed mode, likely for showing part of long content
Unclear What makes Claude Code switch to the windowed display is not known.
Remote session rows and failure reasons now use the post-turn summary's status detail, recent action and needs-action fields
Pressing left arrow to background a session now stops the turn in progress, and queued messages that arrive late are not sent
Unclear It is not known which sessions meet the condition that lets Claude Code stop the running turn.
A cloud Bash dialog can now appear during remote sessions, and the Background tasks dialog has its own cloud tasks context
Unclear It is not clear exactly which remote-session setting makes the cloud Bash dialog appear.
Claude Code can send a request to fetch a skill from an MCP server, and skill sync stops early when nothing is readable
Unclear It is not clear how or when Claude Code sends the skills/get request, so whether you can fetch skills from an MCP server yet is unknown.
For the Chrome auto-mode classifier floor, remote sessions now read the server setting a different way, while CLAUDE_CHROME_CLASSIFIER_FLOOR still overrides it
Unclear What the different way of reading the setting changes for remote sessions is not stated.
Claude Code can now block a hook that did not run and name its plugin, unless a remote switch makes this non-blocking
Unclear It is not clear what causes a hook to be refused this way.
Skills run as separate tasks and hooks from synced plugins can now be refused when plugin sync is off or has no folder
CLAUDE.md files, rules and project memory can now be skipped under a new condition, not only when CLAUDE_CODE_DISABLE_CLAUDE_MDS is set
Unclear What the new condition is, and so which sessions skip these files, is not known.
The instructions for consolidating Claude's memory no longer point it at session logs, only at a sessions folder if one exists
Unclear Whether this memory consolidation step runs for any given user is not known.
.mcp.json files can now be skipped when loading MCP servers#Claude Code has a new condition under which it skips reading project .mcp.json files and uses a server list from elsewhere
Unclear What the check tests, and so when project .mcp.json files are skipped, is not stated.
Claude Code's file upload code now checks files after sending, refuses credential files, limits sizes and is tied to an organization policy
Unclear It is not known how this upload is started or whether the allow_cobalt_plinth policy allows it by default.
A new denyMatchNames list adds extra names for deny-rule matching on prompt-type commands and feeds reserved labels for MCP-loaded tools
Unclear Which commands get these extra names, and what the names are, is not known.
Plugin dependency matching now accounts for same-name plugins synced from your claude.ai account, and its errors explain the conflict and how to fix it
Runners now check remote.origin.url in cloned repositories, correct or log it, and fetch through a temporary remote with credentials passed separately
Hooks are labelled personal or checkout by source, checkout hooks cannot silently replace settings hooks, and a failing matcher stops its hook running
Unclear What the personal and checkout labels change about how those hooks run is not shown.
Self-hosted runners now check governed repositories' origin, note pending settings, and report several new diagnostic stages
project_write by local_path now attaches machine-readable codes to its errors and refusals, and reports when a file is saved
Claude Code now checks pasted-text placeholders before sending and refuses with a "Not sent" message when the paste is blank or no longer available
Unclear Which part of Claude Code shows this refusal, such as the VS Code extension, is not stated.
Compaction plans now compute reclaimableBytes, a new check can skip compactions that would free little, and transcript loading reports scan details
When a resumed remote session cannot read its saved state, Claude Code now exits with an error instead of starting a turn with nothing
Each task in a task list now carries its owner, what it blocks and what blocks it, and optionally extra metadata
Claude Code's internal web search call now has a longer prompt telling it to search once and reply with a brief summary
If a delayed run hands your prompt back, Claude Code now restores your text, pasted content and editor selection
Claude Code now checks at startup whether your own settings differ from those cloud sessions use, and points to claude apply-project-settings
Unclear What decides whether this startup check runs at all was not identified.
File reads can raise their size and token limits to fit remaining context, and cut-off messages now say how much room there was
Unclear It is not stated what allows a read to use the raised limits, so whether this is active is unknown.
Claude Code now refuses a permission answer from a host program if it would lose deny or ask rules that answer added
Errors for files or line ranges too large to read can now have extra text added, though their wording is unchanged for now
With refuse_if_usage_limited, a rewind_conversation request now gets a separate turn_running refusal if a turn is still running
Sending a message to a project's MCP servers after they are closed now fails with "The repo's MCP servers have been closed"
In fleet view, failed or taken jobs no longer become the next tab, and quitting with no session left to open prints a quit line
Unclear Whether these session tabs are available to all users, or only behind a setting, is not stated.
On resume, Claude Code now traces back through the transcript by response, so some turns it used to call interrupted now count as complete
Unclear Which interrupted turns now count as complete is not spelled out.
Claude Code now works out which facts about an artifact changed, such as its owner or audience, and describes them in plain wording
Resuming a background agent now goes through a handover step and stops with an error if it is closed or not taken
If an allow answer's updatedPermissions are too large or malformed and might hide a deny or ask rule, the call is now refused
When a remote session catches up on saving its conversation history, Claude Code now retries and no longer starts saving after a failed catch-up
In hosted sessions, an incoming update over a size limit is now swapped for a smaller version or dropped instead of handled as is
Unclear Whether a size limit is ever actually set for hosted sessions is not known, and without one this changes nothing.
Claude Code can now tell the other end of a remote tools connection it is still working when a tool call runs quietly for a while
Unclear Whether the still-here signal is switched on by default is not known.
Claude Code now ships a 'from quickstart' onboarding guide for managed agents in place of the 'from URL' one
Claude Code now tracks remote tool calls that go silent and can reapply an approval answer given before a restart, running PermissionRequest hooks first
Unclear When this recovery applies and whether it is switched on are not stated.
Claude Code can remove hooks from synced plugin skills, and records why a plugin skill or command was refused or run without its hooks
Unclear It is not stated what turns plugin sync on, so whether this applies to a given account is unknown.
Under some conditions, the Bash tool's instructions now say nohup or & processes are usually stopped a few minutes after the reply
Unclear The conditions that decide whether this sentence is added are not described.
When a PermissionRequest hook allows with permission updates, Claude Code withholds the allow if any stricter rule was lost
Unclear It is not stated what Claude Code does instead of allowing, such as asking or denying.
A permission answer that replaces rules with deny or ask rules now adds them for the session instead of dropping them
Claude Code now refuses paths in its temp folders or in folders background sessions write to freely, including for cloud credential settings
Unclear It is not clear exactly when this check runs or what Claude Code does when it refuses a path.
The background task count in Claude Code's footer now adds remote background tasks, and the agents hint is hidden in some states
Claude Code can tell you a background agent cannot start or resume in this session, and suggests the foreground or a new agent
Unclear What makes a session unable to run background agents, and whether that happens in normal use, is not stated.
The background tasks label at the bottom of the screen now adds another count and hides only when both counts are zero
Unclear It is not clear what the extra count measures, so which items now make the label appear is unknown.
Claude Code no longer grants a permission when saving it would discard a rule that blocks or restricts an action
Unclear It is not clear what happens to the request after the approval is withheld, for example whether you are asked again.
For some background MCP tasks, Claude is told the result may only arrive after you send another message, and not to promise to report back
Unclear It is not clear when Claude Code uses this wording rather than the earlier note.
Starting a remote session now names the session only when no built-in tools are given, after a check that can stop it
Unclear It is not clear what the new environment check refuses, or what a reader starting a remote session will notice.
On Windows, path checks now also try /c/... and /cygdrive/c/... paths converted to the usual C:\... form
Claude Code shows notices saying why hooks are off, covering unreadable managed settings, bare mode, safe mode, managed policy and --settings
NotebookEdit now refuses to edit a notebook that is not valid UTF-8 and leaves the file untouched
When two plugins share a name only the first runs hooks; the other's skipped hooks are now recorded and shown alongside plugin warnings
Repository setup now adds origin with the plain upstream URL instead of the credentialed one, and git subprocesses no longer inherit GIT_CONFIG
Unclear What Claude Code does when the ANTHROPIC_CUSTOM_HEADERS values match or differ is not known.
Large MCP results are no longer cut inside subagents, cut text moves to a shared store, and the restore step no longer assumes MCP
Task update and notification messages in the SDK stream now include a failureKind and have output paths and raw error text sanitized
Unclear It is not clear exactly where the failure kind appears in these messages.
Resuming now handles overdue or summarized permission prompts and rate limits, records more detail on cancelled prompts, and drops a classifier event
Unclear What the new handling of overdue or summarized prompts changes for someone resuming a session is not stated.
Task owner, metadata and blocks/blockedBy dependency links now survive task creation, hand-off, compaction and resume
If the hooks module that made a call has since been reloaded or removed, Claude Code now refuses the call with an error
Claude Code now skips looking for skill folders deeper in your project when a new check is true
Unclear What condition turns off this skill folder search is not known.
Claude Code's Edit tool has a new condition under which it refuses an edit, checked after its stale-file checks
Unclear The condition behind the new refusal and the message it shows are not stated.
Published verbatim by Anthropic for v2.1.296. 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 79 bullets, 15 name something an entry on this page also names, 27 name something no entry here does, and 37 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.
code key to the Claude apps gateway's managed.policies[]: the same settings as cli, also applied in Claude Desktop's Code tab; beside desktop, it turns on Claude Desktop's gateway mode
No entry names this autoCompactWindow to subagent frontmatter and --agents definitions, so a subagent can auto-compact earlier than the main conversation's window
Probably agent-definitions-can-carry-an-autocompactwindow-passed-to, agent-auto-compact-threshold-now-takes-a-second-window-sourc, plugin-agent-frontmatter-gains-autocompactwindow-passed-thr CLAUDE_CODE_WORKFLOW_SUBAGENT_MODEL to run every workflow agent on one model while other subagents keep theirs
Probably new-env-var-claude-code-workflow-subagent-model-overrides-th CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS environment variable to set a longer maximum delay for the backoff when retrying an overloaded (529) request
No entry names this /plugin on a plugin whose hooks are left out because another enabled plugin has the same name
No entry names this allow_large option to the Read tool so Claude can read a text file past the usual size limits in one call when it needs the whole file and the context has room
Probably read-tool-allow-large-override-messages-added, read-tool-gains-allow-large-option-sized-to-remaining-contex PreToolUse hooks that deny a tool call with "continue": false, and managed prompt hooks that block one, refusing the call but not ending the turn
Probably pretooluse-hook-stop-reason-is-appended-correctly, pretooluse-hook-stop-attachment-helper-and-422-to-400-error updatedMCPToolOutput in some sessions
Probably posttooluse-hooks-can-rewrite-mcp-tool-output-via-updatedmcp .mcp.json or plugin MCP server that was switched off for that folder, after changing directory or reloading plugins
Probably telemetry-for-auto-approved-project-mcp-servers, project-mcp-config-skipped-when-a-flag-checked-helper-is-on allowedProviders with "gateway" locking out laptops that name that gateway in user settings
Probably managed-settings-env-now-injects-a-computed-anthropic-base-u forceLoginMethod to gateway with no forceLoginGatewayUrl (regression in 2.1.295)
Probably gateway-login-method-now-counts-as-gateway-required --teleport opening an empty conversation when the session's history could not be read
No entry names this #95873[BUG] Regression since ~2026-09-12: Remote Control sessions have 0 teleport-events, so opening them from another machine (VS Code Remote tab / --teleport) yields an empty session (worked 2026-08-26 → 09-11) Open
52;c;… escape sequence printed on screen after copying in older VTE-based terminals such as MATE Terminal
No entry names this /diff panel or dialog was open
No entry names this #99026[BUG] Built-in /diff pane holds mod toasts, undocumented and contradicted by the screen map Open
UserPromptSubmit hook or a mod's prompt.submit hook ending headless sessions, clearing the typed prompt, or letting the unchecked prompt through
Probably prompt-submit-hook-interrupted-while-checking, input-hooks-interrupted-by-abort-now-return-interrupted-resu, prompt-submit-hook-handling-blocked-prompt-tracking-and-abo, interrupted-while-prompt-checked-handling, userpromptsubmit-hook-cancellation-distinguishes-interrupts claude self-hosted-runner printing misleading errors when registration is refused: it now names the org admin setting, or says a restarted on-demand runner needs a fresh work order
No entry names this --capacity above 1
No entry names this $ methods running in the main session's working directory instead of the calling agent's, and ignoring the turn their calling hook holds
Nothing to match on $.agent.register succeeding from a mod's hook that was still running after the mod was reloaded or removed; the call is now refused
No entry names this $.http.fetch in mods refusing a HEAD request when the response declares a Content-Length over the 4 MiB body limit
Probably custom-headers-check-function-replaced-and-git-subprocess-e, api-client-adds-per-request-extra-headers-and-awaits-a-heade, git-commit-detection-from-reflog, web-fetch-head-requests-skip-the-content-length-size-refusa, new-x-claude-code-wiggly-dove-request-header-off-by-default claude plugin marketplace add, marketplace update and plugin install failing with an internal error for a marketplace named like constructor; add now refuses such names clearly
Probably plugin-install-gets-replace-for-same-repo-duplicates-mcp constructor or prototype being deleted by the next save of that plugin's options
Nothing to match on CLAUDE_CODE_RESUME_INTERRUPTED_TURN re-running a finished turn after a restart when an MCP tool result had ended that turn
No entry names this BASH_ARGV0 shell variable and then use it; these now prompt for approval
Probably bash-variable-list-adds-bash-argv0 ← being run twice, once unseen in the foreground, when the background service was slow to answer
Nothing to match on ← moved the session to the background; it is still not sent, but ↑ now brings it back
Nothing to match on /clear in headless sessions
No entry names this /mcp or claude mcp when an MCP server needs authentication
Probably auth-needed-prompt-for-remote-sessions-without-oauth, remote-environment-classifier-and-ccr-artifacts-endpoints /code-review ending on a raw JSON array in cloud sessions, the Agent SDK and IDE integrations; the findings now print as a numbered list
No entry names this tool_result events missing git_commit_id after git commit -q or git -C <dir> commit
No entry names this #95934OTel: git_commit_id still silently dropped for `git -C <path> commit` and quiet/redirected commits (2.1.278) — recurrence of #77237 Open
CLAUDE_CODE_TRANSCRIPT_LOCAL_GC dropping a message from the saved transcript when its text held a Unicode line or paragraph separator (U+2028, U+2029)
No entry names this claude purge leaving a prompt in the history file when its text held a Unicode line or paragraph separator (U+2028, U+2029)
No entry names this claude plugin install failing for GitHub owner/repo plugin sources on machines with no GitHub SSH key; the clone is now retried over HTTPS
No entry names this rm -rf /c/Users/<name> in Git Bash not asking in bypass permissions mode
No entry names this code settings, the reason shows as a reply
Nothing to match on --debug output to name unrecognized frontmatter fields in custom agent files, with a hint for likely typos
No entry names this --debug output for hooks: command hooks on tool calls, prompts, SessionStart and Stop now log their command, plugin, outcome and duration when they finish, so a slow hook is identifiable
No entry names this CLAUDE_CODE_TRANSCRIPT_LOCAL_GC: a large transcript file is no longer rewritten when that would free under 10% of it
No entry names this /cost, the status line, --max-budget-usd and the SDK's cost figures to price Sonnet 5.5 cache reads at $0.10 per million tokens (was $0.20)
No entry names this ←: a turn or ! command started while the session moves to the background is now stopped instead of finishing out of sight
Nothing to match on claudeCode.spinnerVerbs missing from the extension's settings, so settings.json now completes and validates it; a malformed value no longer breaks the chat panel
No entry names this @browser, as the terminal does; allowing a site for the session stops repeat asks
No entry names this prefersReducedMotion setting
No entry names this A model matched these bullets to the GitHub issues they fix, so a link can be wrong.
910 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 43 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?