Follow Discord
Sweep 09 Oct 2026 · 17:27Z Build v2.1.296 517 read Stable v2.1.287 Latest v2.1.296 Next v2.1.296 Feeds RSS JSON llms.txt llms-full.txt Unofficial
One change · api

Archive Session changedapi/beta/sessions/archive

Nearest release: v2.1.296, published 4 hours before 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+571added
Lines−437removed
From line 213 where the diff opens
First seen 14 Aug 2026 this site's first read of the page
Recorded edits18to this page, all time

The whole hunk

from line 213, old and new numbered
/
lines

The two sides of this change are more than 400 edits apart, too far apart to line up, so this is the differ's own diff of it and the words inside a line are not marked.

from line 213
213213 
214214 - `"claude-haiku-4-5"`
215215 
216 Fastest model with near-frontier intelligence
217 
218216 - `"claude-haiku-4-5-20251001"`
219217 
220 Fastest model with near-frontier intelligence
221 
222218 - `"claude-opus-4-5"`
223219 
224220 Powerful intelligence for long-running agents and coding
from line 283
287283 
288284 - `"fast"`
289285 
290 - `multiagent: BetaManagedAgentsSessionMultiagentCoordinator or null`
286 - `multiagent: BetaManagedAgentsSessionMultiagent or null`
291287 
292288 Resolved multiagent orchestration configuration. Null when the agent is single-threaded.
293289 
294 - `type: "coordinator"`
295 
296 - `agents: array of BetaManagedAgentsSessionThreadAgent or BetaManagedAgentsAdvisor`
297 
298 Full `agent` definitions the coordinator may spawn as session threads.
299 
300 - `BetaManagedAgentsSessionThreadAgent object`
301 
302 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`.
303 
304 - `type: "agent"`
305 
306 - `id: string`
307 
308 - `description: string or null`
309 
310 - `mcp_servers: array of BetaManagedAgentsMCPServerURLDefinition`
311 
312 - `type: "url"`
290 - `BetaManagedAgentsSessionMultiagentCoordinator object`
291 
292 Resolved coordinator topology with full agent definitions for each roster member.
293 
294 - `type: "coordinator"`
295 
296 - `agents: array of BetaManagedAgentsSessionThreadAgent or BetaManagedAgentsAdvisor`
297 
298 Full `agent` definitions the coordinator may spawn as session threads.
299 
300 - `BetaManagedAgentsSessionThreadAgent object`
301 
302 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`.
303 
304 - `type: "agent"`
305 
306 - `id: string`
307 
308 - `description: string or null`
309 
310 - `mcp_servers: array of BetaManagedAgentsMCPServerURLDefinition`
311 
312 - `type: "url"`
313 
314 - `name: string`
315 
316 - `url: string`
317 
318 - `model: BetaManagedAgentsModelConfig`
319 
320 Model identifier and configuration.
313321 
314322 - `name: string`
315323 
316 - `url: string`
317 
318 - `model: BetaManagedAgentsModelConfig`
319 
320 Model identifier and configuration.
321 
322 - `name: string`
323 
324 - `skills: array of BetaManagedAgentsAnthropicSkill or BetaManagedAgentsCustomSkill`
325 
326 - `BetaManagedAgentsAnthropicSkill object`
327 
328 A resolved Anthropic-managed skill.
329 
330 - `type: "anthropic"`
331 
332 - `skill_id: string`
333 
334 - `version: string`
335 
336 - `BetaManagedAgentsCustomSkill object`
337 
338 A resolved user-created custom skill.
339 
340 - `type: "custom"`
341 
342 - `skill_id: string`
343 
344 - `version: string`
345 
346 - `system: string or null`
347 
348 - `tools: array of BetaManagedAgentsAgentToolset20260401 or BetaManagedAgentsMCPToolset or BetaManagedAgentsCustomTool`
349 
350 - `BetaManagedAgentsAgentToolset20260401 object`
351 
352 - `type: "agent_toolset_20260401"`
353 
354 - `configs: array of BetaManagedAgentsAgentToolConfig`
355 
356 - `BetaManagedAgentsBashToolConfig object`
357 
358 Configuration for the bash tool.
359 
360 - `type: "bash"`
324 - `skills: array of BetaManagedAgentsAnthropicSkill or BetaManagedAgentsCustomSkill`
325 
326 - `BetaManagedAgentsAnthropicSkill object`
327 
328 A resolved Anthropic-managed skill.
329 
330 - `type: "anthropic"`
331 
332 - `skill_id: string`
333 
334 - `version: string`
335 
336 - `BetaManagedAgentsCustomSkill object`
337 
338 A resolved user-created custom skill.
339 
340 - `type: "custom"`
341 
342 - `skill_id: string`
343 
344 - `version: string`
345 
346 - `system: string or null`
347 
348 - `tools: array of BetaManagedAgentsAgentToolset20260401 or BetaManagedAgentsMCPToolset or BetaManagedAgentsCustomTool`
349 
350 - `BetaManagedAgentsAgentToolset20260401 object`
351 
352 - `type: "agent_toolset_20260401"`
353 
354 - `configs: array of BetaManagedAgentsAgentToolConfig`
355 
356 - `BetaManagedAgentsBashToolConfig object`
357 
358 Configuration for the bash tool.
359 
360 - `type: "bash"`
361 
362 - `enabled: boolean`
363 
364 - `name: "bash"`
365 
366 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
367 
368 Permission policy for tool execution.
369 
370 - `BetaManagedAgentsAlwaysAllowPolicy object`
371 
372 Tool calls are automatically approved without user confirmation.
373 
374 - `type: "always_allow"`
375 
376 - `BetaManagedAgentsAlwaysAskPolicy object`
377 
378 Tool calls require user confirmation before execution.
379 
380 - `type: "always_ask"`
381 
382 - `BetaManagedAgentsAutoPolicy object`
383 
384 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.
385 
386 - `type: "auto"`
387 
388 - `BetaManagedAgentsEditToolConfig object`
389 
390 Configuration for the edit tool.
391 
392 - `type: "edit"`
393 
394 - `enabled: boolean`
395 
396 - `name: "edit"`
397 
398 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
399 
400 Permission policy for tool execution.
401 
402 - `BetaManagedAgentsAlwaysAllowPolicy object`
403 
404 Tool calls are automatically approved without user confirmation.
405 
406 - `BetaManagedAgentsAlwaysAskPolicy object`
407 
408 Tool calls require user confirmation before execution.
409 
410 - `BetaManagedAgentsAutoPolicy object`
411 
412 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.
413 
414 - `BetaManagedAgentsReadToolConfig object`
415 
416 Configuration for the read tool.
417 
418 - `type: "read"`
419 
420 - `enabled: boolean`
421 
422 - `name: "read"`
423 
424 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
425 
426 Permission policy for tool execution.
427 
428 - `BetaManagedAgentsAlwaysAllowPolicy object`
429 
430 Tool calls are automatically approved without user confirmation.
431 
432 - `BetaManagedAgentsAlwaysAskPolicy object`
433 
434 Tool calls require user confirmation before execution.
435 
436 - `BetaManagedAgentsAutoPolicy object`
437 
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 - `BetaManagedAgentsWriteToolConfig object`
441 
442 Configuration for the write tool.
443 
444 - `type: "write"`
445 
446 - `enabled: boolean`
447 
448 - `name: "write"`
449 
450 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
451 
452 Permission policy for tool execution.
453 
454 - `BetaManagedAgentsAlwaysAllowPolicy object`
455 
456 Tool calls are automatically approved without user confirmation.
457 
458 - `BetaManagedAgentsAlwaysAskPolicy object`
459 
460 Tool calls require user confirmation before execution.
461 
462 - `BetaManagedAgentsAutoPolicy object`
463 
464 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.
465 
466 - `BetaManagedAgentsGlobToolConfig object`
467 
468 Configuration for the glob tool.
469 
470 - `type: "glob"`
471 
472 - `enabled: boolean`
473 
474 - `name: "glob"`
475 
476 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
477 
478 Permission policy for tool execution.
479 
480 - `BetaManagedAgentsAlwaysAllowPolicy object`
481 
482 Tool calls are automatically approved without user confirmation.
483 
484 - `BetaManagedAgentsAlwaysAskPolicy object`
485 
486 Tool calls require user confirmation before execution.
487 
488 - `BetaManagedAgentsAutoPolicy object`
489 
490 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.
491 
492 - `BetaManagedAgentsGrepToolConfig object`
493 
494 Configuration for the grep tool.
495 
496 - `type: "grep"`
497 
498 - `enabled: boolean`
499 
500 - `name: "grep"`
501 
502 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
503 
504 Permission policy for tool execution.
505 
506 - `BetaManagedAgentsAlwaysAllowPolicy object`
507 
508 Tool calls are automatically approved without user confirmation.
509 
510 - `BetaManagedAgentsAlwaysAskPolicy object`
511 
512 Tool calls require user confirmation before execution.
513 
514 - `BetaManagedAgentsAutoPolicy object`
515 
516 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.
517 
518 - `BetaManagedAgentsWebFetchToolConfig object`
519 
520 Configuration for the web_fetch tool.
521 
522 - `type: "web_fetch"`
523 
524 - `enabled: boolean`
525 
526 - `name: "web_fetch"`
527 
528 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
529 
530 Permission policy for tool execution.
531 
532 - `BetaManagedAgentsAlwaysAllowPolicy object`
533 
534 Tool calls are automatically approved without user confirmation.
535 
536 - `BetaManagedAgentsAlwaysAskPolicy object`
537 
538 Tool calls require user confirmation before execution.
539 
540 - `BetaManagedAgentsAutoPolicy object`
541 
542 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.
543 
544 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
545 
546 Which sources contribute URLs the tool may fetch, always in the object form. Null when not set, which allows every source.
547 
548 - `client_tool_results: BetaManagedAgentsWebFetchURLSourceToolFilter or null`
549 
550 Which custom tools' results contribute URLs that may be fetched. Null when not set, which allows every custom tool's results.
551 
552 - `BetaManagedAgentsWebFetchURLSourceAll object`
553 
554 Every URL from this source may be fetched. This is the default.
555 
556 - `type: "all"`
557 
558 - `BetaManagedAgentsWebFetchURLSourceNone object`
559 
560 This source contributes no URLs that may be fetched.
561 
562 - `type: "none"`
563 
564 - `BetaManagedAgentsWebFetchURLSourceOnly object`
565 
566 Only the named tools' results contribute URLs that may be fetched.
567 
568 - `type: "only"`
569 
570 - `tools: array of BetaManagedAgentsWebFetchURLSourceToolReference`
571 
572 The tools whose results contribute. Between 1 and 128 entries, each with a different name. An empty list is rejected; use "none" to allow no tool's results.
573 
574 - `type: "tool_reference"`
575 
576 Must be "tool_reference".
577 
578 - `name: string`
579 
580 Name of the tool. Compared exactly, so upper and lower case letters are different.
581 
582 minLength: 1, maxLength: 128
583 
584 - `BetaManagedAgentsWebFetchURLSourceExcept object`
585 
586 Every tool's results contribute URLs that may be fetched, except the named tools' results.
587 
588 - `type: "except"`
589 
590 - `tools: array of BetaManagedAgentsWebFetchURLSourceToolReference`
591 
592 The tools whose results do not contribute. Between 1 and 128 entries, each with a different name. An empty list is rejected; use "all" to leave out no tool's results.
593 
594 - `type: "tool_reference"`
595 
596 Must be "tool_reference".
597 
598 - `name: string`
599 
600 Name of the tool. Compared exactly, so upper and lower case letters are different.
601 
602 minLength: 1, maxLength: 128
603 
604 - `server_tool_results: BetaManagedAgentsWebFetchURLSourceToolFilter or null`
605 
606 Which of the web_search and web_fetch tools' results contribute URLs that may be fetched. Null when not set, which allows both.
607 
608 - `user_input: BetaManagedAgentsWebFetchURLSourceUserInput or null`
609 
610 Whether URLs in the text of user messages may be fetched. Null when not set, which allows them.
611 
612 - `BetaManagedAgentsWebFetchURLSourceAll object`
613 
614 Every URL from this source may be fetched. This is the default.
615 
616 - `BetaManagedAgentsWebFetchURLSourceNone object`
617 
618 This source contributes no URLs that may be fetched.
619 
620 - `allowed_domains: optional array of string`
621 
622 - `blocked_domains: optional array of string`
623 
624 - `max_content_tokens: optional number or null`
625 
626 format: int32
627 
628 - `BetaManagedAgentsWebSearchToolConfig object`
629 
630 Configuration for the web_search tool.
631 
632 - `type: "web_search"`
633 
634 - `enabled: boolean`
635 
636 - `name: "web_search"`
637 
638 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
639 
640 Permission policy for tool execution.
641 
642 - `BetaManagedAgentsAlwaysAllowPolicy object`
643 
644 Tool calls are automatically approved without user confirmation.
645 
646 - `BetaManagedAgentsAlwaysAskPolicy object`
647 
648 Tool calls require user confirmation before execution.
649 
650 - `BetaManagedAgentsAutoPolicy object`
651 
652 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.
653 
654 - `allowed_domains: optional array of string`
655 
656 - `blocked_domains: optional array of string`
657 
658 - `user_location: optional BetaManagedAgentsUserLocation or null`
659 
660 Approximate user location for search result localization.
661 
662 - `type: "approximate"`
663 
664 Location precision. Only "approximate" is supported.
665 
666 - `city: optional string or null`
667 
668 City name.
669 
670 minLength: 1, maxLength: 255
671 
672 - `country: optional string or null`
673 
674 Two-letter ISO 3166-1 country code, uppercase.
675 
676 - `region: optional string or null`
677 
678 Region or state name.
679 
680 minLength: 1, maxLength: 255
681 
682 - `timezone: optional string or null`
683 
684 IANA timezone identifier, e.g. "America/Los_Angeles".
685 
686 minLength: 1, maxLength: 255
687 
688 - `default_config: BetaManagedAgentsAgentToolsetDefaultConfig`
689 
690 Resolved default configuration for agent tools.
361691 
362692 - `enabled: boolean`
363693 
364 - `name: "bash"`
365 
366694 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
367695 
368696 Permission policy for tool execution.
from line 699
371699 
372700 Tool calls are automatically approved without user confirmation.
373701 
374 - `type: "always_allow"`
375 
376702 - `BetaManagedAgentsAlwaysAskPolicy object`
377703 
378704 Tool calls require user confirmation before execution.
379705 
380 - `type: "always_ask"`
381 
382706 - `BetaManagedAgentsAutoPolicy object`
383707 
384708 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.
385709 
386 - `type: "auto"`
387 
388 - `BetaManagedAgentsEditToolConfig object`
389 
390 Configuration for the edit tool.
391 
392 - `type: "edit"`
710 - `BetaManagedAgentsMCPToolset object`
711 
712 - `type: "mcp_toolset"`
713 
714 - `configs: array of BetaManagedAgentsMCPToolConfig`
393715 
394716 - `enabled: boolean`
395717 
396 - `name: "edit"`
718 - `name: string`
397719 
398720 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
399721 
from line 733
411733 
412734 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.
413735 
414 - `BetaManagedAgentsReadToolConfig object`
415 
416 Configuration for the read tool.
417 
418 - `type: "read"`
736 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
737 
738 Resolved default configuration for all tools from an MCP server.
419739 
420740 - `enabled: boolean`
421741 
422 - `name: "read"`
423 
424742 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
425743 
426744 Permission policy for tool execution.
from line 755
437755 
438756 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.
439757 
440 - `BetaManagedAgentsWriteToolConfig object`
441 
442 Configuration for the write tool.
443 
444 - `type: "write"`
445 
446 - `enabled: boolean`
447 
448 - `name: "write"`
449 
450 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
451 
452 Permission policy for tool execution.
453 
454 - `BetaManagedAgentsAlwaysAllowPolicy object`
455 
456 Tool calls are automatically approved without user confirmation.
457 
458 - `BetaManagedAgentsAlwaysAskPolicy object`
459 
460 Tool calls require user confirmation before execution.
461 
462 - `BetaManagedAgentsAutoPolicy object`
463 
464 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.
465 
466 - `BetaManagedAgentsGlobToolConfig object`
467 
468 Configuration for the glob tool.
469 
470 - `type: "glob"`
471 
472 - `enabled: boolean`
473 
474 - `name: "glob"`
475 
476 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
477 
478 Permission policy for tool execution.
479 
480 - `BetaManagedAgentsAlwaysAllowPolicy object`
481 
482 Tool calls are automatically approved without user confirmation.
483 
484 - `BetaManagedAgentsAlwaysAskPolicy object`
485 
486 Tool calls require user confirmation before execution.
487 
488 - `BetaManagedAgentsAutoPolicy object`
489 
490 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.
491 
492 - `BetaManagedAgentsGrepToolConfig object`
493 
494 Configuration for the grep tool.
495 
496 - `type: "grep"`
497 
498 - `enabled: boolean`
499 
500 - `name: "grep"`
501 
502 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
503 
504 Permission policy for tool execution.
505 
506 - `BetaManagedAgentsAlwaysAllowPolicy object`
507 
508 Tool calls are automatically approved without user confirmation.
509 
510 - `BetaManagedAgentsAlwaysAskPolicy object`
511 
512 Tool calls require user confirmation before execution.
513 
514 - `BetaManagedAgentsAutoPolicy object`
515 
516 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.
517 
518 - `BetaManagedAgentsWebFetchToolConfig object`
519 
520 Configuration for the web_fetch tool.
521 
522 - `type: "web_fetch"`
523 
524 - `enabled: boolean`
525 
526 - `name: "web_fetch"`
527 
528 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
529 
530 Permission policy for tool execution.
531 
532 - `BetaManagedAgentsAlwaysAllowPolicy object`
533 
534 Tool calls are automatically approved without user confirmation.
535 
536 - `BetaManagedAgentsAlwaysAskPolicy object`
537 
538 Tool calls require user confirmation before execution.
539 
540 - `BetaManagedAgentsAutoPolicy object`
541 
542 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.
543 
544 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
545 
546 Which sources contribute URLs the tool may fetch, always in the object form. Null when not set, which allows every source.
547 
548 - `client_tool_results: BetaManagedAgentsWebFetchURLSourceToolFilter or null`
549 
550 Which custom tools' results contribute URLs that may be fetched. Null when not set, which allows every custom tool's results.
551 
552 - `BetaManagedAgentsWebFetchURLSourceAll object`
553 
554 Every URL from this source may be fetched. This is the default.
555 
556 - `type: "all"`
557 
558 - `BetaManagedAgentsWebFetchURLSourceNone object`
559 
560 This source contributes no URLs that may be fetched.
561 
562 - `type: "none"`
563 
564 - `BetaManagedAgentsWebFetchURLSourceOnly object`
565 
566 Only the named tools' results contribute URLs that may be fetched.
567 
568 - `type: "only"`
569 
570 - `tools: array of BetaManagedAgentsWebFetchURLSourceToolReference`
571 
572 The tools whose results contribute. Between 1 and 128 entries, each with a different name. An empty list is rejected; use "none" to allow no tool's results.
573 
574 - `type: "tool_reference"`
575 
576 Must be "tool_reference".
577 
578 - `name: string`
579 
580 Name of the tool. Compared exactly, so upper and lower case letters are different.
581 
582 minLength: 1, maxLength: 128
583 
584 - `BetaManagedAgentsWebFetchURLSourceExcept object`
585 
586 Every tool's results contribute URLs that may be fetched, except the named tools' results.
587 
588 - `type: "except"`
589 
590 - `tools: array of BetaManagedAgentsWebFetchURLSourceToolReference`
591 
592 The tools whose results do not contribute. Between 1 and 128 entries, each with a different name. An empty list is rejected; use "all" to leave out no tool's results.
593 
594 - `type: "tool_reference"`
595 
596 Must be "tool_reference".
597 
598 - `name: string`
599 
600 Name of the tool. Compared exactly, so upper and lower case letters are different.
601 
602 minLength: 1, maxLength: 128
603 
604 - `server_tool_results: BetaManagedAgentsWebFetchURLSourceToolFilter or null`
605 
606 Which of the web_search and web_fetch tools' results contribute URLs that may be fetched. Null when not set, which allows both.
607 
608 - `user_input: BetaManagedAgentsWebFetchURLSourceUserInput or null`
609 
610 Whether URLs in the text of user messages may be fetched. Null when not set, which allows them.
611 
612 - `BetaManagedAgentsWebFetchURLSourceAll object`
613 
614 Every URL from this source may be fetched. This is the default.
615 
616 - `BetaManagedAgentsWebFetchURLSourceNone object`
617 
618 This source contributes no URLs that may be fetched.
619 
620 - `allowed_domains: optional array of string`
621 
622 - `blocked_domains: optional array of string`
623 
624 - `max_content_tokens: optional number or null`
625 
626 format: int32
627 
628 - `BetaManagedAgentsWebSearchToolConfig object`
629 
630 Configuration for the web_search tool.
631 
632 - `type: "web_search"`
633 
634 - `enabled: boolean`
635 
636 - `name: "web_search"`
637 
638 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
639 
640 Permission policy for tool execution.
641 
642 - `BetaManagedAgentsAlwaysAllowPolicy object`
643 
644 Tool calls are automatically approved without user confirmation.
645 
646 - `BetaManagedAgentsAlwaysAskPolicy object`
647 
648 Tool calls require user confirmation before execution.
649 
650 - `BetaManagedAgentsAutoPolicy object`
651 
652 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.
653 
654 - `allowed_domains: optional array of string`
655 
656 - `blocked_domains: optional array of string`
657 
658 - `user_location: optional BetaManagedAgentsUserLocation or null`
659 
660 Approximate user location for search result localization.
661 
662 - `type: "approximate"`
663 
664 Location precision. Only "approximate" is supported.
665 
666 - `city: optional string or null`
667 
668 City name.
669 
670 minLength: 1, maxLength: 255
671 
672 - `country: optional string or null`
673 
674 Two-letter ISO 3166-1 country code, uppercase.
675 
676 - `region: optional string or null`
677 
678 Region or state name.
679 
680 minLength: 1, maxLength: 255
681 
682 - `timezone: optional string or null`
683 
684 IANA timezone identifier, e.g. "America/Los_Angeles".
685 
686 minLength: 1, maxLength: 255
687 
688 - `default_config: BetaManagedAgentsAgentToolsetDefaultConfig`
689 
690 Resolved default configuration for agent tools.
691 
692 - `enabled: boolean`
693 
694 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
695 
696 Permission policy for tool execution.
697 
698 - `BetaManagedAgentsAlwaysAllowPolicy object`
699 
700 Tool calls are automatically approved without user confirmation.
701 
702 - `BetaManagedAgentsAlwaysAskPolicy object`
703 
704 Tool calls require user confirmation before execution.
705 
706 - `BetaManagedAgentsAutoPolicy object`
707 
708 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.
709 
710 - `BetaManagedAgentsMCPToolset object`
711 
712 - `type: "mcp_toolset"`
713 
714 - `configs: array of BetaManagedAgentsMCPToolConfig`
715 
716 - `enabled: boolean`
758 - `mcp_server_name: string`
759 
760 - `BetaManagedAgentsCustomTool object`
761 
762 A custom tool as returned in API responses.
763 
764 - `type: "custom"`
765 
766 - `description: string`
767 
768 - `input_schema: BetaManagedAgentsCustomToolInputSchema`
769 
770 JSON Schema for custom tool input parameters.
771 
772 - `type: "object"`
773 
774 - `properties: optional map[unknown] or null`
775 
776 - `required: optional array of string or null`
717777 
718778 - `name: string`
719779 
720 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
721 
722 Permission policy for tool execution.
723 
724 - `BetaManagedAgentsAlwaysAllowPolicy object`
725 
726 Tool calls are automatically approved without user confirmation.
727 
728 - `BetaManagedAgentsAlwaysAskPolicy object`
729 
730 Tool calls require user confirmation before execution.
731 
732 - `BetaManagedAgentsAutoPolicy object`
733 
734 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.
735 
736 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
737 
738 Resolved default configuration for all tools from an MCP server.
739 
740 - `enabled: boolean`
741 
742 - `permission_policy: BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy`
743 
744 Permission policy for tool execution.
745 
746 - `BetaManagedAgentsAlwaysAllowPolicy object`
747 
748 Tool calls are automatically approved without user confirmation.
749 
750 - `BetaManagedAgentsAlwaysAskPolicy object`
751 
752 Tool calls require user confirmation before execution.
753 
754 - `BetaManagedAgentsAutoPolicy object`
755 
756 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.
757 
758 - `mcp_server_name: string`
759 
760 - `BetaManagedAgentsCustomTool object`
761 
762 A custom tool as returned in API responses.
763 
764 - `type: "custom"`
765 
766 - `description: string`
767 
768 - `input_schema: BetaManagedAgentsCustomToolInputSchema`
769 
770 JSON Schema for custom tool input parameters.
771 
772 - `type: "object"`
773 
774 - `properties: optional map[unknown] or null`
775 
776 - `required: optional array of string or null`
780 - `version: number`
781 
782 format: int32
783 
784 - `BetaManagedAgentsAdvisor object`
785 
786 Platform advisor roster entry: a model the session's primary thread may consult mid-turn.
787 
788 - `type: "advisor"`
789 
790 - `model: string`
791 
792 The advisor model id.
793 
794 - `BetaManagedAgentsSessionMultiagent20261001 object`
795 
796 Resolved multiagent configuration with three members, as copied to the `session` at creation.
797 
798 - `type: "multiagent_20261001"`
799 
800 - `advisor: BetaManagedAgentsMultiagentAdvisor`
801 
802 Whether the session's primary thread can consult an advisor model.
803 
804 - `BetaManagedAgentsMultiagentAdvisorEnabled object`
805 
806 The session's primary thread can consult `model` mid-turn.
807 
808 - `type: "enabled"`
809 
810 - `model: string`
811 
812 The advisor model id.
813 
814 - `BetaManagedAgentsMultiagentAdvisorDisabled object`
815 
816 The agent has no advisor.
817 
818 - `type: "disabled"`
819 
820 - `subagents: BetaManagedAgentsSessionMultiagentSubagents`
821 
822 Whether the agent can spawn session threads.
823 
824 - `BetaManagedAgentsSessionMultiagentSubagentsEnabled object`
825 
826 The agent can spawn session threads.
827 
828 - `type: "enabled"`
829 
830 - `inline_agents: BetaManagedAgentsMultiagentInlineAgents`
831 
832 Whether the agent can define inline agents, which are not saved, when it spawns session threads.
833 
834 - `BetaManagedAgentsMultiagentInlineAgentsEnabled object`
835 
836 The agent can define inline agents.
837 
838 - `type: "enabled"`
839 
840 - `BetaManagedAgentsMultiagentInlineAgentsDisabled object`
841 
842 The agent cannot define inline agents.
843 
844 - `type: "disabled"`
845 
846 - `predefined_agents: array of BetaManagedAgentsSessionThreadAgent`
847 
848 Full `agent` definitions of the predefined agents, which are saved agents that this agent can spawn as session threads.
849 
850 - `type: "agent"`
851 
852 - `id: string`
853 
854 - `description: string or null`
855 
856 - `mcp_servers: array of BetaManagedAgentsMCPServerURLDefinition`
857 
858 - `model: BetaManagedAgentsModelConfig`
859 
860 Model identifier and configuration.
777861 
778862 - `name: string`
779863 
780 - `version: number`
781 
782 format: int32
783 
784 - `BetaManagedAgentsAdvisor object`
785 
786 Platform advisor roster entry: a model the session's primary thread may consult mid-turn.
787 
788 - `type: "advisor"`
789 
790 - `model: string`
791 
792 The advisor model id.
864 - `skills: array of BetaManagedAgentsAnthropicSkill or BetaManagedAgentsCustomSkill`
865 
866 - `system: string or null`
867 
868 - `tools: array of BetaManagedAgentsAgentToolset20260401 or BetaManagedAgentsMCPToolset or BetaManagedAgentsCustomTool`
869 
870 - `version: number`
871 
872 format: int32
873 
874 - `BetaManagedAgentsMultiagentSubagentsDisabled object`
875 
876 The agent cannot spawn session threads.
877 
878 - `type: "disabled"`
879 
880 - `workflows: BetaManagedAgentsSessionMultiagentWorkflows`
881 
882 Whether the agent can start workflow runs.
883 
884 - `BetaManagedAgentsSessionMultiagentWorkflowsEnabled object`
885 
886 The agent can start workflow runs.
887 
888 - `type: "enabled"`
889 
890 - `inline_agents: BetaManagedAgentsMultiagentInlineAgents`
891 
892 Whether a run's plan can define inline agents, which are not saved.
893 
894 - `predefined_agents: array of BetaManagedAgentsSessionThreadAgent`
895 
896 Full `agent` definitions of the predefined agents, which are saved agents that a run's plan can use.
897 
898 - `type: "agent"`
899 
900 - `id: string`
901 
902 - `description: string or null`
903 
904 - `mcp_servers: array of BetaManagedAgentsMCPServerURLDefinition`
905 
906 - `model: BetaManagedAgentsModelConfig`
907 
908 Model identifier and configuration.
909 
910 - `name: string`
911 
912 - `skills: array of BetaManagedAgentsAnthropicSkill or BetaManagedAgentsCustomSkill`
913 
914 - `system: string or null`
915 
916 - `tools: array of BetaManagedAgentsAgentToolset20260401 or BetaManagedAgentsMCPToolset or BetaManagedAgentsCustomTool`
917 
918 - `version: number`
919 
920 format: int32
921 
922 - `BetaManagedAgentsMultiagentWorkflowsDisabled object`
923 
924 The agent cannot start workflow runs.
925 
926 - `type: "disabled"`
793927 
794928 - `name: string`
795929 
796930 
Feedback