chore(deps): update dependency @anthropic-ai/claude-agent-sdk to v0.3.263 - #44
chore(deps): update dependency @anthropic-ai/claude-agent-sdk to v0.3.263#44renovate[bot] wants to merge 1 commit into
Conversation
|
PR author is in the excluded authors list. |
|
|
Review the following changes in direct dependencies. Learn more about Socket for GitHub.
|
1ad176e to
9f06cac
Compare
🟢 WAVE BugBot — clearNo confident findings on the changed lines. Local review · $0 inference · wave-dispatch |
9f06cac to
1ea24a4
Compare
|
All alerts resolved. Learn more about Socket for GitHub. This PR previously contained dependency changes with security issues that have been resolved, removed, or ignored. |
1ea24a4 to
cda8484
Compare
There was a problem hiding this comment.
Risk: low. Cursor Bugbot is not configured on this PR; evaluated using remaining signals. Approving this lockfile-only @anthropic-ai/claude-agent-sdk bump (0.3.168→0.3.197) with clean WAVE BugBot and Socket security results. No reviewers assigned.
Sent by Cursor Approval Agent: Pull Request Router and Approver
cda8484 to
946235b
Compare
946235b to
1f5215e
Compare
1f5215e to
4e0d82e
Compare
b528e55 to
9464eb6
Compare
beead6e to
677b05f
Compare
239e120 to
d0a45f0
Compare
d0a45f0 to
37ed908
Compare
|
Skipping PR review because a bot author is detected. If you want to trigger CodeAnt AI, comment |
89d5cb6 to
c1e690d
Compare
|
Your free trial PR review limit of 300 PRs has been reached. Please upgrade your plan to continue using CodeAnt AI. |
c1e690d to
d8be250
Compare
|
Your free trial PR review limit of 300 PRs has been reached. Please upgrade your plan to continue using CodeAnt AI. |
d8be250 to
6418e5e
Compare
|
Your free trial PR review limit of 300 PRs has been reached. Please upgrade your plan to continue using CodeAnt AI. |
6418e5e to
aab7ab2
Compare
|
Your free trial PR review limit of 300 PRs has been reached. Please upgrade your plan to continue using CodeAnt AI. |
aab7ab2 to
d708178
Compare
|
Your free trial PR review limit of 300 PRs has been reached. Please upgrade your plan to continue using CodeAnt AI. |
d708178 to
7aec23a
Compare
|
Your free trial PR review limit of 300 PRs has been reached. Please upgrade your plan to continue using CodeAnt AI. |
7aec23a to
cd1e06f
Compare
|
Your free trial PR review limit of 300 PRs has been reached. Please upgrade your plan to continue using CodeAnt AI. |
cd1e06f to
4b07529
Compare
|
Your free trial PR review limit of 300 PRs has been reached. Please upgrade your plan to continue using CodeAnt AI. |
4b07529 to
678c404
Compare
|
Your free trial PR review limit of 300 PRs has been reached. Please upgrade your plan to continue using CodeAnt AI. |
678c404 to
4daf3e3
Compare


This PR contains the following updates:
0.3.168→0.3.263Release Notes
anthropics/claude-agent-sdk-typescript (@anthropic-ai/claude-agent-sdk)
v0.3.263Compare Source
v0.3.261Compare Source
pluginDelivery: 'initialize'to sendpluginsover stdin so the launch command line no longer grows with the plugin count (fixes Windows start failures with many plugins)query()throwing "Object not disposable" in runtimes without a nativeSymbol.dispose, such as Node ≤22vmcontexts (Jest'snodeenvironment, vitestvmThreads/vmForks) and Node <18.18v0.3.260Compare Source
user_message_uuidtothinking_tokenssystem messages, linking thinking progress to the user message that triggered the turnfirst_content_frame_ms,first_stream_post_ms,first_stream_post_ack_msandfirst_stream_post_wall_msfields to the success result message for remote-session latency breakdownsmanagedSettingsdisableAutoMode: "disable"(either spelling) being dropped by the restrictive-only filter instead of turning auto mode off for the spawned sessionrewindFiles()reporting success when no files could be restored (for example when checkpoint backups are missing); it now failserror_max_structured_output_retriesresults to append the last StructuredOutput tool error; validation errors now name the offending key, allowed values, and actual length or countrate_limit_eventto also re-emit during an exceeded window on repeat 429s (about once per 30 seconds per limit window), so stream consumers can refresh stale rate-limit statev0.3.259Compare Source
user_message_uuidsbesideuser_message_uuidon a turn's first reply frame and result: every user message the turn answered, so a reply to several merged messages can be matched to eachpermissionPrompts: 'none'option to auto-deny permission prompts in sessions with nobody to answer them, without disabling auto mode's classifierv0.3.258Compare Source
v0.3.257Compare Source
thinkingTokenstoModelUsage(a subset ofoutputTokens), and fixed result-messageusage.output_tokens_details.thinking_tokensreporting 0 instead of the session's real counttool_use_result.resourceLinkson user messages carrying MCP tool results: theresource_linkblocks the tool returned, so hosts can render returned files without parsing the result textresource_linkstotask_notificationfor an auto-backgrounded MCP tool call that completed, listing the files it returned by reference; join to the call viatool_use_idmcp_reconnectandmcp_toggleacting on a same-named.mcp.json/~/.claude.jsonserver instead of the--mcp-configormcp_set_serversonemcp_toggledisable also removing the tools of a sibling MCP server whose name extends the disabled one's (disablingfoodroppedfoo__bar's tools)mcp_set_serversto also list a server whose connection attempt throws underadded(with afailedrow inmcp_status), not only undererrorstool_progressheartbeat (heartbeat: true) like other long tools; heartbeat frames never clear asubagent_retryindicator@anthropic-ai/claude-agent-sdk/browser) never streaming any messages on engines without nativeSymbol.dispose(Safari/iOS, Firefox ESR, older Chromium)task_notification-pgiving up on a long-running background subagent without actually stopping it, sobackground_tasks_changedkept listing it and events for it arrived after itsstoppednotificationdetailoption toQuery.getContextUsage():'summary'answers from the last response's usage and local estimates without per-category token-count API calls (default'full')v0.3.252Compare Source
v0.3.251Compare Source
v0.3.250Compare Source
v0.3.248Compare Source
timeoutfor SDK-hosted MCP servers (createSdkMcpServer({ timeout })), overridingMCP_TOOL_TIMEOUTfor that server's tool callsv0.3.247Compare Source
ambientflag totask_started,task_notificationandbackground_tasks_changedtask entries so hosts can exclude housekeeping tasks from activity indicatorspermissionModeon per-turnsystem/initframes reporting the mode at turn start instead of the live mode, so a mode switch right after submitting no longer sends a stale valuev0.3.246Compare Source
user_message_uuidto error result messages and to the first assistant message orstream_eventof each turn, linking a reply or failure to the user message that triggered itmodelUsage[*].costBasis('list' | 'managed' | 'unknown') reporting which price table each model'scostUSDwas computed frommodelPricingsupport in themanagedSettingsoption for hosts that setCLAUDE_CODE_PROVIDER_MANAGED_BY_HOST; an admin-managed settings source that setsmodelPricingstill winsperTaskStopAffordanceoption: when set,interrupt()aborts only the current turn and keeps background agents and workflows running; otherwise (and for one-shot string prompts) they stopv0.3.245Compare Source
v0.3.243queued_turn_countto result messages: the number of queued user sends still pending when the result was produced, so hosts know whether another turn and result will followmcp_statusreporting a remote MCP server as connected after its connection dropped; it now reports pending while reconnecting, then connected or faileddisableAllHooksalso disabling hook callbacks registered through thehooksoption; they now keep running, matchingallowManagedHooksOnlydocumentblock (or pageimageblocks forpagesreads) now arrives inside thetool_resultcontent instead of as a separateusermessage after itv0.3.242v0.3.241Compare Source
v0.3.240Compare Source
v0.3.239Compare Source
total_cost_usd/modelUsage.costUSDnow include the 1.1× US-only-inference (data residency) multiplier when the response reportsinference_geo: "us"total_cost_usd,duration_api_msandmodelUsageas of its release, not the turn-end snapshotSYSTEM_PROMPT_DYNAMIC_BOUNDARYin an arraysystemPromptbeing sent to the model as literal text on Bedrock, Vertex, Foundry, and gateway providersinitializeon a running process is now followed by abackground_tasks_changedsnapshot of the live background tasks, so reconnecting hosts see work that is still runningv0.3.238Compare Source
is_backgroundedandspawn_depthtotask_startedevents for subagent tasks (is_backgroundedalso on background Bash tasks)suppressOriginalPrompttoUserPromptExpansionhook output, matchingUserPromptSubmitcommand_lifecyclestaterefused: a cross-session peer message the session's receive-side policy declines now reports this terminal state instead of producing no lifecycle framesinitializeto an already-running CLI; the response now reportshooks_appliedCLAUDE_CODE_ENABLE_PROMPT_SUGGESTION=truenot keepingprompt_suggestionmessages on when the account is near, but not over, its usage limitvcs_state_changedpush events to emit one event per pushed branchv0.3.237Compare Source
v0.3.236Compare Source
PostToolUsehooks can returnhookSpecificOutput.classifierContext, a short host-asserted note about a tool call's result that the auto mode permission classifier reads alongside that resultv0.3.235Compare Source
v0.3.234Compare Source
bypass_permissions_disabledfromExitReasontype; the value was never emitted — TypeScript consumers with an explicitcasebranch get a compile error on upgrade (runtime unaffected)ApiKeySourcetype to include the valuessystem/initactually reports (ANTHROPIC_API_KEY,apiKeyHelper,/login managed key,none)vcs_state_changedevents report the directory the shell finished in (an innercdis reflected)origininjected by the host may declare the sending session's permission class (fromMode) so a same-class message is delivered to a recipient that runs without askingSDKSystemMessage(system/init) gains an optionaleffortfield: the session's applied effort level, ornullwhen none is sent. Set on Remote Control bridge init framesv0.3.233Compare Source
TaskCreate/TaskGet/TaskUpdate/TaskList,TodoWrite) are no longer in the default tool surface on Opus 4.8, Sonnet 5, Fable 5, Mythos 5, and newer models; name them in thetoolsoption or reference them inallowedTools(or setCLAUDE_CODE_ENABLE_TODO_TOOLS=1) to keep themv0.3.232Compare Source
tool_resultframes whose result carries_metanow emittool_use_resultas{ content, _meta }(matching main-loop frames) instead of a bare value/contextresult messages now carry a structuredcontext_usagepayload (newSDKContextUsagetype), so consumers can render the context-usage card without parsing the markdown tablevcs_state_changedevents now populate thebranchfield for push operations, sourced from the pushed refv0.3.231Compare Source
v0.3.229Compare Source
terminal_slash_commandsto the system init message so Remote Control clients can hide terminal-oriented commandsterminal_reason"api_error"instead of"image_error";StopFailureerror_detailsis"request_body_over_limit: …"v0.3.228Compare Source
AgentOutput):usage.output_tokens_detailsis now carried throughv0.3.227Compare Source
v0.3.226Compare Source
v0.3.225Compare Source
v0.3.224Compare Source
crossSessionInboundanddialogExpirysettings: cross-session messages sent to a session running with bypassed permissions are held for your approval, and messages to other sessions auto-deliversubkind: 'peer-send-message'to thetask-notificationmember ofSDKMessageOrigin, marking a notification raised by a cross-sessionSendMessagesource: 'archive'plugin config variant toSettings, withurland optionalsha256, for installing plugins from a zip over HTTPSSettings:decode: 'jwt'withmaskClaims,extract/onExtractNoMatchonenvVars, andawsPairs/sigv4for AWS SigV4 re-signing/resumeno longer cross projectsv0.3.223Compare Source
resumeDropsTurnoption: withresumeSessionAt, declares the turn a truncating resume intends to drop; the CLI refuses the resume if anything else would be discardedapi_error_status: 529, so SDK consumers can detect overload terminations structurally instead of matching message text-p/ SDKquery()withoutcanUseTool) now emitssystem/permission_deniedstream events when a tool call is auto-deniedusagevsmodelUsageon stream-json results:usageis main-loop-only and per-turn;modelUsageis cumulative, covers all query-pipeline calls, and is the field for cost accountingv0.3.222Compare Source
query({ sessionStore, resume })not carrying usersettings.json(apiKeyHelper,env,hooks,permissions) into the resumed subprocessv0.3.221Compare Source
skillsoption validation: malformed names (delimiters or control characters) and wildcard-form names are rejected with a clear error; useskills: 'all'to enable every skillmcpServersoption not being connected before the first turn, which caused the model to emit tool calls as literal textv0.3.220Compare Source
v0.3.219Compare Source
cancel_queuedto the interrupt control request (capabilityinterrupt_cancel_queued_v1): cancels queued and pending-dispatch messages alongside the abortfast_mode_disabled_reasonto result and init messages so SDK hosts can explain why fast mode is offDirectoryAddedlifecycle hook event to the control protocol, fired when a new working directory is registered mid-sessionfast_mode_statefrom the spawn-time model after a model switchsandbox.network.strictAllowlistto SDK settings types for deterministically denying non-allowlisted hosts in sandboxed commandsworkflowSizeGuidelineto SDK settings types for setting the advisory dynamic-workflow size guidelinev0.3.218Compare Source
SkillToolOutputnow reportsbackground: truewhen a forked skill was dispatched as a detached background agentapi_error_statusreporting null for rate-limit and overloaded errors delivered mid-stream; it now reports 429/529canonicalModelandproviderto eachmodelUsageentry in result messages so downstream billing can look up the correct rate table forcostUSDv0.3.217Compare Source
CLAUDE_CODE_MAX_SUBAGENT_SPAWN_DEPTHto allow deeper nestingCLAUDE_CODE_MAX_CONCURRENT_SUBAGENTS)v0.3.216Compare Source
skippedLinkscount torewindFilesresponses for paths the rewind safety guards refused to restore or deletetool_result_metasidecar to user messages (non_execution_kind,user_feedback) so consumers can classify denied, interrupted, or cancelled tool calls without string-matching result proseuser_message_uuidandrequest_sent_wall_msfields to the success result message for cross-host request-latency correlationv0.3.215Compare Source
v0.3.214set_permission_modenow rejects unrecognized permission modes with an error instead of silently adopting them; the'manual'alias is accepted at every ingresssubkind: 'scheduled-trigger'to thetask-notificationmember ofSDKMessageOrigin, marking deliveries that are the fired prompt of a user-configured scheduled taskapplyFlagSettings({effortLevel})now accepts'max'in its TypeScript type (runtime already supported it)interrupt()now carryaborted: true, so consumers can distinguish a mid-stream partial from a completed messagesubagent_typeandsubagent_retryfields totool_progressmessages so clients can show a subagent waiting out an API rate-limit retrysystem/initmessage'spluginsentries and thereload_pluginsresponse now include each plugin's manifestversion"fork"instead of"resume"when the session begins as a forkv0.3.213v0.3.212Compare Source
resumeSessionAtandsessionIdvalues being passed to the CLI as separate argv tokens; both now use equals-form (--flag=value)v0.3.211Compare Source
--replay-user-messageswith--include-partial-messagesemitting the turn-start user replay after the first content block instead of before the turn's content eventsSDKAssistantMessage.timestamp(ISO-8601) to the live stream, matchingSDKUserMessage; older emitters omit it, consumers should fall back to receive timeUSAGE_LIMIT_ERROR_PREFIXESand siblings) as@alphaexports for classifying rate-limit messages without hand-mirrored listsv0.3.210Compare Source
timedOutAfterMstoBashToolOutput, set when a command is auto-backgrounded on timeoutv0.3.209Compare Source
v0.3.208Compare Source
UserPromptSubmithook callback exceeding its timeout killing the entire query with an empty error; it now blocks the prompt with a clear timeout message and the session continuesextraArgsvalues that look like flags (e.g.resume: '--version') being parsed as their own CLI flags; dash-leading values are now bound with equals-form argvAbortControllerno longer accumulateabortlisteners on its signal after each completed querycreateSdkMcpServerdocs pointing at a nonexistent env var; the MCP tool-call timeout knob isMCP_TOOL_TIMEOUTv0.3.207Compare Source
canUseToolreturning{behavior: 'allow'}withoutupdatedInputbeing rejected as a deny with a raw ZodError message; the tool now runs with the original input per the documented contractAgentToolCompletedOutput) that matches the emitted object exactlyv0.3.206Compare Source
command_lifecycleframes to stream-json and SDK sessions, reporting each uuid-stamped message's terminal state (queued/started/completed/cancelled/discarded); zero-API results no longer report staleduration_api_msv0.3.205Compare Source
still_queued(UUIDs of queued async messages that will still run),Query.interrupt()returns the typed receipt, andsystem/initadvertises aninterrupt_receipt_v1capability for feature detectionnameandbodyfields to peer-message session events, exposing the sender display name and decoded message bodyv0.3.204Compare Source
terminal_reasonvaluestool_deferred_unavailable(deferred tool resume found the tool gone — previously anis_errorresult with no reason, read as a clean completion by lifecycle sweeps) andturn_setup_failed(the turn-input builder threw before the turn started). Both classify as dead turns, so commands consumed by them reportcommand_lifecyclestatecancelledcancelled— on remote transports that acknowledged them as processed, silently dropping messages nobody cancelled)terminal_reasonvaluesapi_error,malformed_tool_use_exhausted,budget_exhausted, andstructured_output_retry_exhausted. Turns that die on an exhausted-API-retry or malformed-tool-use give-up previously reportedcompleted; budget and structured-output exhaustion results previously omittedterminal_reason. Commands consumed by such turns now reportcommand_lifecyclestatecancelledinstead ofcompleted(dup-over-loss)v0.3.203Compare Source
background_tasks_changedsystem message with the full set of live background tasks on every membership change, so consumers can track background activity as a level instead of pairingtask_started/task_notificationedgessdk.d.tswith unresolved type references that broke consumer typechecking withskipLibCheckdisabledv0.3.202Compare Source
parent_agent_idfield to subagent session messages for building depth-2+ agent trees from disk-persisted metadataapply_flag_settingswith a non-object settings value crashing the session instead of returning a control errorv0.3.201Compare Source
v0.3.200Compare Source
'manual'as an accepted alias for the'default'permission mode in SDK inputsonSetPermissionModecallback not firing for SDK-hosted Remote Control sessionsset_modelcontrol request accepting unrecognized model strings; invalid models are now rejected before latchingv0.3.199Compare Source
requestIdtocanUseToolcallback options for correlating out-of-band permission responses, and support for returningnullto suppress the SDK's automatic control responseblockedfield toworkflow_agentprogress events indicating when an agent was blocked by the auto-mode safety classifiermode:"mask"and per-credentialinjectHoststosandbox.credentialssettings types for injecting masked credentials into sandboxed commandsv0.3.198Compare Source
canUseToolis configured alongsideallowedToolsorbypassPermissions, which shadow the callbackrequest_timeout_msoption tomcp_set_serverscontrol requestSDKUserMessage.isSyntheticnot being mapped toisMetaon ingestion, which could cause synthetic messages to be treated as real user messagesv0.3.197Compare Source
v0.3.196Compare Source
prompt_idfield to hook input payloads for correlating hook events with OpenTelemetry prompt-level eventstool_resultdeliveries in long-running sessionsv0.3.195Compare Source
Query.reinitialize()to re-send the initialize control request and redeliver pending permission/dialog prompts after a transport gapcommands_changedevent not being emitted for synced skills when the skill list resolves before the change-detector subscribesv0.3.193Compare Source
promptSuggestionsoption to Browser SDKquery()to opt the remote CLI into emitting follow-up suggestionsv0.3.191Compare Source
old_sourcefield toNotebookEdittool results forreplaceanddeleteoperations, enabling inline diffsseven_day_overage_includedtoSDKRateLimitInfo.rateLimitTypefor per-model weekly usage limitsmodel_scopedarray to usage response for per-model weekly limit windows with utilization and reset timessettingSourcesincludes user/project settingsv0.3.190Compare Source
v0.3.187Compare Source
sandbox.credentialsto SDK settings types for configuring credential file and environment variable denial in sandboxed commandsv0.3.186Compare Source
agent_idfield tocan_use_toolcontrol requests — background agents now forward permission prompts tocanUseToolinstead of auto-denying, and stdin stays open while background tasks are runningReadMcpResourceDirTooltool type to SDK schemas — MCP resource directory listing is now a dedicated tool instead of a fallback insideReadMcpResourceToolrewind_conversationcontrol request for rewinding a conversation to a previous point with durable resume anchor supportv0.3.185Compare Source
v0.3.183v0.3.182v0.3.181Compare Source
errorCode,canUserPurchaseCredits, andhasChargeableSavedPaymentMethodfields toSDKRateLimitInfofor detecting credits-required rate limitstool_use_meta.icon_urlto assistant messages, populated from MCP server directory metadatafile_attachmentsfrom inbound user messagesv0.3.179Compare Source
tool_use_metasidecar to assistant messages with display-friendly names for tool calls, so SDK consumers can render human-readable labels instead of raw wire names-pmode exiting before a completed background agent's notification was delivered, causing interim text to ship as the final resultv0.3.178Compare Source
options.pathToClaudeCodeExecutablesafetyCheck,asyncAgent), enabling SDK consumers to programmatically match denial causesUserPromptSubmithook block feedback not being emitted to the SDK event stream — consumers can now see why a prompt was blocked by a hook instead of a silent hangworker_shutting_downsystem message on graceful exit so remote clients can show why the session endedmcp__server,mcp__server__*) indisallowedToolsbeing silently ignored — they now correctly remove all tools from the named serverv0.3.177Compare Source
v0.3.176Compare Source
resultmessages being dropped when multiple turns complete while a background agent or workflow is runningv0.3.175Compare Source
v0.3.174Compare Source
system/model_fallbackmessage for all fallback triggers —overloaded,server_error, andlast_resortin addition tomodel_not_foundandpermission_denied— and the message'striggerfield gained theserver_errorandlast_resortvaluesv0.3.173Compare Source
v0.3.172Compare Source
pluginsoption now acceptsskipMcpDiscovery: trueper plugin, so a host that manages a plugin's MCP connections itself can load skills/hooks from the plugin path without the engine re-reading its.mcp.json/ add tests) being silently dropped instead of treated as a plain promptv0.3.170Compare Source
v0.3.169Compare Source
usage_EXPERIMENTAL_MAY_CHANGE_DO_NOT_RELY_ON_THIS_API_YET()method onQueryreturning structured session cost, plan rate-limit, and local usage-behaviors datasseoption (SSEOptions) toBrowserQueryOptionsas an alternative towebsocket, for browser SDK consumers who prefer Server-Sent EventsConfiguration
📅 Schedule: (UTC)
🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.
♻ Rebasing: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox.
🔕 Ignore: Close this PR and you won't be reminded about this update again.
This PR was generated by Mend Renovate. View the repository job log.