The whole hunk
from line 1, old and new numbered
/
lines
from line 1
1---
2title: Update Agent
3url: https://platform.claude.com/docs/en/api/beta/agents/update
4---
5
16# Update Agent
27
38**POST** `/v1/agents/{agent_id}`
from line 21
1621
1722 - `string`
1823
19 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 41 more`
24 - `"message-batches-2024-09-24" or "prompt-caching-2024-07-31" or "computer-use-2024-10-22" or 42 more`
2025
2126 - `"message-batches-2024-09-24"`
2227
from line 69
6469
6570 - `"user-profiles-2026-08-18"`
6671
72 - `"user-profiles-2026-09-04"`
73
6774 - `"advisor-tool-2026-03-01"`
6875
6976 - `"managed-agents-2026-04-01"`
from line 113
106113
107114 - `"mid-conversation-system-clear-at-2026-08-21"`
108115
116- `"anthropic-workspace-id": optional string`
117
109118## Body parameters
110119
111120- `description: optional string or null`
from line 127
118127
119128 MCP servers. Full replacement. Omit to preserve; send empty array or `null` to clear. Names must be unique. Maximum 20. Every server must be referenced by an `mcp_toolset` in the agent's resulting `tools`; unreferenced servers are rejected. See the [MCP connector guide](https://platform.claude.com/docs/en/managed-agents/mcp-connector).
120129
130 - `type: "url"`
131
121132 - `name: string`
122133
123134 Unique name for this server, referenced by mcp_toolset configurations. 1-255 characters.
from line 135
124135
125136 minLength: 1, maxLength: 255
126137
127 - `type: "url"`
128
129138 - `url: string`
130139
131140 Endpoint URL for the MCP server.
from line 293
284293
285294 A coordinator topology: the session's primary thread orchestrates work by spawning session threads, each running an agent drawn from the `agents` roster.
286295
296 - `type: "coordinator"`
297
287298 - `agents: array of BetaManagedAgentsMultiagentRosterEntryParams`
288299
289300 Agents the coordinator may spawn as session threads. 1–20 entries. Each entry is an agent ID string, a versioned `{"type":"agent","id","version"}` reference, or `{"type":"self"}` to allow recursive self-invocation. Entries must reference distinct agents (after resolving `self` and string forms); at most one `self`. Referenced agents must exist, must not be archived, and must not themselves have `multiagent` set (depth limit 1).
from line 305
294305
295306 Specification for an Agent. Provide a specific `version` or use the short-form `agent="agent_id"` for the most recent version
296307
308 - `type: "agent"`
309
297310 - `id: string`
298311
299312 The `agent` ID.
from line 313
300313
301314 minLength: 1, maxLength: 128
302315
303 - `type: "agent"`
304
305316 - `version: optional number`
306317
307318 The specific `agent` version to use. Omit to use the latest version. Must be at least 1 if specified.
from line 329
318329
319330 Platform advisor roster entry: a model the session's primary thread may consult mid-turn. At most one per roster; the entry occupies the roster name `anthropic.advisor`.
320331
332 - `type: "advisor"`
333
321334 - `model: string`
322335
323336 A Claude model id. The model must be permitted as an advisor for this agent's model — see the sessions/threads/advisor spec.
from line 337
324337
325338 minLength: 1, maxLength: 256
326339
327 - `type: "advisor"`
328
329 - `type: "coordinator"`
330
331340- `name: optional string`
332341
333342 Human-readable name. Must be non-empty. Omit to preserve. Cannot be cleared.
from line 351
342351
343352 An Anthropic-managed skill.
344353
354 - `type: "anthropic"`
355
345356 - `skill_id: string`
346357
347358 Identifier of the Anthropic skill (e.g., "xlsx").
from line 359
348359
349360 minLength: 1, maxLength: 64
350361
351 - `type: "anthropic"`
352
353362 - `version: optional string or null`
354363
355364 Version to pin. Defaults to latest if omitted.
from line 369
360369
361370 A user-created custom skill.
362371
372 - `type: "custom"`
373
363374 - `skill_id: string`
364375
365376 Tagged ID of the custom skill (e.g., "skill_01XJ5...").
from line 377
366377
367378 minLength: 1, maxLength: 64
368379
369 - `type: "custom"`
370
371380 - `version: optional string or null`
372381
373382 Version to pin. Defaults to latest if omitted.
from line 407
398407
399408 Configuration override for the bash tool.
400409
410 - `type: optional "bash"`
411
401412 - `name: "bash"`
402413
403414 Must be "bash".
from line 417
406417
407418 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
408419
409 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
420 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
410421
411422 Permission policy for tool execution.
412423
from line 433
422433
423434 - `type: "always_ask"`
424435
425 - `type: optional "bash"`
436 - `BetaManagedAgentsAutoPolicy object`
426437
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
440 - `type: "auto"`
441
427442 - `BetaManagedAgentsEditToolConfigParams object`
428443
429444 Configuration override for the edit tool.
430445
446 - `type: optional "edit"`
447
431448 - `name: "edit"`
432449
433450 Must be "edit".
from line 453
436453
437454 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
438455
439 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
456 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
440457
441458 Permission policy for tool execution.
442459
from line 465
448465
449466 Tool calls require user confirmation before execution.
450467
451 - `type: optional "edit"`
468 - `BetaManagedAgentsAutoPolicy object`
452469
470 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.
471
453472 - `BetaManagedAgentsReadToolConfigParams object`
454473
455474 Configuration override for the read tool.
456475
476 - `type: optional "read"`
477
457478 - `name: "read"`
458479
459480 Must be "read".
from line 483
462483
463484 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
464485
465 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
486 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
466487
467488 Permission policy for tool execution.
468489
from line 495
474495
475496 Tool calls require user confirmation before execution.
476497
477 - `type: optional "read"`
498 - `BetaManagedAgentsAutoPolicy object`
478499
500 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.
501
479502 - `BetaManagedAgentsWriteToolConfigParams object`
480503
481504 Configuration override for the write tool.
482505
506 - `type: optional "write"`
507
483508 - `name: "write"`
484509
485510 Must be "write".
from line 513
488513
489514 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
490515
491 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
516 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
492517
493518 Permission policy for tool execution.
494519
from line 525
500525
501526 Tool calls require user confirmation before execution.
502527
503 - `type: optional "write"`
528 - `BetaManagedAgentsAutoPolicy object`
504529
530 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.
531
505532 - `BetaManagedAgentsGlobToolConfigParams object`
506533
507534 Configuration override for the glob tool.
508535
536 - `type: optional "glob"`
537
509538 - `name: "glob"`
510539
511540 Must be "glob".
from line 543
514543
515544 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
516545
517 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
546 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
518547
519548 Permission policy for tool execution.
520549
from line 555
526555
527556 Tool calls require user confirmation before execution.
528557
529 - `type: optional "glob"`
558 - `BetaManagedAgentsAutoPolicy object`
530559
560 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.
561
531562 - `BetaManagedAgentsGrepToolConfigParams object`
532563
533564 Configuration override for the grep tool.
534565
566 - `type: optional "grep"`
567
535568 - `name: "grep"`
536569
537570 Must be "grep".
from line 573
540573
541574 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
542575
543 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
576 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
544577
545578 Permission policy for tool execution.
546579
from line 585
552585
553586 Tool calls require user confirmation before execution.
554587
555 - `type: optional "grep"`
588 - `BetaManagedAgentsAutoPolicy object`
556589
590 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.
591
557592 - `BetaManagedAgentsWebFetchToolConfigParams object`
558593
559594 Configuration override for the web_fetch tool.
560595
596 - `type: optional "web_fetch"`
597
561598 - `name: "web_fetch"`
562599
563600 Must be "web_fetch".
from line 617
580617
581618 format: int32
582619
583 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
620 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
584621
585622 Permission policy for tool execution.
586623
from line 629
592629
593630 Tool calls require user confirmation before execution.
594631
595 - `type: optional "web_fetch"`
632 - `BetaManagedAgentsAutoPolicy object`
596633
634 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.
635
597636 - `BetaManagedAgentsWebSearchToolConfigParams object`
598637
599638 Configuration override for the web_search tool.
600639
640 - `type: optional "web_search"`
641
601642 - `name: "web_search"`
602643
603644 Must be "web_search".
from line 655
614655
615656 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
616657
617 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
658 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
618659
619660 Permission policy for tool execution.
620661
from line 667
626667
627668 Tool calls require user confirmation before execution.
628669
629 - `type: optional "web_search"`
670 - `BetaManagedAgentsAutoPolicy object`
630671
672 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.
673
631674 - `user_location: optional BetaManagedAgentsUserLocation or null`
632675
633676 Approximate user location for search result localization.
from line 709
666709
667710 Whether tools are enabled and available to Claude by default. Defaults to true if not specified.
668711
669 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
712 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
670713
671714 Permission policy for tool execution.
672715
from line 721
678721
679722 Tool calls require user confirmation before execution.
680723
724 - `BetaManagedAgentsAutoPolicy object`
725
726 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.
727
681728 - `BetaManagedAgentsMCPToolsetParams object`
682729
683730 Configuration for tools from an MCP server defined in `mcp_servers`.
684731
732 - `type: "mcp_toolset"`
733
685734 - `mcp_server_name: string`
686735
687736 Name of the MCP server. Must match a server name from the mcp_servers array. 1-255 characters.
from line 737
688737
689738 minLength: 1, maxLength: 255
690739
691 - `type: "mcp_toolset"`
692
693740 - `configs: optional array of BetaManagedAgentsMCPToolConfigParams`
694741
695742 Per-tool configuration overrides.
from line 751
704751
705752 Whether this tool is enabled. Overrides the `default_config` setting.
706753
707 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
754 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
708755
709756 Permission policy for tool execution.
710757
from line 763
716763
717764 Tool calls require user confirmation before execution.
718765
766 - `BetaManagedAgentsAutoPolicy object`
767
768 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.
769
719770 - `default_config: optional BetaManagedAgentsMCPToolsetDefaultConfigParams or null`
720771
721772 Default configuration for all tools from an MCP server.
from line 775
724775
725776 Whether tools are enabled by default. Defaults to true if not specified.
726777
727 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or null`
778 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
728779
729780 Permission policy for tool execution.
730781
from line 787
736787
737788 Tool calls require user confirmation before execution.
738789
790 - `BetaManagedAgentsAutoPolicy object`
791
792 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.
793
739794 - `BetaManagedAgentsCustomToolParams object`
740795
741796 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.
742797
798 - `type: "custom"`
799
743800 - `description: string`
744801
745802 Description of what the tool does, shown to the agent to help it decide when to use the tool.
from line 819
762819
763820 minLength: 1, maxLength: 128
764821
765 - `type: "custom"`
766
767822- `version: optional number`
768823
769824 The agent's current version, used to prevent concurrent overwrites. Obtain this value from a create or retrieve response. Must be at least 1 if specified. When supplied, the request fails if it does not match the server's current version; omit to apply the update unconditionally.
from line 831
776831
777832 A Managed Agents `agent`.
778833
834 - `type: "agent"`
835
779836 - `id: string`
780837
781838 - `archived_at: string or null`
from line 851
794851
795852 - `mcp_servers: array of BetaManagedAgentsMCPServerURLDefinition`
796853
797 - `name: string`
798
799854 - `type: "url"`
800855
856 - `name: string`
857
801858 - `url: string`
802859
803860 - `metadata: map[string]`
from line 983
926983
927984 Resolved coordinator topology with a concrete agent roster.
928985
986 - `type: "coordinator"`
987
929988 - `agents: array of BetaManagedAgentsAgentReference or BetaManagedAgentsAdvisor`
930989
931990 Agents the coordinator may spawn as session threads, each resolved to a specific version.
from line 993
934993
935994 A resolved agent reference with a concrete version.
936995
937 - `id: string`
938
939996 - `type: "agent"`
940997
998 - `id: string`
999
9411000 - `version: number`
9421001
9431002 format: int32
from line 1005
9461005
9471006 Platform advisor roster entry: a model the session's primary thread may consult mid-turn.
9481007
1008 - `type: "advisor"`
1009
9491010 - `model: string`
9501011
9511012 The advisor model id.
9521013
953 - `type: "advisor"`
954
955 - `type: "coordinator"`
956
9571014 - `name: string`
9581015
9591016 - `skills: array of BetaManagedAgentsAnthropicSkill or BetaManagedAgentsCustomSkill`
from line 1019
9621019
9631020 A resolved Anthropic-managed skill.
9641021
965 - `skill_id: string`
966
9671022 - `type: "anthropic"`
9681023
1024 - `skill_id: string`
1025
9691026 - `version: string`
9701027
9711028 - `BetaManagedAgentsCustomSkill object`
from line 1029
9721029
9731030 A resolved user-created custom skill.
9741031
975 - `skill_id: string`
976
9771032 - `type: "custom"`
9781033
1034 - `skill_id: string`
1035
9791036 - `version: string`
9801037
9811038 - `system: string or null`
from line 1041
9841041
9851042 - `BetaManagedAgentsAgentToolset20260401 object`
9861043
1044 - `type: "agent_toolset_20260401"`
1045
9871046 - `configs: array of BetaManagedAgentsAgentToolConfig`
9881047
9891048 - `BetaManagedAgentsBashToolConfig object`
from line 1049
9901049
9911050 Configuration for the bash tool.
9921051
1052 - `type: "bash"`
1053
9931054 - `enabled: boolean`
9941055
9951056 - `name: "bash"`
9961057
997 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1058 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
9981059
9991060 Permission policy for tool execution.
10001061
from line 1071
10101071
10111072 - `type: "always_ask"`
10121073
1013 - `type: "bash"`
1074 - `BetaManagedAgentsAutoPolicy object`
10141075
1076 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.
1077
1078 - `type: "auto"`
1079
10151080 - `BetaManagedAgentsEditToolConfig object`
10161081
10171082 Configuration for the edit tool.
10181083
1084 - `type: "edit"`
1085
10191086 - `enabled: boolean`
10201087
10211088 - `name: "edit"`
10221089
1023 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1090 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
10241091
10251092 Permission policy for tool execution.
10261093
from line 1099
10321099
10331100 Tool calls require user confirmation before execution.
10341101
1035 - `type: "edit"`
1102 - `BetaManagedAgentsAutoPolicy object`
10361103
1104 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.
1105
10371106 - `BetaManagedAgentsReadToolConfig object`
10381107
10391108 Configuration for the read tool.
10401109
1110 - `type: "read"`
1111
10411112 - `enabled: boolean`
10421113
10431114 - `name: "read"`
10441115
1045 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1116 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
10461117
10471118 Permission policy for tool execution.
10481119
from line 1125
10541125
10551126 Tool calls require user confirmation before execution.
10561127
1057 - `type: "read"`
1128 - `BetaManagedAgentsAutoPolicy object`
10581129
1130 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.
1131
10591132 - `BetaManagedAgentsWriteToolConfig object`
10601133
10611134 Configuration for the write tool.
10621135
1136 - `type: "write"`
1137
10631138 - `enabled: boolean`
10641139
10651140 - `name: "write"`
10661141
1067 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1142 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
10681143
10691144 Permission policy for tool execution.
10701145
from line 1151
10761151
10771152 Tool calls require user confirmation before execution.
10781153
1079 - `type: "write"`
1154 - `BetaManagedAgentsAutoPolicy object`
10801155
1156 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.
1157
10811158 - `BetaManagedAgentsGlobToolConfig object`
10821159
10831160 Configuration for the glob tool.
10841161
1162 - `type: "glob"`
1163
10851164 - `enabled: boolean`
10861165
10871166 - `name: "glob"`
10881167
1089 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1168 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
10901169
10911170 Permission policy for tool execution.
10921171
from line 1177
10981177
10991178 Tool calls require user confirmation before execution.
11001179
1101 - `type: "glob"`
1180 - `BetaManagedAgentsAutoPolicy object`
11021181
1182 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.
1183
11031184 - `BetaManagedAgentsGrepToolConfig object`
11041185
11051186 Configuration for the grep tool.
11061187
1188 - `type: "grep"`
1189
11071190 - `enabled: boolean`
11081191
11091192 - `name: "grep"`
11101193
1111 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1194 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
11121195
11131196 Permission policy for tool execution.
11141197
from line 1203
11201203
11211204 Tool calls require user confirmation before execution.
11221205
1123 - `type: "grep"`
1206 - `BetaManagedAgentsAutoPolicy object`
11241207
1208 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.
1209
11251210 - `BetaManagedAgentsWebFetchToolConfig object`
11261211
11271212 Configuration for the web_fetch tool.
11281213
1214 - `type: "web_fetch"`
1215
11291216 - `enabled: boolean`
11301217
11311218 - `name: "web_fetch"`
11321219
1133 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1220 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
11341221
11351222 Permission policy for tool execution.
11361223
from line 1229
11421229
11431230 Tool calls require user confirmation before execution.
11441231
1145 - `type: "web_fetch"`
1232 - `BetaManagedAgentsAutoPolicy object`
11461233
1234 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.
1235
11471236 - `allowed_domains: optional array of string`
11481237
11491238 - `blocked_domains: optional array of string`
from line 1245
11561245
11571246 Configuration for the web_search tool.
11581247
1248 - `type: "web_search"`
1249
11591250 - `enabled: boolean`
11601251
11611252 - `name: "web_search"`
11621253
1163 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1254 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
11641255
11651256 Permission policy for tool execution.
11661257
from line 1263
11721263
11731264 Tool calls require user confirmation before execution.
11741265
1175 - `type: "web_search"`
1266 - `BetaManagedAgentsAutoPolicy object`
11761267
1268 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.
1269
11771270 - `allowed_domains: optional array of string`
11781271
11791272 - `blocked_domains: optional array of string`
from line 1307
12141307
12151308 - `enabled: boolean`
12161309
1217 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1310 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
12181311
12191312 Permission policy for tool execution.
12201313
from line 1319
12261319
12271320 Tool calls require user confirmation before execution.
12281321
1229 - `type: "agent_toolset_20260401"`
1322 - `BetaManagedAgentsAutoPolicy object`
12301323
1324 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.
1325
12311326 - `BetaManagedAgentsMCPToolset object`
12321327
1328 - `type: "mcp_toolset"`
1329
12331330 - `configs: array of BetaManagedAgentsMCPToolConfig`
12341331
12351332 - `enabled: boolean`
from line 1333
12361333
12371334 - `name: string`
12381335
1239 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1336 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
12401337
12411338 Permission policy for tool execution.
12421339
from line 1345
12481345
12491346 Tool calls require user confirmation before execution.
12501347
1348 - `BetaManagedAgentsAutoPolicy object`
1349
1350 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.
1351
12511352 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
12521353
12531354 Resolved default configuration for all tools from an MCP server.
from line 1355
12541355
12551356 - `enabled: boolean`
12561357
1257 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy`
1358 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
12581359
12591360 Permission policy for tool execution.
12601361
from line 1367
12661367
12671368 Tool calls require user confirmation before execution.
12681369
1269 - `mcp_server_name: string`
1370 - `BetaManagedAgentsAutoPolicy object`
12701371
1271 - `type: "mcp_toolset"`
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.
12721373
1374 - `mcp_server_name: string`
1375
12731376 - `BetaManagedAgentsCustomTool object`
12741377
12751378 A custom tool as returned in API responses.
12761379
1380 - `type: "custom"`
1381
12771382 - `description: string`
12781383
12791384 - `input_schema: BetaManagedAgentsCustomToolInputSchema`
from line 1392
12871392 - `required: optional array of string or null`
12881393
12891394 - `name: string`
1290
1291 - `type: "custom"`
1292
1293 - `type: "agent"`
12941395
12951396 - `updated_at: string`
12961397