Follow Discord
Sweep 22 Sep 2026 · 17:19Z Build v2.1.280 501 read Stable v2.1.267 Latest v2.1.280 Next v2.1.280 Feeds RSS JSON llms.txt Unofficial
One change · api

Create Session changed

api/beta/sessions/create

Nearest release: v2.1.268, published under an hour after this site recorded the change. Shown because the two are within 24 hours of each other. Nothing here says the release caused the edit.

Recorded here
Lines+1,722added
Lines−1,621removed
From line 1 where the diff opens
First seen 14 Aug 2026 this site's first read of the page
Recorded edits9to this page, all time

The whole hunk

from line 1, old and new numbered
/
lines
from line 1
1---
2title: Create Session
3url: https://platform.claude.com/docs/en/api/beta/sessions/create
4---
5 
16# Create Session
27 
38**POST** `/v1/sessions`
from line 17
1217 
1318 - `string`
1419 
15 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 41 more`
20 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 42 more`
1621 
1722 - `"message-batches-2024-09-24"`
1823 
from line 65
6065 
6166 - `"user-profiles-2026-08-18"`
6267 
68 - `"user-profiles-2026-09-04"`
69 
6370 - `"advisor-tool-2026-03-01"`
6471 
6572 - `"managed-agents-2026-04-01"`
from line 109
102109 
103110 - `"mid-conversation-system-clear-at-2026-08-21"`
104111 
112- `"anthropic-workspace-id": optional string`
113 
105114## Body parameters
106115 
107116- `agent: string or BetaManagedAgentsAgentParams or BetaManagedAgentsAgentWithOverridesParams`
from line 123
114123 
115124 Specification for an Agent. Provide a specific `version` or use the short-form `agent="agent_id"` for the most recent version
116125 
126 - `type: "agent"`
127 
117128 - `id: string`
118129 
119130 The `agent` ID.
from line 131
120131 
121132 minLength: 1, maxLength: 128
122133 
123 - `type: "agent"`
124 
125134 - `version: optional number`
126135 
127136 The specific `agent` version to use. Omit to use the latest version. Must be at least 1 if specified.
from line 141
132141 
133142 Reference to an `agent` plus optional configuration overrides. Each provided field replaces the agent's value for the caller's use; the agent resource is unchanged.
134143 
144 - `type: "agent_with_overrides"`
145 
135146 - `id: string`
136147 
137148 The `agent` ID.
from line 149
138149 
139150 minLength: 1, maxLength: 128
140151 
141 - `type: "agent_with_overrides"`
142 
143152 - `mcp_servers: optional array of BetaManagedAgentsURLMCPServerParams`
144153 
145154 Replacement MCP server list. Full replacement: the provided array becomes the MCP servers. Send an empty array to clear; omit to preserve the agent's servers.
146155 
156 - `type: "url"`
157 
147158 - `name: string`
148159 
149160 Unique name for this server, referenced by mcp_toolset configurations. 1-255 characters.
from line 161
150161 
151162 minLength: 1, maxLength: 255
152163 
153 - `type: "url"`
154 
155164 - `url: string`
156165 
157166 Endpoint URL for the MCP server.
from line 319
310319 
311320 An Anthropic-managed skill.
312321 
322 - `type: "anthropic"`
323 
313324 - `skill_id: string`
314325 
315326 Identifier of the Anthropic skill (e.g., "xlsx").
from line 327
316327 
317328 minLength: 1, maxLength: 64
318329 
319 - `type: "anthropic"`
320 
321330 - `version: optional string or null`
322331 
323332 Version to pin. Defaults to latest if omitted.
from line 337
328337 
329338 A user-created custom skill.
330339 
340 - `type: "custom"`
341 
331342 - `skill_id: string`
332343 
333344 Tagged ID of the custom skill (e.g., "skill_01XJ5...").
from line 345
334345 
335346 minLength: 1, maxLength: 64
336347 
337 - `type: "custom"`
338 
339348 - `version: optional string or null`
340349 
341350 Version to pin. Defaults to latest if omitted.
from line 375
366375 
367376 Configuration override for the bash tool.
368377 
378 - `type: optional "bash"`
379 
369380 - `name: "bash"`
370381 
371382 Must be "bash".
from line 385
374385 
375386 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
376387 
377 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
388 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
378389 
379390 Permission policy for tool execution.
380391 
from line 401
390401 
391402 - `type: "always_ask"`
392403 
393 - `type: optional "bash"`
404 - `BetaManagedAgentsAutoPolicy object`
394405 
406 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
407 
408 - `type: "auto"`
409 
395410 - `BetaManagedAgentsEditToolConfigParams object`
396411 
397412 Configuration override for the edit tool.
398413 
414 - `type: optional "edit"`
415 
399416 - `name: "edit"`
400417 
401418 Must be "edit".
from line 421
404421 
405422 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
406423 
407 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
424 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
408425 
409426 Permission policy for tool execution.
410427 
from line 433
416433 
417434 Tool calls require user confirmation before execution.
418435 
419 - `type: optional "edit"`
436 - `BetaManagedAgentsAutoPolicy object`
420437 
438 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
439 
421440 - `BetaManagedAgentsReadToolConfigParams object`
422441 
423442 Configuration override for the read tool.
424443 
444 - `type: optional "read"`
445 
425446 - `name: "read"`
426447 
427448 Must be "read".
from line 451
430451 
431452 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
432453 
433 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
454 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
434455 
435456 Permission policy for tool execution.
436457 
from line 463
442463 
443464 Tool calls require user confirmation before execution.
444465 
445 - `type: optional "read"`
466 - `BetaManagedAgentsAutoPolicy object`
446467 
468 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
469 
447470 - `BetaManagedAgentsWriteToolConfigParams object`
448471 
449472 Configuration override for the write tool.
450473 
474 - `type: optional "write"`
475 
451476 - `name: "write"`
452477 
453478 Must be "write".
from line 481
456481 
457482 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
458483 
459 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
484 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
460485 
461486 Permission policy for tool execution.
462487 
from line 493
468493 
469494 Tool calls require user confirmation before execution.
470495 
471 - `type: optional "write"`
496 - `BetaManagedAgentsAutoPolicy object`
472497 
498 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
499 
473500 - `BetaManagedAgentsGlobToolConfigParams object`
474501 
475502 Configuration override for the glob tool.
476503 
504 - `type: optional "glob"`
505 
477506 - `name: "glob"`
478507 
479508 Must be "glob".
from line 511
482511 
483512 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
484513 
485 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
514 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
486515 
487516 Permission policy for tool execution.
488517 
from line 523
494523 
495524 Tool calls require user confirmation before execution.
496525 
497 - `type: optional "glob"`
526 - `BetaManagedAgentsAutoPolicy object`
498527 
528 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
529 
499530 - `BetaManagedAgentsGrepToolConfigParams object`
500531 
501532 Configuration override for the grep tool.
502533 
534 - `type: optional "grep"`
535 
503536 - `name: "grep"`
504537 
505538 Must be "grep".
from line 541
508541 
509542 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
510543 
511 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
544 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
512545 
513546 Permission policy for tool execution.
514547 
from line 553
520553 
521554 Tool calls require user confirmation before execution.
522555 
523 - `type: optional "grep"`
556 - `BetaManagedAgentsAutoPolicy object`
524557 
558 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
559 
525560 - `BetaManagedAgentsWebFetchToolConfigParams object`
526561 
527562 Configuration override for the web_fetch tool.
528563 
564 - `type: optional "web_fetch"`
565 
529566 - `name: "web_fetch"`
530567 
531568 Must be "web_fetch".
from line 585
548585 
549586 format: int32
550587 
551 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
588 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
552589 
553590 Permission policy for tool execution.
554591 
from line 597
560597 
561598 Tool calls require user confirmation before execution.
562599 
563 - `type: optional "web_fetch"`
600 - `BetaManagedAgentsAutoPolicy object`
564601 
602 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
603 
565604 - `BetaManagedAgentsWebSearchToolConfigParams object`
566605 
567606 Configuration override for the web_search tool.
568607 
608 - `type: optional "web_search"`
609 
569610 - `name: "web_search"`
570611 
571612 Must be "web_search".
from line 623
582623 
583624 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
584625 
585 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
626 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
586627 
587628 Permission policy for tool execution.
588629 
from line 635
594635 
595636 Tool calls require user confirmation before execution.
596637 
597 - `type: optional "web_search"`
638 - `BetaManagedAgentsAutoPolicy object`
598639 
640 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
641 
599642 - `user_location: optional BetaManagedAgentsUserLocation or null`
600643 
601644 Approximate user location for search result localization.
from line 677
634677 
635678 Whether tools are enabled and available to Claude by default. Defaults to true if not specified.
636679 
637 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
680 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
638681 
639682 Permission policy for tool execution.
640683 
from line 689
646689 
647690 Tool calls require user confirmation before execution.
648691 
692 - `BetaManagedAgentsAutoPolicy object`
693 
694 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
695 
649696 - `BetaManagedAgentsMCPToolsetParams object`
650697 
651698 Configuration for tools from an MCP server defined in `mcp_servers`.
652699 
700 - `type: "mcp_toolset"`
701 
653702 - `mcp_server_name: string`
654703 
655704 Name of the MCP server. Must match a server name from the mcp_servers array. 1-255 characters.
from line 705
656705 
657706 minLength: 1, maxLength: 255
658707 
659 - `type: "mcp_toolset"`
660 
661708 - `configs: optional array of BetaManagedAgentsMCPToolConfigParams`
662709 
663710 Per-tool configuration overrides.
from line 719
672719 
673720 Whether this tool is enabled. Overrides the `default_config` setting.
674721 
675 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
722 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
676723 
677724 Permission policy for tool execution.
678725 
from line 731
684731 
685732 Tool calls require user confirmation before execution.
686733 
734 - `BetaManagedAgentsAutoPolicy object`
735 
736 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
737 
687738 - `default_config: optional BetaManagedAgentsMCPToolsetDefaultConfigParams or null`
688739 
689740 Default configuration for all tools from an MCP server.
from line 743
692743 
693744 Whether tools are enabled by default. Defaults to true if not specified.
694745 
695 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
746 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
696747 
697748 Permission policy for tool execution.
698749 
from line 755
704755 
705756 Tool calls require user confirmation before execution.
706757 
758 - `BetaManagedAgentsAutoPolicy object`
759 
760 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
761 
707762 - `BetaManagedAgentsCustomToolParams object`
708763 
709764 A custom tool that is executed by the API client rather than the agent. When the agent calls this tool, an `agent.custom_tool_use` event is emitted and the session goes idle, waiting for the client to provide the result via a `user.custom_tool_result` event.
710765 
766 - `type: "custom"`
767 
711768 - `description: string`
712769 
713770 Description of what the tool does, shown to the agent to help it decide when to use the tool.
from line 787
730787 
731788 minLength: 1, maxLength: 128
732789 
733 - `type: "custom"`
734 
735790 - `version: optional number`
736791 
737792 The specific `agent` version to use. Omit to use the latest version.
from line 803
748803 
749804 A hard spend ceiling. The session stops issuing new model requests once the tracked list cost reaches `max_list_cost`.
750805 
806 - `type: "limit"`
807 
751808 - `max_list_cost: BetaMonetaryAmount`
752809 
753810 A monetary amount in a specific currency.
from line 817
760817 
761818 Uppercase ISO-4217 currency code. `USD` is the only currency currently supported; the accepted set is closed and grows only when a new currency is priced.
762819 
763 - `type: "limit"`
764 
765820- `initial_events: optional array of BetaManagedAgentsUserMessageEventParams or BetaManagedAgentsUserDefineOutcomeEventParams`
766821 
767822 Initial events to send to the `session` at creation, processed in order. Supports `user.message` and `user.define_outcome` events. Maximum 50 events.
from line 825
770825 
771826 Parameters for sending a user message to the session.
772827 
828 - `type: "user.message"`
829 
773830 - `content: array of BetaManagedAgentsTextBlock or BetaManagedAgentsImageBlock or BetaManagedAgentsDocumentBlock or BetaManagedAgentsRedactedBlock`
774831 
775832 Array of content blocks for the user message.
from line 835
778835 
779836 Regular text content.
780837 
838 - `type: "text"`
839 
781840 - `text: string`
782841 
783842 The text content.
from line 843
784843 
785844 minLength: 1
786845 
787 - `type: "text"`
788 
789846 - `BetaManagedAgentsImageBlock object`
790847 
791848 Image content specified directly as base64 data or as a reference via a URL.
792849 
850 - `type: "image"`
851 
793852 - `source: BetaManagedAgentsBase64ImageSource or BetaManagedAgentsURLImageSource or BetaManagedAgentsFileImageSource`
794853 
795854 Union type for image source variants.
from line 857
798857 
799858 Base64-encoded image data.
800859 
860 - `type: "base64"`
861 
801862 - `data: string`
802863 
803864 Base64-encoded image data.
from line 871
810871 
811872 minLength: 1
812873 
813 - `type: "base64"`
814 
815874 - `BetaManagedAgentsURLImageSource object`
816875 
817876 Image referenced by URL.
from line 887
828887 
829888 Image referenced by file ID.
830889 
890 - `type: "file"`
891 
831892 - `file_id: string`
832893 
833894 ID of a previously uploaded file.
from line 895
834895 
835896 minLength: 1
836897 
837 - `type: "file"`
838 
839 - `type: "image"`
840 
841898 - `BetaManagedAgentsDocumentBlock object`
842899 
843900 Document content, either specified directly as base64 data, as text, or as a reference via a URL.
844901 
902 - `type: "document"`
903 
845904 - `source: BetaManagedAgentsBase64DocumentSource or BetaManagedAgentsPlainTextDocumentSource or BetaManagedAgentsURLDocumentSource or BetaManagedAgentsFileDocumentSource`
846905 
847906 Union type for document source variants.
from line 909
850909 
851910 Base64-encoded document data.
852911 
912 - `type: "base64"`
913 
853914 - `data: string`
854915 
855916 Base64-encoded document data.
from line 923
862923 
863924 minLength: 1
864925 
865 - `type: "base64"`
866 
867926 - `BetaManagedAgentsPlainTextDocumentSource object`
868927 
869928 Plain text document content.
870929 
930 - `type: "text"`
931 
871932 - `data: string`
872933 
873934 The plain text content.
from line 939
878939 
879940 MIME type of the text content. Must be "text/plain".
880941 
881 - `type: "text"`
882 
883942 - `BetaManagedAgentsURLDocumentSource object`
884943 
885944 Document referenced by URL.
from line 955
896955 
897956 Document referenced by file ID.
898957 
958 - `type: "file"`
959 
899960 - `file_id: string`
900961 
901962 ID of a previously uploaded file.
from line 963
902963 
903964 minLength: 1
904965 
905 - `type: "file"`
906 
907 - `type: "document"`
908 
909966 - `context: optional string or null`
910967 
911968 Additional context about the document for the model.
from line 977
920977 
921978 - `type: "redacted"`
922979 
923 - `type: "user.message"`
924 
925980 - `BetaManagedAgentsUserDefineOutcomeEventParams object`
926981 
927982 Parameters for defining an outcome the agent should work toward. The agent begins work on receipt.
928983 
984 - `type: "user.define_outcome"`
985 
929986 - `description: string`
930987 
931988 What the agent should produce. This is the task specification.
from line 995
938995 
939996 Rubric referenced by a file uploaded via the Files API.
940997 
998 - `type: "file"`
999 
9411000 - `file_id: string`
9421001 
9431002 ID of the rubric file.
9441003 
945 - `type: "file"`
946 
9471004 - `BetaManagedAgentsTextRubricParams object`
9481005 
9491006 Rubric content provided inline as text.
9501007 
1008 - `type: "text"`
1009 
9511010 - `content: string`
9521011 
9531012 Rubric content. Plain text or markdown — the grader treats it as freeform text. Maximum 262144 characters.
from line 1013
9541013 
9551014 maxLength: 262144
9561015 
957 - `type: "text"`
958 
959 - `type: "user.define_outcome"`
960 
9611016 - `max_iterations: optional number or null`
9621017 
9631018 Eval→revision cycles before giving up. Default 3, max 20.
from line 1031
9761031 
9771032 Mount a GitHub repository into the session's container.
9781033 
979 - `authorization_token: string`
980 
981 GitHub authorization token used to clone the repository.
982 
983 minLength: 1, maxLength: 4096
984 
9851034 - `type: "github_repository"`
9861035 
9871036 - `url: string`
from line 1039
9901039 
9911040 minLength: 1, maxLength: 2048
9921041 
1042 - `authorization_token: optional string`
1043 
1044 GitHub authorization token used to clone the repository. Required for private repositories; optional for public ones.
1045 
1046 minLength: 1, maxLength: 4096
1047 
9931048 - `checkout: optional BetaManagedAgentsBranchCheckout or BetaManagedAgentsCommitCheckout or null`
9941049 
9951050 Branch or commit to check out. Defaults to the repository's default branch.
from line 1051
9961051 
9971052 - `BetaManagedAgentsBranchCheckout object`
9981053 
1054 - `type: "branch"`
1055 
9991056 - `name: string`
10001057 
10011058 Branch name to check out.
from line 1059
10021059 
10031060 minLength: 1, maxLength: 255
10041061 
1005 - `type: "branch"`
1006 
10071062 - `BetaManagedAgentsCommitCheckout object`
10081063 
1064 - `type: "commit"`
1065 
10091066 - `sha: string`
10101067 
10111068 Full commit SHA to check out.
from line 1069
10121069 
10131070 minLength: 7, maxLength: 64
10141071 
1015 - `type: "commit"`
1016 
10171072 - `mount_path: optional string or null`
10181073 
10191074 Mount path in the container. Defaults to `/workspace/<repo-name>`.
from line 1079
10241079 
10251080 Mount a file uploaded via the Files API into the session.
10261081 
1082 - `type: "file"`
1083 
10271084 - `file_id: string`
10281085 
10291086 ID of a previously uploaded file.
from line 1087
10301087 
10311088 minLength: 1, maxLength: 128
10321089 
1033 - `type: "file"`
1034 
10351090 - `mount_path: optional string or null`
10361091 
10371092 Mount path in the container. Defaults to `/mnt/session/uploads/<file_id>`.
from line 1097
10421097 
10431098 Parameters for attaching a memory store to an agent session.
10441099 
1100 - `type: "memory_store"`
1101 
10451102 - `memory_store_id: string`
10461103 
10471104 The memory store ID (memstore_...). Must belong to the caller's organization and workspace.
10481105 
1049 - `type: "memory_store"`
1050 
10511106 - `access: optional "read_write" or "read_only" or null`
10521107 
10531108 Access mode for an attached memory store.
from line 1133
10781133 
10791134 A Managed Agents `session`.
10801135 
1136 - `type: "session"`
1137 
10811138 - `id: string`
10821139 
10831140 - `agent: BetaManagedAgentsSessionAgent`
from line 1141
10841141 
10851142 Resolved `agent` definition for a `session`. Snapshot of the `agent` at `session` creation time.
10861143 
1144 - `type: "agent"`
1145 
10871146 - `id: string`
10881147 
10891148 - `description: string or null`
from line 1149
10901149 
10911150 - `mcp_servers: array of BetaManagedAgentsMCPServerURLDefinition`
10921151 
1093 - `name: string`
1094 
10951152 - `type: "url"`
10961153 
1154 - `name: string`
1155 
10971156 - `url: string`
10981157 
10991158 - `model: BetaManagedAgentsModelConfig`
from line 1279
12201279 
12211280 Resolved coordinator topology with full agent definitions for each roster member.
12221281 
1282 - `type: "coordinator"`
1283 
12231284 - `agents: array of BetaManagedAgentsSessionThreadAgent or BetaManagedAgentsAdvisor`
12241285 
12251286 Full `agent` definitions the coordinator may spawn as session threads.
from line 1289
12281289 
12291290 Resolved `agent` definition for a single `session_thread`. Snapshot of the agent at thread creation time. The multiagent roster is not repeated here; read it from `Session.agent`.
12301291 
1292 - `type: "agent"`
1293 
12311294 - `id: string`
12321295 
12331296 - `description: string or null`
from line 1297
12341297 
12351298 - `mcp_servers: array of BetaManagedAgentsMCPServerURLDefinition`
12361299 
1237 - `name: string`
1238 
12391300 - `type: "url"`
12401301 
1302 - `name: string`
1303 
12411304 - `url: string`
12421305 
12431306 - `model: BetaManagedAgentsModelConfig`
from line 1315
12521315 
12531316 A resolved Anthropic-managed skill.
12541317 
1255 - `skill_id: string`
1256 
12571318 - `type: "anthropic"`
12581319 
1320 - `skill_id: string`
1321 
12591322 - `version: string`
12601323 
12611324 - `BetaManagedAgentsCustomSkill object`
from line 1325
12621325 
12631326 A resolved user-created custom skill.
12641327 
1265 - `skill_id: string`
1266 
12671328 - `type: "custom"`
12681329 
1330 - `skill_id: string`
1331 
12691332 - `version: string`
12701333 
12711334 - `system: string or null`
from line 1337
12741337 
12751338 - `BetaManagedAgentsAgentToolset20260401 object`
12761339 
1340 - `type: "agent_toolset_20260401"`
1341 
12771342 - `configs: array of BetaManagedAgentsAgentToolConfig`
12781343 
12791344 - `BetaManagedAgentsBashToolConfig object`
from line 1345
12801345 
12811346 Configuration for the bash tool.
12821347 
1348 - `type: "bash"`
1349 
12831350 - `enabled: boolean`
12841351 
12851352 - `name: "bash"`
12861353 
1287 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1354 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
12881355 
12891356 Permission policy for tool execution.
12901357 
from line 1367
13001367 
13011368 - `type: "always_ask"`
13021369 
1303 - `type: "bash"`
1370 - `BetaManagedAgentsAutoPolicy object`
13041371 
1372 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
1373 
1374 - `type: "auto"`
1375 
13051376 - `BetaManagedAgentsEditToolConfig object`
13061377 
13071378 Configuration for the edit tool.
13081379 
1380 - `type: "edit"`
1381 
13091382 - `enabled: boolean`
13101383 
13111384 - `name: "edit"`
13121385 
1313 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1386 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
13141387 
13151388 Permission policy for tool execution.
13161389 
from line 1395
13221395 
13231396 Tool calls require user confirmation before execution.
13241397 
1325 - `type: "edit"`
1398 - `BetaManagedAgentsAutoPolicy object`
13261399 
1400 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
1401 
13271402 - `BetaManagedAgentsReadToolConfig object`
13281403 
13291404 Configuration for the read tool.
13301405 
1406 - `type: "read"`
1407 
13311408 - `enabled: boolean`
13321409 
13331410 - `name: "read"`
13341411 
1335 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1412 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
13361413 
13371414 Permission policy for tool execution.
13381415 
from line 1421
13441421 
13451422 Tool calls require user confirmation before execution.
13461423 
1347 - `type: "read"`
1424 - `BetaManagedAgentsAutoPolicy object`
13481425 
1426 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
1427 
13491428 - `BetaManagedAgentsWriteToolConfig object`
13501429 
13511430 Configuration for the write tool.
13521431 
1432 - `type: "write"`
1433 
13531434 - `enabled: boolean`
13541435 
13551436 - `name: "write"`
13561437 
1357 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1438 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
13581439 
13591440 Permission policy for tool execution.
13601441 
from line 1447
13661447 
13671448 Tool calls require user confirmation before execution.
13681449 
1369 - `type: "write"`
1450 - `BetaManagedAgentsAutoPolicy object`
13701451 
1452 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
1453 
13711454 - `BetaManagedAgentsGlobToolConfig object`
13721455 
13731456 Configuration for the glob tool.
13741457 
1458 - `type: "glob"`
1459 
13751460 - `enabled: boolean`
13761461 
13771462 - `name: "glob"`
13781463 
1379 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1464 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
13801465 
13811466 Permission policy for tool execution.
13821467 
from line 1473
13881473 
13891474 Tool calls require user confirmation before execution.
13901475 
1391 - `type: "glob"`
1476 - `BetaManagedAgentsAutoPolicy object`
13921477 
1478 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
1479 
13931480 - `BetaManagedAgentsGrepToolConfig object`
13941481 
13951482 Configuration for the grep tool.
13961483 
1484 - `type: "grep"`
1485 
13971486 - `enabled: boolean`
13981487 
13991488 - `name: "grep"`
14001489 
1401 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1490 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
14021491 
14031492 Permission policy for tool execution.
14041493 
from line 1499
14101499 
14111500 Tool calls require user confirmation before execution.
14121501 
1413 - `type: "grep"`
1502 - `BetaManagedAgentsAutoPolicy object`
14141503 
1504 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
1505 
14151506 - `BetaManagedAgentsWebFetchToolConfig object`
14161507 
14171508 Configuration for the web_fetch tool.
14181509 
1510 - `type: "web_fetch"`
1511 
14191512 - `enabled: boolean`
14201513 
14211514 - `name: "web_fetch"`
14221515 
1423 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1516 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
14241517 
14251518 Permission policy for tool execution.
14261519 
from line 1525
14321525 
14331526 Tool calls require user confirmation before execution.
14341527 
1435 - `type: "web_fetch"`
1528 - `BetaManagedAgentsAutoPolicy object`
14361529 
1530 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
1531 
14371532 - `allowed_domains: optional array of string`
14381533 
14391534 - `blocked_domains: optional array of string`
from line 1541
14461541 
14471542 Configuration for the web_search tool.
14481543 
1544 - `type: "web_search"`
1545 
14491546 - `enabled: boolean`
14501547 
14511548 - `name: "web_search"`
14521549 
1453 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1550 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
14541551 
14551552 Permission policy for tool execution.
14561553 
from line 1559
14621559 
14631560 Tool calls require user confirmation before execution.
14641561 
1465 - `type: "web_search"`
1562 - `BetaManagedAgentsAutoPolicy object`
14661563 
1564 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
1565 
14671566 - `allowed_domains: optional array of string`
14681567 
14691568 - `blocked_domains: optional array of string`
from line 1603
15041603 
15051604 - `enabled: boolean`
15061605 
1507 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1606 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
15081607 
15091608 Permission policy for tool execution.
15101609 
from line 1615
15161615 
15171616 Tool calls require user confirmation before execution.
15181617 
1519 - `type: "agent_toolset_20260401"`
1618 - `BetaManagedAgentsAutoPolicy object`
15201619 
1620 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
1621 
15211622 - `BetaManagedAgentsMCPToolset object`
15221623 
1624 - `type: "mcp_toolset"`
1625 
15231626 - `configs: array of BetaManagedAgentsMCPToolConfig`
15241627 
15251628 - `enabled: boolean`
from line 1629
15261629 
15271630 - `name: string`
15281631 
1529 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1632 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
15301633 
15311634 Permission policy for tool execution.
15321635 
from line 1641
15381641 
15391642 Tool calls require user confirmation before execution.
15401643 
1644 - `BetaManagedAgentsAutoPolicy object`
1645 
1646 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
1647 
15411648 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
15421649 
15431650 Resolved default configuration for all tools from an MCP server.
from line 1651
15441651 
15451652 - `enabled: boolean`
15461653 
1547 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1654 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
15481655 
15491656 Permission policy for tool execution.
15501657 
from line 1663
15561663 
15571664 Tool calls require user confirmation before execution.
15581665 
1559 - `mcp_server_name: string`
1666 - `BetaManagedAgentsAutoPolicy object`
15601667 
1561 - `type: "mcp_toolset"`
1668 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
15621669 
1670 - `mcp_server_name: string`
1671 
15631672 - `BetaManagedAgentsCustomTool object`
15641673 
15651674 A custom tool as returned in API responses.
15661675 
1676 - `type: "custom"`
1677 
15671678 - `description: string`
15681679 
15691680 - `input_schema: BetaManagedAgentsCustomToolInputSchema`
from line 1689
15781689 
15791690 - `name: string`
15801691 
1581 - `type: "custom"`
1582 
1583 - `type: "agent"`
1584 
15851692 - `version: number`
15861693 
15871694 format: int32
from line 1697
15901697 
15911698 Platform advisor roster entry: a model the session's primary thread may consult mid-turn.
15921699 
1700 - `type: "advisor"`
1701 
15931702 - `model: string`
15941703 
15951704 The advisor model id.
15961705 
1597 - `type: "advisor"`
1598 
1599 - `type: "coordinator"`
1600 
16011706 - `name: string`
16021707 
16031708 - `skills: array of BetaManagedAgentsAnthropicSkill or BetaManagedAgentsCustomSkill`
from line 1727
16221727 
16231728 A custom tool as returned in API responses.
16241729 
1625 - `type: "agent"`
1626 
16271730 - `version: number`
16281731 
16291732 format: int32
from line 1741
16381741 
16391742 A hard spend ceiling. The session stops issuing new model requests once the tracked list cost reaches `max_list_cost`.
16401743 
1744 - `type: "limit"`
1745 
16411746 - `max_list_cost: BetaMonetaryAmount`
16421747 
16431748 A monetary amount in a specific currency.
from line 1755
16501755 
16511756 Uppercase ISO-4217 currency code. `USD` is the only currency currently supported; the accepted set is closed and grows only when a new currency is priced.
16521757 
1653 - `type: "limit"`
1654 
16551758 - `created_at: string`
16561759 
16571760 A timestamp in RFC 3339 format
from line 1769
16661769 
16671770 Per-outcome evaluation state. One entry per `define_outcome` event sent to the session.
16681771 
1772 - `type: "outcome_evaluation"`
1773 
16691774 - `completed_at: string or null`
16701775 
16711776 A timestamp in RFC 3339 format
from line 1799
16941799 
16951800 Current evaluation state. `pending` before the agent begins work; `running` while producing or revising; `evaluating` while the grader scores; `satisfied`/`max_iterations_reached`/`failed`/`interrupted` are terminal.
16961801 
1697 - `type: "outcome_evaluation"`
1698 
16991802 - `resources: array of BetaManagedAgentsSessionResource`
17001803 
17011804 - `BetaManagedAgentsGitHubRepositoryResource object`
17021805 
1806 - `type: "github_repository"`
1807 
17031808 - `id: string`
17041809 
17051810 - `created_at: string`
from line 1815
17101815 
17111816 - `mount_path: string`
17121817 
1713 - `type: "github_repository"`
1714 
17151818 - `updated_at: string`
17161819 
17171820 A timestamp in RFC 3339 format
from line 1827
17241827 
17251828 - `BetaManagedAgentsBranchCheckout object`
17261829 
1830 - `type: "branch"`
1831 
17271832 - `name: string`
17281833 
17291834 Branch name to check out.
from line 1835
17301835 
17311836 minLength: 1, maxLength: 255
17321837 
1733 - `type: "branch"`
1734 
17351838 - `BetaManagedAgentsCommitCheckout object`
17361839 
1840 - `type: "commit"`
1841 
17371842 - `sha: string`
17381843 
17391844 Full commit SHA to check out.
from line 1845
17401845 
17411846 minLength: 7, maxLength: 64
17421847 
1743 - `type: "commit"`
1744 
17451848 - `BetaManagedAgentsFileResource object`
17461849 
1850 - `type: "file"`
1851 
17471852 - `id: string`
17481853 
17491854 - `created_at: string`
from line 1861
17561861 
17571862 - `mount_path: string`
17581863 
1759 - `type: "file"`
1760 
17611864 - `updated_at: string`
17621865 
17631866 A timestamp in RFC 3339 format
from line 1871
17681871 
17691872 A memory store attached to an agent session.
17701873 
1874 - `type: "memory_store"`
1875 
17711876 - `memory_store_id: string`
17721877 
17731878 The memory store ID (memstore_...). Must belong to the caller's organization and workspace.
17741879 
1775 - `type: "memory_store"`
1776 
17771880 - `access: optional "read_write" or "read_only" or null`
17781881 
17791882 Access mode for an attached memory store.
from line 1932
18291932 - `"terminated"`
18301933 
18311934 - `title: string or null`
1832 
1833 - `type: "session"`
18341935 
18351936 - `updated_at: string`
18361937 
Feedback