One read of Claude Developer Platformapi-20261112T203713Z
139 pages moved out of 916 read.
Pages moved
139
significant first
Pages read
916
in this capture
Captured
20:37 UTC
Corpus hash
59cd06ab8c5c
corpus-hash
What this read moved
26-50 of 139, page 2 of 6This capture is too large to show at once. Changes 26-50 of 139 are below, significant first; the rest are on the following screens.
api/beta/sessions/retrieve Changed · +11 / -11 lines
from line 425
425425
426426 - `BetaManagedAgentsAutoPolicy object`
427427
428 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.
428 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
429429
430430 - `type: "auto"`
431431
from line 453
453453
454454 - `BetaManagedAgentsAutoPolicy object`
455455
456 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.
456 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
457457
458458 - `BetaManagedAgentsReadToolConfig object`
459459
from line 479
479479
480480 - `BetaManagedAgentsAutoPolicy object`
481481
482 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.
482 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
483483
484484 - `BetaManagedAgentsWriteToolConfig object`
485485
from line 505
505505
506506 - `BetaManagedAgentsAutoPolicy object`
507507
508 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.
508 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
509509
510510 - `BetaManagedAgentsGlobToolConfig object`
511511
from line 531
531531
532532 - `BetaManagedAgentsAutoPolicy object`
533533
534 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.
534 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
535535
536536 - `BetaManagedAgentsGrepToolConfig object`
537537
from line 557
557557
558558 - `BetaManagedAgentsAutoPolicy object`
559559
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.
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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
561561
562562 - `BetaManagedAgentsWebFetchToolConfig object`
563563
from line 583
583583
584584 - `BetaManagedAgentsAutoPolicy object`
585585
586 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.
586 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
587587
588588 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
589589
from line 693
693693
694694 - `BetaManagedAgentsAutoPolicy object`
695695
696 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.
696 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
697697
698698 - `allowed_domains: optional array of string`
699699
from line 751
751751
752752 - `BetaManagedAgentsAutoPolicy object`
753753
754 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.
754 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
755755
756756 - `BetaManagedAgentsMCPToolset object`
757757
from line 777
777777
778778 - `BetaManagedAgentsAutoPolicy object`
779779
780 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.
780 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
781781
782782 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
783783
from line 799
799799
800800 - `BetaManagedAgentsAutoPolicy object`
801801
802 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.
802 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
803803
804804 - `mcp_server_name: string`
805805
api/beta/sessions/threads Changed · +94 / -94 lines
This page is larger than the 256 KiB this site keeps, so one side of the diff below stops where the stored text does.
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 419
419419
420420 - `BetaManagedAgentsAutoPolicy object`
421421
422 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.
422 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
423423
424424 - `type: "auto"`
425425
from line 447
447447
448448 - `BetaManagedAgentsAutoPolicy object`
449449
450 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.
450 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
451451
452452 - `BetaManagedAgentsReadToolConfig object`
453453
from line 473
473473
474474 - `BetaManagedAgentsAutoPolicy object`
475475
476 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.
476 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
477477
478478 - `BetaManagedAgentsWriteToolConfig object`
479479
from line 499
499499
500500 - `BetaManagedAgentsAutoPolicy object`
501501
502 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.
502 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
503503
504504 - `BetaManagedAgentsGlobToolConfig object`
505505
from line 525
525525
526526 - `BetaManagedAgentsAutoPolicy object`
527527
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.
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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
529529
530530 - `BetaManagedAgentsGrepToolConfig object`
531531
from line 551
551551
552552 - `BetaManagedAgentsAutoPolicy object`
553553
554 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.
554 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
555555
556556 - `BetaManagedAgentsWebFetchToolConfig object`
557557
from line 577
577577
578578 - `BetaManagedAgentsAutoPolicy object`
579579
580 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.
580 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
581581
582582 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
583583
from line 687
687687
688688 - `BetaManagedAgentsAutoPolicy object`
689689
690 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.
690 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
691691
692692 - `allowed_domains: optional array of string`
693693
from line 745
745745
746746 - `BetaManagedAgentsAutoPolicy object`
747747
748 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.
748 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
749749
750750 - `BetaManagedAgentsMCPToolset object`
751751
from line 771
771771
772772 - `BetaManagedAgentsAutoPolicy object`
773773
774 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.
774 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
775775
776776 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
777777
from line 793
793793
794794 - `BetaManagedAgentsAutoPolicy object`
795795
796 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.
796 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
797797
798798 - `mcp_server_name: string`
799799
from line 1520
15201520
15211521 - `BetaManagedAgentsAutoPolicy object`
15221522
1523 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.
1523 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
15241524
15251525 - `type: "auto"`
15261526
from line 1548
15481548
15491549 - `BetaManagedAgentsAutoPolicy object`
15501550
1551 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.
1551 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
15521552
15531553 - `BetaManagedAgentsReadToolConfig object`
15541554
from line 1574
15741574
15751575 - `BetaManagedAgentsAutoPolicy object`
15761576
1577 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.
1577 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
15781578
15791579 - `BetaManagedAgentsWriteToolConfig object`
15801580
from line 1600
16001600
16011601 - `BetaManagedAgentsAutoPolicy object`
16021602
1603 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.
1603 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
16041604
16051605 - `BetaManagedAgentsGlobToolConfig object`
16061606
from line 1626
16261626
16271627 - `BetaManagedAgentsAutoPolicy object`
16281628
1629 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.
1629 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
16301630
16311631 - `BetaManagedAgentsGrepToolConfig object`
16321632
from line 1652
16521652
16531653 - `BetaManagedAgentsAutoPolicy object`
16541654
1655 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.
1655 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
16561656
16571657 - `BetaManagedAgentsWebFetchToolConfig object`
16581658
from line 1678
16781678
16791679 - `BetaManagedAgentsAutoPolicy object`
16801680
1681 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.
1681 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
16821682
16831683 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
16841684
from line 1788
17881788
17891789 - `BetaManagedAgentsAutoPolicy object`
17901790
1791 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.
1791 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
17921792
17931793 - `allowed_domains: optional array of string`
17941794
from line 1846
18461846
18471847 - `BetaManagedAgentsAutoPolicy object`
18481848
1849 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.
1849 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
18501850
18511851 - `BetaManagedAgentsMCPToolset object`
18521852
from line 1872
18721872
18731873 - `BetaManagedAgentsAutoPolicy object`
18741874
1875 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.
1875 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
18761876
18771877 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
18781878
from line 1894
18941894
18951895 - `BetaManagedAgentsAutoPolicy object`
18961896
1897 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.
1897 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
18981898
18991899 - `mcp_server_name: string`
19001900
from line 2612
26122612
26132613 - `BetaManagedAgentsAutoPolicy object`
26142614
2615 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.
2615 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
26162616
26172617 - `type: "auto"`
26182618
from line 2640
26402640
26412641 - `BetaManagedAgentsAutoPolicy object`
26422642
2643 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.
2643 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
26442644
26452645 - `BetaManagedAgentsReadToolConfig object`
26462646
from line 2666
26662666
26672667 - `BetaManagedAgentsAutoPolicy object`
26682668
2669 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.
2669 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
26702670
26712671 - `BetaManagedAgentsWriteToolConfig object`
26722672
from line 2692
26922692
26932693 - `BetaManagedAgentsAutoPolicy object`
26942694
2695 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.
2695 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
26962696
26972697 - `BetaManagedAgentsGlobToolConfig object`
26982698
from line 2718
27182718
27192719 - `BetaManagedAgentsAutoPolicy object`
27202720
2721 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.
2721 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
27222722
27232723 - `BetaManagedAgentsGrepToolConfig object`
27242724
from line 2744
27442744
27452745 - `BetaManagedAgentsAutoPolicy object`
27462746
2747 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.
2747 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
27482748
27492749 - `BetaManagedAgentsWebFetchToolConfig object`
27502750
from line 2770
27702770
27712771 - `BetaManagedAgentsAutoPolicy object`
27722772
2773 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.
2773 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
27742774
27752775 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
27762776
from line 2880
28802880
28812881 - `BetaManagedAgentsAutoPolicy object`
28822882
2883 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.
2883 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
28842884
28852885 - `allowed_domains: optional array of string`
28862886
from line 2938
29382938
29392939 - `BetaManagedAgentsAutoPolicy object`
29402940
2941 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.
2941 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
29422942
29432943 - `BetaManagedAgentsMCPToolset object`
29442944
from line 2964
29642964
29652965 - `BetaManagedAgentsAutoPolicy object`
29662966
2967 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.
2967 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
29682968
29692969 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
29702970
from line 2986
29862986
29872987 - `BetaManagedAgentsAutoPolicy object`
29882988
2989 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.
2989 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
29902990
29912991 - `mcp_server_name: string`
29922992
from line 3549
35493549
35503550 - `BetaManagedAgentsAutoPolicy object`
35513551
3552 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.
3552 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
35533553
35543554 - `type: "auto"`
35553555
from line 3577
35773577
35783578 - `BetaManagedAgentsAutoPolicy object`
35793579
3580 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.
3580 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
35813581
35823582 - `BetaManagedAgentsReadToolConfig object`
35833583
from line 3603
36033603
36043604 - `BetaManagedAgentsAutoPolicy object`
36053605
3606 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.
3606 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
36073607
36083608 - `BetaManagedAgentsWriteToolConfig object`
36093609
from line 3629
36293629
36303630 - `BetaManagedAgentsAutoPolicy object`
36313631
3632 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.
3632 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
36333633
36343634 - `BetaManagedAgentsGlobToolConfig object`
36353635
from line 3655
36553655
36563656 - `BetaManagedAgentsAutoPolicy object`
36573657
3658 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.
3658 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
36593659
36603660 - `BetaManagedAgentsGrepToolConfig object`
36613661
from line 3681
36813681
36823682 - `BetaManagedAgentsAutoPolicy object`
36833683
3684 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.
3684 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
36853685
36863686 - `BetaManagedAgentsWebFetchToolConfig object`
36873687
from line 3707
37073707
37083708 - `BetaManagedAgentsAutoPolicy object`
37093709
3710 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.
3710 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
37113711
37123712 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
37133713
from line 3817
38173817
38183818 - `BetaManagedAgentsAutoPolicy object`
38193819
3820 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.
3820 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
38213821
38223822 - `allowed_domains: optional array of string`
38233823
from line 3875
38753875
38763876 - `BetaManagedAgentsAutoPolicy object`
38773877
3878 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.
3878 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
38793879
38803880 - `BetaManagedAgentsMCPToolset object`
38813881
from line 3901
39013901
39023902 - `BetaManagedAgentsAutoPolicy object`
39033903
3904 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.
3904 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
39053905
39063906 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
39073907
from line 3923
39233923
39243924 - `BetaManagedAgentsAutoPolicy object`
39253925
3926 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.
3926 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
39273927
39283928 - `mcp_server_name: string`
39293929
from line 4193
41934193
41944194 - `BetaManagedAgentsAutoPolicy object`
41954195
4196 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.
4196 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
41974197
41984198 - `type: "auto"`
41994199
from line 4221
42214221
42224222 - `BetaManagedAgentsAutoPolicy object`
42234223
4224 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.
4224 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
42254225
42264226 - `BetaManagedAgentsReadToolConfig object`
42274227
from line 4247
42474247
42484248 - `BetaManagedAgentsAutoPolicy object`
42494249
4250 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.
4250 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
42514251
42524252 - `BetaManagedAgentsWriteToolConfig object`
42534253
from line 4273
42734273
42744274 - `BetaManagedAgentsAutoPolicy object`
42754275
4276 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.
4276 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
42774277
42784278 - `BetaManagedAgentsGlobToolConfig object`
42794279
from line 4299
42994299
43004300 - `BetaManagedAgentsAutoPolicy object`
43014301
4302 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.
4302 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
43034303
43044304 - `BetaManagedAgentsGrepToolConfig object`
43054305
from line 4325
43254325
43264326 - `BetaManagedAgentsAutoPolicy object`
43274327
4328 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.
4328 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
43294329
43304330 - `BetaManagedAgentsWebFetchToolConfig object`
43314331
from line 4351
43514351
43524352 - `BetaManagedAgentsAutoPolicy object`
43534353
4354 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.
4354 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
43554355
43564356 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
43574357
from line 4461
44614461
44624462 - `BetaManagedAgentsAutoPolicy object`
44634463
4464 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.
4464 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
44654465
44664466 - `allowed_domains: optional array of string`
44674467
from line 4519
45194519
45204520 - `BetaManagedAgentsAutoPolicy object`
45214521
4522 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.
4522 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
45234523
45244524 - `BetaManagedAgentsMCPToolset object`
45254525
from line 4545
45454545
45464546 - `BetaManagedAgentsAutoPolicy object`
45474547
4548 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.
4548 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
45494549
45504550 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
45514551
from line 4567
45674567
45684568 - `BetaManagedAgentsAutoPolicy object`
45694569
4570 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.
4570 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
45714571
45724572 - `mcp_server_name: string`
45734573
from line 5373
53735373
53745374 - `BetaManagedAgentsAgentAutoEvaluatedPermissionDeny object`
53755375
5376 The server judged the invocation high-risk; it does not execute and a synthetic error tool result is appended.
5376 The server judged the invocation high-risk, or reached no judgement and did not hold it for client approval (for example, because the organization is out of credits or has reached a spend limit); it does not execute and a synthetic error tool result is appended.
53775377
53785378 - `type: "deny"`
53795379
53805380 - `reason_code: string`
53815381
5382 The judgement's grounds in registry-bound terms. Open registry; currently "high_risk" (judged high-risk; the call does not run). Clients must tolerate values outside this set.
5382 The judgement's grounds in registry-bound terms. Open registry; currently "high_risk" (judged high-risk; the call does not run) and "indeterminate" (no judgement was reached; the call does not run). Clients must tolerate values outside this set.
53835383
53845384 maxLength: 64
53855385
from line 6913
69136913
69146914 - `BetaManagedAgentsAutoPolicy object`
69156915
6916 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.
6916 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
69176917
69186918 - `type: "auto"`
69196919
from line 6941
69416941
69426942 - `BetaManagedAgentsAutoPolicy object`
69436943
6944 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.
6944 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
69456945
69466946 - `BetaManagedAgentsReadToolConfig object`
69476947
from line 6967
69676967
69686968 - `BetaManagedAgentsAutoPolicy object`
69696969
6970 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.
6970 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
69716971
69726972 - `BetaManagedAgentsWriteToolConfig object`
69736973
from line 6993
69936993
69946994 - `BetaManagedAgentsAutoPolicy object`
69956995
6996 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.
6996 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
69976997
69986998 - `BetaManagedAgentsGlobToolConfig object`
69996999
from line 7019
70197019
70207020 - `BetaManagedAgentsAutoPolicy object`
70217021
7022 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.
7022 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
70237023
70247024 - `BetaManagedAgentsGrepToolConfig object`
70257025
from line 7045
70457045
70467046 - `BetaManagedAgentsAutoPolicy object`
70477047
7048 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.
7048 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
70497049
70507050 - `BetaManagedAgentsWebFetchToolConfig object`
70517051
from line 7071
70717071
70727072 - `BetaManagedAgentsAutoPolicy object`
70737073
7074 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.
7074 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
70757075
70767076 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
70777077
from line 7181
71817181
71827182 - `BetaManagedAgentsAutoPolicy object`
71837183
7184 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.
7184 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
71857185
71867186 - `allowed_domains: optional array of string`
71877187
from line 7239
72397239
72407240 - `BetaManagedAgentsAutoPolicy object`
72417241
7242 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.
7242 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
72437243
72447244 - `BetaManagedAgentsMCPToolset object`
72457245
from line 7265
72657265
72667266 - `BetaManagedAgentsAutoPolicy object`
72677267
7268 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.
7268 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
72697269
72707270 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
72717271
from line 7287
72877287
72887288 - `BetaManagedAgentsAutoPolicy object`
72897289
7290 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.
7290 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
72917291
72927292 - `mcp_server_name: string`
72937293
from line 8563
85638563
85648564 - `BetaManagedAgentsAgentAutoEvaluatedPermissionDeny object`
85658565
8566 The server judged the invocation high-risk; it does not execute and a synthetic error tool result is appended.
8566 The server judged the invocation high-risk, or reached no judgement and did not hold it for client approval (for example, because the organization is out of credits or has reached a spend limit); it does not execute and a synthetic error tool result is appended.
85678567
85688568 - `type: "deny"`
85698569
85708570 - `reason_code: string`
85718571
8572 The judgement's grounds in registry-bound terms. Open registry; currently "high_risk" (judged high-risk; the call does not run). Clients must tolerate values outside this set.
8572 The judgement's grounds in registry-bound terms. Open registry; currently "high_risk" (judged high-risk; the call does not run) and "indeterminate" (no judgement was reached; the call does not run). Clients must tolerate values outside this set.
85738573
85748574 maxLength: 64
85758575
from line 10103
1010310103
1010410104 - `BetaManagedAgentsAutoPolicy object`
1010510105
10106 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.
10106 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1010710107
1010810108 - `type: "auto"`
1010910109
from line 10131
1013110131
1013210132 - `BetaManagedAgentsAutoPolicy object`
1013310133
10134 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.
10134 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1013510135
1013610136 - `BetaManagedAgentsReadToolConfig object`
1013710137
from line 10157
1015710157
1015810158 - `BetaManagedAgentsAutoPolicy object`
1015910159
10160 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.
10160 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1016110161
1016210162 - `BetaManagedAgentsWriteToolConfig object`
1016310163
from line 10183
1018310183
1018410184 - `BetaManagedAgentsAutoPolicy object`
1018510185
10186 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.
10186 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1018710187
1018810188 - `BetaManagedAgentsGlobToolConfig object`
1018910189
from line 10209
1020910209
1021010210 - `BetaManagedAgentsAutoPolicy object`
1021110211
10212 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.
10212 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1021310213
1021410214 - `BetaManagedAgentsGrepToolConfig object`
1021510215
from line 10235
1023510235
1023610236 - `BetaManagedAgentsAutoPolicy object`
1023710237
10238 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.
10238 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1023910239
1024010240 - `BetaManagedAgentsWebFetchToolConfig object`
1024110241
from line 10261
1026110261
1026210262 - `BetaManagedAgentsAutoPolicy object`
1026310263
10264 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.
10264 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1026510265
1026610266 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
1026710267
from line 10371
1037110371
1037210372 - `BetaManagedAgentsAutoPolicy object`
1037310373
10374 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.
10374 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1037510375
1037610376 - `allowed_domains: optional array of string`
1037710377
from line 10429
1042910429
1043010430 - `BetaManagedAgentsAutoPolicy object`
1043110431
10432 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.
10432 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1043310433
1043410434 - `BetaManagedAgentsMCPToolset object`
1043510435
from line 10455
1045510455
1045610456 - `BetaManagedAgentsAutoPolicy object`
1045710457
10458 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.
10458 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1045910459
1046010460 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
1046110461
from line 10477
1047710477
1047810478 - `BetaManagedAgentsAutoPolicy object`
1047910479
10480 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.
10480 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1048110481
1048210482 - `mcp_server_name: string`
1048310483
from line 11748
1174811748
1174911749 - `BetaManagedAgentsAgentAutoEvaluatedPermissionDeny object`
1175011750
11751 The server judged the invocation high-risk; it does not execute and a synthetic error tool result is appended.
11751 The server judged the invocation high-risk, or reached no judgement and did not hold it for client approval (for example, because the organization is out of credits or has reached a spend limit); it does not execute and a synthetic error tool result is appended.
1175211752
1175311753 - `type: "deny"`
1175411754
1175511755 - `reason_code: string`
1175611756
11757 The judgement's grounds in registry-bound terms. Open registry; currently "high_risk" (judged high-risk; the call does not run). Clients must tolerate values outside this set.
11757 The judgement's grounds in registry-bound terms. Open registry; currently "high_risk" (judged high-risk; the call does not run) and "indeterminate" (no judgement was reached; the call does not run). Clients must tolerate values outside this set.
1175811758
1175911759 maxLength: 64
1176011760
from line 13288
1328813288
1328913289 - `BetaManagedAgentsAutoPolicy object`
1329013290
13291 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.
13291 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1329213292
1329313293 - `type: "auto"`
1329413294
from line 13316
1331613316
1331713317 - `BetaManagedAgentsAutoPolicy object`
1331813318
13319 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.
13319 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1332013320
1332113321 - `BetaManagedAgentsReadToolConfig object`
1332213322
from line 13342
1334213342
1334313343 - `BetaManagedAgentsAutoPolicy object`
1334413344
13345 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.
13345 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1334613346
1334713347 - `BetaManagedAgentsWriteToolConfig object`
1334813348
from line 13368
1336813368
1336913369 - `BetaManagedAgentsAutoPolicy object`
1337013370
13371 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.
13371 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1337213372
1337313373 - `BetaManagedAgentsGlobToolConfig object`
1337413374
from line 13394
1339413394
1339513395 - `BetaManagedAgentsAutoPolicy object`
1339613396
13397 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.
13397 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1339813398
1339913399 - `BetaManagedAgentsGrepToolConfig object`
1340013400
from line 13420
1342013420
1342113421 - `BetaManagedAgentsAutoPolicy object`
1342213422
13423 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.
13423 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1342413424
1342513425 - `BetaManagedAgentsWebFetchToolConfig object`
1342613426
from line 13446
1344613446
1344713447 - `BetaManagedAgentsAutoPolicy object`
1344813448
13449 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.
13449 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1345013450
1345113451 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
1345213452
from line 13556
1355613556
1355713557 - `BetaManagedAgentsAutoPolicy object`
1355813558
13559 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.
13559 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1356013560
1356113561 - `allowed_domains: optional array of string`
1356213562
from line 13614
1361413614
1361513615 - `BetaManagedAgentsAutoPolicy object`
1361613616
13617 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.
13617 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1361813618
1361913619 - `BetaManagedAgentsMCPToolset object`
1362013620
from line 13640
1364013640
1364113641 - `BetaManagedAgentsAutoPolicy object`
1364213642
13643 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.
13643 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1364413644
1364513645 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
1364613646
from line 13662
1366213662
1366313663 - `BetaManagedAgentsAutoPolicy object`
1366413664
13665 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.
13665 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
1366613666
1366713667 - `mcp_server_name: string`
1366813668
1366913669
api/beta/sessions/threads/archive Changed · +11 / -11 lines
from line 393
393393
394394 - `BetaManagedAgentsAutoPolicy object`
395395
396 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.
396 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
397397
398398 - `type: "auto"`
399399
from line 421
421421
422422 - `BetaManagedAgentsAutoPolicy object`
423423
424 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.
424 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
425425
426426 - `BetaManagedAgentsReadToolConfig object`
427427
from line 447
447447
448448 - `BetaManagedAgentsAutoPolicy object`
449449
450 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.
450 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
451451
452452 - `BetaManagedAgentsWriteToolConfig object`
453453
from line 473
473473
474474 - `BetaManagedAgentsAutoPolicy object`
475475
476 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.
476 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
477477
478478 - `BetaManagedAgentsGlobToolConfig object`
479479
from line 499
499499
500500 - `BetaManagedAgentsAutoPolicy object`
501501
502 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.
502 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
503503
504504 - `BetaManagedAgentsGrepToolConfig object`
505505
from line 525
525525
526526 - `BetaManagedAgentsAutoPolicy object`
527527
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.
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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
529529
530530 - `BetaManagedAgentsWebFetchToolConfig object`
531531
from line 551
551551
552552 - `BetaManagedAgentsAutoPolicy object`
553553
554 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.
554 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
555555
556556 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
557557
from line 661
661661
662662 - `BetaManagedAgentsAutoPolicy object`
663663
664 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.
664 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
665665
666666 - `allowed_domains: optional array of string`
667667
from line 719
719719
720720 - `BetaManagedAgentsAutoPolicy object`
721721
722 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.
722 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
723723
724724 - `BetaManagedAgentsMCPToolset object`
725725
from line 745
745745
746746 - `BetaManagedAgentsAutoPolicy object`
747747
748 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.
748 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
749749
750750 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
751751
from line 767
767767
768768 - `BetaManagedAgentsAutoPolicy object`
769769
770 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.
770 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
771771
772772 - `mcp_server_name: string`
773773
api/beta/sessions/threads/events Changed · +26 / -26 lines
from line 625
625625
626626 - `BetaManagedAgentsAgentAutoEvaluatedPermissionDeny object`
627627
628 The server judged the invocation high-risk; it does not execute and a synthetic error tool result is appended.
628 The server judged the invocation high-risk, or reached no judgement and did not hold it for client approval (for example, because the organization is out of credits or has reached a spend limit); it does not execute and a synthetic error tool result is appended.
629629
630630 - `type: "deny"`
631631
632632 - `reason_code: string`
633633
634 The judgement's grounds in registry-bound terms. Open registry; currently "high_risk" (judged high-risk; the call does not run). Clients must tolerate values outside this set.
634 The judgement's grounds in registry-bound terms. Open registry; currently "high_risk" (judged high-risk; the call does not run) and "indeterminate" (no judgement was reached; the call does not run). Clients must tolerate values outside this set.
635635
636636 maxLength: 64
637637
from line 2165
21652165
21662166 - `BetaManagedAgentsAutoPolicy object`
21672167
2168 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.
2168 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
21692169
21702170 - `type: "auto"`
21712171
from line 2193
21932193
21942194 - `BetaManagedAgentsAutoPolicy object`
21952195
2196 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.
2196 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
21972197
21982198 - `BetaManagedAgentsReadToolConfig object`
21992199
from line 2219
22192219
22202220 - `BetaManagedAgentsAutoPolicy object`
22212221
2222 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.
2222 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
22232223
22242224 - `BetaManagedAgentsWriteToolConfig object`
22252225
from line 2245
22452245
22462246 - `BetaManagedAgentsAutoPolicy object`
22472247
2248 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.
2248 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
22492249
22502250 - `BetaManagedAgentsGlobToolConfig object`
22512251
from line 2271
22712271
22722272 - `BetaManagedAgentsAutoPolicy object`
22732273
2274 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.
2274 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
22752275
22762276 - `BetaManagedAgentsGrepToolConfig object`
22772277
from line 2297
22972297
22982298 - `BetaManagedAgentsAutoPolicy object`
22992299
2300 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.
2300 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
23012301
23022302 - `BetaManagedAgentsWebFetchToolConfig object`
23032303
from line 2323
23232323
23242324 - `BetaManagedAgentsAutoPolicy object`
23252325
2326 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.
2326 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
23272327
23282328 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
23292329
from line 2433
24332433
24342434 - `BetaManagedAgentsAutoPolicy object`
24352435
2436 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.
2436 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
24372437
24382438 - `allowed_domains: optional array of string`
24392439
from line 2491
24912491
24922492 - `BetaManagedAgentsAutoPolicy object`
24932493
2494 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.
2494 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
24952495
24962496 - `BetaManagedAgentsMCPToolset object`
24972497
from line 2517
25172517
25182518 - `BetaManagedAgentsAutoPolicy object`
25192519
2520 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.
2520 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
25212521
25222522 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
25232523
from line 2539
25392539
25402540 - `BetaManagedAgentsAutoPolicy object`
25412541
2542 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.
2542 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
25432543
25442544 - `mcp_server_name: string`
25452545
from line 3810
38103810
38113811 - `BetaManagedAgentsAgentAutoEvaluatedPermissionDeny object`
38123812
3813 The server judged the invocation high-risk; it does not execute and a synthetic error tool result is appended.
3813 The server judged the invocation high-risk, or reached no judgement and did not hold it for client approval (for example, because the organization is out of credits or has reached a spend limit); it does not execute and a synthetic error tool result is appended.
38143814
38153815 - `type: "deny"`
38163816
38173817 - `reason_code: string`
38183818
3819 The judgement's grounds in registry-bound terms. Open registry; currently "high_risk" (judged high-risk; the call does not run). Clients must tolerate values outside this set.
3819 The judgement's grounds in registry-bound terms. Open registry; currently "high_risk" (judged high-risk; the call does not run) and "indeterminate" (no judgement was reached; the call does not run). Clients must tolerate values outside this set.
38203820
38213821 maxLength: 64
38223822
from line 5350
53505350
53515351 - `BetaManagedAgentsAutoPolicy object`
53525352
5353 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.
5353 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
53545354
53555355 - `type: "auto"`
53565356
from line 5378
53785378
53795379 - `BetaManagedAgentsAutoPolicy object`
53805380
5381 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.
5381 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
53825382
53835383 - `BetaManagedAgentsReadToolConfig object`
53845384
from line 5404
54045404
54055405 - `BetaManagedAgentsAutoPolicy object`
54065406
5407 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.
5407 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
54085408
54095409 - `BetaManagedAgentsWriteToolConfig object`
54105410
from line 5430
54305430
54315431 - `BetaManagedAgentsAutoPolicy object`
54325432
5433 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.
5433 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
54345434
54355435 - `BetaManagedAgentsGlobToolConfig object`
54365436
from line 5456
54565456
54575457 - `BetaManagedAgentsAutoPolicy object`
54585458
5459 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.
5459 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
54605460
54615461 - `BetaManagedAgentsGrepToolConfig object`
54625462
from line 5482
54825482
54835483 - `BetaManagedAgentsAutoPolicy object`
54845484
5485 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.
5485 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
54865486
54875487 - `BetaManagedAgentsWebFetchToolConfig object`
54885488
from line 5508
55085508
55095509 - `BetaManagedAgentsAutoPolicy object`
55105510
5511 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.
5511 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
55125512
55135513 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
55145514
from line 5618
56185618
56195619 - `BetaManagedAgentsAutoPolicy object`
56205620
5621 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.
5621 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
56225622
56235623 - `allowed_domains: optional array of string`
56245624
from line 5676
56765676
56775677 - `BetaManagedAgentsAutoPolicy object`
56785678
5679 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.
5679 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
56805680
56815681 - `BetaManagedAgentsMCPToolset object`
56825682
from line 5702
57025702
57035703 - `BetaManagedAgentsAutoPolicy object`
57045704
5705 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.
5705 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
57065706
57075707 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
57085708
from line 5724
57245724
57255725 - `BetaManagedAgentsAutoPolicy object`
57265726
5727 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.
5727 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
57285728
57295729 - `mcp_server_name: string`
57305730
api/beta/sessions/threads/events/list Changed · +13 / -13 lines
from line 623
623623
624624 - `BetaManagedAgentsAgentAutoEvaluatedPermissionDeny object`
625625
626 The server judged the invocation high-risk; it does not execute and a synthetic error tool result is appended.
626 The server judged the invocation high-risk, or reached no judgement and did not hold it for client approval (for example, because the organization is out of credits or has reached a spend limit); it does not execute and a synthetic error tool result is appended.
627627
628628 - `type: "deny"`
629629
630630 - `reason_code: string`
631631
632 The judgement's grounds in registry-bound terms. Open registry; currently "high_risk" (judged high-risk; the call does not run). Clients must tolerate values outside this set.
632 The judgement's grounds in registry-bound terms. Open registry; currently "high_risk" (judged high-risk; the call does not run) and "indeterminate" (no judgement was reached; the call does not run). Clients must tolerate values outside this set.
633633
634634 maxLength: 64
635635
from line 2163
21632163
21642164 - `BetaManagedAgentsAutoPolicy object`
21652165
2166 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.
2166 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
21672167
21682168 - `type: "auto"`
21692169
from line 2191
21912191
21922192 - `BetaManagedAgentsAutoPolicy object`
21932193
2194 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.
2194 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
21952195
21962196 - `BetaManagedAgentsReadToolConfig object`
21972197
from line 2217
22172217
22182218 - `BetaManagedAgentsAutoPolicy object`
22192219
2220 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.
2220 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
22212221
22222222 - `BetaManagedAgentsWriteToolConfig object`
22232223
from line 2243
22432243
22442244 - `BetaManagedAgentsAutoPolicy object`
22452245
2246 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.
2246 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
22472247
22482248 - `BetaManagedAgentsGlobToolConfig object`
22492249
from line 2269
22692269
22702270 - `BetaManagedAgentsAutoPolicy object`
22712271
2272 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.
2272 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
22732273
22742274 - `BetaManagedAgentsGrepToolConfig object`
22752275
from line 2295
22952295
22962296 - `BetaManagedAgentsAutoPolicy object`
22972297
2298 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.
2298 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
22992299
23002300 - `BetaManagedAgentsWebFetchToolConfig object`
23012301
from line 2321
23212321
23222322 - `BetaManagedAgentsAutoPolicy object`
23232323
2324 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.
2324 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
23252325
23262326 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
23272327
from line 2431
24312431
24322432 - `BetaManagedAgentsAutoPolicy object`
24332433
2434 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.
2434 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
24352435
24362436 - `allowed_domains: optional array of string`
24372437
from line 2489
24892489
24902490 - `BetaManagedAgentsAutoPolicy object`
24912491
2492 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.
2492 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
24932493
24942494 - `BetaManagedAgentsMCPToolset object`
24952495
from line 2515
25152515
25162516 - `BetaManagedAgentsAutoPolicy object`
25172517
2518 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.
2518 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
25192519
25202520 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
25212521
from line 2537
25372537
25382538 - `BetaManagedAgentsAutoPolicy object`
25392539
2540 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.
2540 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
25412541
25422542 - `mcp_server_name: string`
25432543
api/beta/sessions/threads/events/stream Changed · +13 / -13 lines
from line 625
625625
626626 - `BetaManagedAgentsAgentAutoEvaluatedPermissionDeny object`
627627
628 The server judged the invocation high-risk; it does not execute and a synthetic error tool result is appended.
628 The server judged the invocation high-risk, or reached no judgement and did not hold it for client approval (for example, because the organization is out of credits or has reached a spend limit); it does not execute and a synthetic error tool result is appended.
629629
630630 - `type: "deny"`
631631
632632 - `reason_code: string`
633633
634 The judgement's grounds in registry-bound terms. Open registry; currently "high_risk" (judged high-risk; the call does not run). Clients must tolerate values outside this set.
634 The judgement's grounds in registry-bound terms. Open registry; currently "high_risk" (judged high-risk; the call does not run) and "indeterminate" (no judgement was reached; the call does not run). Clients must tolerate values outside this set.
635635
636636 maxLength: 64
637637
from line 2165
21652165
21662166 - `BetaManagedAgentsAutoPolicy object`
21672167
2168 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.
2168 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
21692169
21702170 - `type: "auto"`
21712171
from line 2193
21932193
21942194 - `BetaManagedAgentsAutoPolicy object`
21952195
2196 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.
2196 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
21972197
21982198 - `BetaManagedAgentsReadToolConfig object`
21992199
from line 2219
22192219
22202220 - `BetaManagedAgentsAutoPolicy object`
22212221
2222 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.
2222 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
22232223
22242224 - `BetaManagedAgentsWriteToolConfig object`
22252225
from line 2245
22452245
22462246 - `BetaManagedAgentsAutoPolicy object`
22472247
2248 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.
2248 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
22492249
22502250 - `BetaManagedAgentsGlobToolConfig object`
22512251
from line 2271
22712271
22722272 - `BetaManagedAgentsAutoPolicy object`
22732273
2274 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.
2274 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
22752275
22762276 - `BetaManagedAgentsGrepToolConfig object`
22772277
from line 2297
22972297
22982298 - `BetaManagedAgentsAutoPolicy object`
22992299
2300 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.
2300 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
23012301
23022302 - `BetaManagedAgentsWebFetchToolConfig object`
23032303
from line 2323
23232323
23242324 - `BetaManagedAgentsAutoPolicy object`
23252325
2326 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.
2326 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
23272327
23282328 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
23292329
from line 2433
24332433
24342434 - `BetaManagedAgentsAutoPolicy object`
24352435
2436 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.
2436 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
24372437
24382438 - `allowed_domains: optional array of string`
24392439
from line 2491
24912491
24922492 - `BetaManagedAgentsAutoPolicy object`
24932493
2494 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.
2494 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
24952495
24962496 - `BetaManagedAgentsMCPToolset object`
24972497
from line 2517
25172517
25182518 - `BetaManagedAgentsAutoPolicy object`
25192519
2520 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.
2520 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
25212521
25222522 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
25232523
from line 2539
25392539
25402540 - `BetaManagedAgentsAutoPolicy object`
25412541
2542 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.
2542 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
25432543
25442544 - `mcp_server_name: string`
25452545
api/beta/sessions/threads/list Changed · +11 / -11 lines
from line 417
417417
418418 - `BetaManagedAgentsAutoPolicy object`
419419
420 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.
420 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
421421
422422 - `type: "auto"`
423423
from line 445
445445
446446 - `BetaManagedAgentsAutoPolicy object`
447447
448 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.
448 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
449449
450450 - `BetaManagedAgentsReadToolConfig object`
451451
from line 471
471471
472472 - `BetaManagedAgentsAutoPolicy object`
473473
474 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.
474 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
475475
476476 - `BetaManagedAgentsWriteToolConfig object`
477477
from line 497
497497
498498 - `BetaManagedAgentsAutoPolicy object`
499499
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.
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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
501501
502502 - `BetaManagedAgentsGlobToolConfig object`
503503
from line 523
523523
524524 - `BetaManagedAgentsAutoPolicy object`
525525
526 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.
526 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
527527
528528 - `BetaManagedAgentsGrepToolConfig object`
529529
from line 549
549549
550550 - `BetaManagedAgentsAutoPolicy object`
551551
552 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.
552 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
553553
554554 - `BetaManagedAgentsWebFetchToolConfig object`
555555
from line 575
575575
576576 - `BetaManagedAgentsAutoPolicy object`
577577
578 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.
578 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
579579
580580 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
581581
from line 685
685685
686686 - `BetaManagedAgentsAutoPolicy object`
687687
688 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.
688 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
689689
690690 - `allowed_domains: optional array of string`
691691
from line 743
743743
744744 - `BetaManagedAgentsAutoPolicy object`
745745
746 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.
746 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
747747
748748 - `BetaManagedAgentsMCPToolset object`
749749
from line 769
769769
770770 - `BetaManagedAgentsAutoPolicy object`
771771
772 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.
772 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
773773
774774 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
775775
from line 791
791791
792792 - `BetaManagedAgentsAutoPolicy object`
793793
794 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.
794 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
795795
796796 - `mcp_server_name: string`
797797
api/beta/sessions/threads/retrieve Changed · +11 / -11 lines
from line 393
393393
394394 - `BetaManagedAgentsAutoPolicy object`
395395
396 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.
396 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
397397
398398 - `type: "auto"`
399399
from line 421
421421
422422 - `BetaManagedAgentsAutoPolicy object`
423423
424 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.
424 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
425425
426426 - `BetaManagedAgentsReadToolConfig object`
427427
from line 447
447447
448448 - `BetaManagedAgentsAutoPolicy object`
449449
450 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.
450 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
451451
452452 - `BetaManagedAgentsWriteToolConfig object`
453453
from line 473
473473
474474 - `BetaManagedAgentsAutoPolicy object`
475475
476 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.
476 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
477477
478478 - `BetaManagedAgentsGlobToolConfig object`
479479
from line 499
499499
500500 - `BetaManagedAgentsAutoPolicy object`
501501
502 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.
502 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
503503
504504 - `BetaManagedAgentsGrepToolConfig object`
505505
from line 525
525525
526526 - `BetaManagedAgentsAutoPolicy object`
527527
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.
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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
529529
530530 - `BetaManagedAgentsWebFetchToolConfig object`
531531
from line 551
551551
552552 - `BetaManagedAgentsAutoPolicy object`
553553
554 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.
554 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
555555
556556 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
557557
from line 661
661661
662662 - `BetaManagedAgentsAutoPolicy object`
663663
664 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.
664 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
665665
666666 - `allowed_domains: optional array of string`
667667
from line 719
719719
720720 - `BetaManagedAgentsAutoPolicy object`
721721
722 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.
722 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
723723
724724 - `BetaManagedAgentsMCPToolset object`
725725
from line 745
745745
746746 - `BetaManagedAgentsAutoPolicy object`
747747
748 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.
748 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
749749
750750 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
751751
from line 767
767767
768768 - `BetaManagedAgentsAutoPolicy object`
769769
770 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.
770 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
771771
772772 - `mcp_server_name: string`
773773
api/beta/sessions/update Changed · +22 / -22 lines
from line 215
215215
216216 - `BetaManagedAgentsAutoPolicy object`
217217
218 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.
218 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
219219
220220 - `type: "auto"`
221221
from line 247
247247
248248 - `BetaManagedAgentsAutoPolicy object`
249249
250 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.
250 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
251251
252252 - `BetaManagedAgentsReadToolConfigParams object`
253253
from line 277
277277
278278 - `BetaManagedAgentsAutoPolicy object`
279279
280 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.
280 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
281281
282282 - `BetaManagedAgentsWriteToolConfigParams object`
283283
from line 307
307307
308308 - `BetaManagedAgentsAutoPolicy object`
309309
310 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.
310 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
311311
312312 - `BetaManagedAgentsGlobToolConfigParams object`
313313
from line 337
337337
338338 - `BetaManagedAgentsAutoPolicy object`
339339
340 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.
340 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
341341
342342 - `BetaManagedAgentsGrepToolConfigParams object`
343343
from line 367
367367
368368 - `BetaManagedAgentsAutoPolicy object`
369369
370 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.
370 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
371371
372372 - `BetaManagedAgentsWebFetchToolConfigParams object`
373373
from line 411
411411
412412 - `BetaManagedAgentsAutoPolicy object`
413413
414 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.
414 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
415415
416416 - `url_sources: optional BetaManagedAgentsWebFetchURLSourcesParams or null`
417417
from line 545
545545
546546 - `BetaManagedAgentsAutoPolicy object`
547547
548 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.
548 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
549549
550550 - `user_location: optional BetaManagedAgentsUserLocation or null`
551551
from line 601
601601
602602 - `BetaManagedAgentsAutoPolicy object`
603603
604 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.
604 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
605605
606606 - `BetaManagedAgentsMCPToolsetParams object`
607607
from line 643
643643
644644 - `BetaManagedAgentsAutoPolicy object`
645645
646 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.
646 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
647647
648648 - `default_config: optional BetaManagedAgentsMCPToolsetDefaultConfigParams or null`
649649
from line 667
667667
668668 - `BetaManagedAgentsAutoPolicy object`
669669
670 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.
670 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
671671
672672 - `BetaManagedAgentsCustomToolParams object`
673673
from line 1011
10111011
10121012 - `BetaManagedAgentsAutoPolicy object`
10131013
1014 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.
1014 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
10151015
10161016 - `type: "auto"`
10171017
from line 1039
10391039
10401040 - `BetaManagedAgentsAutoPolicy object`
10411041
1042 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.
1042 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
10431043
10441044 - `BetaManagedAgentsReadToolConfig object`
10451045
from line 1065
10651065
10661066 - `BetaManagedAgentsAutoPolicy object`
10671067
1068 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.
1068 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
10691069
10701070 - `BetaManagedAgentsWriteToolConfig object`
10711071
from line 1091
10911091
10921092 - `BetaManagedAgentsAutoPolicy object`
10931093
1094 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.
1094 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
10951095
10961096 - `BetaManagedAgentsGlobToolConfig object`
10971097
from line 1117
11171117
11181118 - `BetaManagedAgentsAutoPolicy object`
11191119
1120 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.
1120 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
11211121
11221122 - `BetaManagedAgentsGrepToolConfig object`
11231123
from line 1143
11431143
11441144 - `BetaManagedAgentsAutoPolicy object`
11451145
1146 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.
1146 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
11471147
11481148 - `BetaManagedAgentsWebFetchToolConfig object`
11491149
from line 1169
11691169
11701170 - `BetaManagedAgentsAutoPolicy object`
11711171
1172 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.
1172 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
11731173
11741174 - `url_sources: BetaManagedAgentsWebFetchURLSources or null`
11751175
from line 1279
12791279
12801280 - `BetaManagedAgentsAutoPolicy object`
12811281
1282 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.
1282 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
12831283
12841284 - `allowed_domains: optional array of string`
12851285
from line 1337
13371337
13381338 - `BetaManagedAgentsAutoPolicy object`
13391339
1340 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.
1340 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
13411341
13421342 - `BetaManagedAgentsMCPToolset object`
13431343
from line 1363
13631363
13641364 - `BetaManagedAgentsAutoPolicy object`
13651365
1366 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.
1366 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
13671367
13681368 - `default_config: BetaManagedAgentsMCPToolsetDefaultConfig`
13691369
from line 1385
13851385
13861386 - `BetaManagedAgentsAutoPolicy object`
13871387
1388 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.
1388 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, or for some causes to deny, for example when the organization is out of credits or has reached a spend limit.
13891389
13901390 - `mcp_server_name: string`
13911391
api/compliance Changed · +280 / -206 lines
This page is larger than the 256 KiB this site keeps, so one side of the diff below stops where the stored text does.
from line 1216
12161216
12171217 - `"mcp_server_managed_auth_token_exchanged"`
12181218
1219 A user attempted to obtain an access token for an MCP server via enterprise managed authorization. This event reports the outcomes of attempted token exchanges. Repeated failures with the same cause may be reported once until the cause changes, and requests refused before a token exchange is attempted are not reported, except the "connector_scope_not_granted" and "identity_assertion_refused" failures described under error_type.
1219 A user attempted to obtain an access token for an MCP server via enterprise managed authorization (auto-connect). This event reports the outcomes of attempted token exchanges. Repeated failures with the same cause may be reported once until the cause changes, and requests refused before a token exchange is attempted are not reported, except the "connector_scope_not_granted" and "identity_assertion_refused" failures described under error_type.
12201220
12211221 - `"mcp_server_managed_auth_updated"`
12221222
1223 An MCP server's enterprise managed authorization settings were set, changed, or cleared, including when they were supplied while the server was being added or edited. Fields without a "previous_" prefix describe the settings after the change and are null when the server has no managed authorization settings afterwards; "previous_" fields describe the settings before the change and are null when the server had none before (always the case for a newly added server).
1223 An MCP server's enterprise managed authorization (auto-connect) settings were set, changed, or cleared, including when they were supplied while the server was being added or edited. Fields without a "previous_" prefix describe the settings after the change and are null when the server has no managed authorization settings afterwards; "previous_" fields describe the settings before the change and are null when the server had none before (always the case for a newly added server).
12241224
12251225 - `"mcp_server_updated"`
12261226
from line 2167
21672167
21682168 - `"rbac_role_permission_added"`
21692169
2170 Admin added a permission to an RBAC custom role.
2170 A permission was added to an RBAC custom role.
21712171
2172 Emitted once per requested permission, including permissions the role
2173 already had, so a retried request still produces a complete audit record.
2172 A request records one activity per permission it asks to add, including any
2173 the role already had, so a retry still leaves a complete audit record.
21742174
21752175 - `"rbac_role_permission_removed"`
21762176
2177 Admin removed a permission from an RBAC custom role.
2177 A permission was removed from an RBAC custom role.
21782178
2179 Emitted once per requested permission, including permissions the role
2180 already lacked, so a retried request still produces a complete audit
2179 A request records one activity per permission it asks to remove, including
2180 any the role already lacked, so a retry still leaves a complete audit
21812181 record.
21822182
21832183 - `"rbac_role_unassigned"`
from line 3646
36463646
36473647 - `"mcp_server_managed_auth_token_exchanged"`
36483648
3649 A user attempted to obtain an access token for an MCP server via enterprise managed authorization. This event reports the outcomes of attempted token exchanges. Repeated failures with the same cause may be reported once until the cause changes, and requests refused before a token exchange is attempted are not reported, except the "connector_scope_not_granted" and "identity_assertion_refused" failures described under error_type.
3649 A user attempted to obtain an access token for an MCP server via enterprise managed authorization (auto-connect). This event reports the outcomes of attempted token exchanges. Repeated failures with the same cause may be reported once until the cause changes, and requests refused before a token exchange is attempted are not reported, except the "connector_scope_not_granted" and "identity_assertion_refused" failures described under error_type.
36503650
36513651 - `"mcp_server_managed_auth_updated"`
36523652
3653 An MCP server's enterprise managed authorization settings were set, changed, or cleared, including when they were supplied while the server was being added or edited. Fields without a "previous_" prefix describe the settings after the change and are null when the server has no managed authorization settings afterwards; "previous_" fields describe the settings before the change and are null when the server had none before (always the case for a newly added server).
3653 An MCP server's enterprise managed authorization (auto-connect) settings were set, changed, or cleared, including when they were supplied while the server was being added or edited. Fields without a "previous_" prefix describe the settings after the change and are null when the server has no managed authorization settings afterwards; "previous_" fields describe the settings before the change and are null when the server had none before (always the case for a newly added server).
36543654
36553655 - `"mcp_server_updated"`
36563656
from line 4597
45974597
45984598 - `"rbac_role_permission_added"`
45994599
4600 Admin added a permission to an RBAC custom role.
4600 A permission was added to an RBAC custom role.
46014601
4602 Emitted once per requested permission, including permissions the role
4603 already had, so a retried request still produces a complete audit record.
4602 A request records one activity per permission it asks to add, including any
4603 the role already had, so a retry still leaves a complete audit record.
46044604
46054605 - `"rbac_role_permission_removed"`
46064606
4607 Admin removed a permission from an RBAC custom role.
4607 A permission was removed from an RBAC custom role.
46084608
4609 Emitted once per requested permission, including permissions the role
4610 already lacked, so a retried request still produces a complete audit
4609 A request records one activity per permission it asks to remove, including
4610 any the role already lacked, so a retry still leaves a complete audit
46114611 record.
46124612
46134613 - `"rbac_role_unassigned"`
from line 4867
48674867
48684868#### Headers
48694869
4870- `"anthropic-version": optional string`
4870- `"anthropic-version": string`
48714871
4872 The version of the Claude API you want to use.
4872 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
48734873
4874 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
4874 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
48754875
48764876- `"x-api-key": optional string`
48774877
from line 8976
89768976
89778977 default: attested_device_actor
89788978
8979 - `external_client_id: string`
8980
8981 - `kid_hash: string`
8982
8983 - `ip_address: optional string or null`
8984
8985 - `user_agent: optional string or null`
8986
8987 - `claude_artifact_id: string`
8988
8989 The artifact's identifier.
8990
8991 - `comment_action: "activate_thread" or "create_thread
8979
api/compliance/activities Changed · +174 / -58 lines
This page is larger than the 256 KiB this site keeps, so one side of the diff below stops where the stored text does.
from line 1214
12141214
12151215 - `"mcp_server_managed_auth_token_exchanged"`
12161216
1217 A user attempted to obtain an access token for an MCP server via enterprise managed authorization. This event reports the outcomes of attempted token exchanges. Repeated failures with the same cause may be reported once until the cause changes, and requests refused before a token exchange is attempted are not reported, except the "connector_scope_not_granted" and "identity_assertion_refused" failures described under error_type.
1217 A user attempted to obtain an access token for an MCP server via enterprise managed authorization (auto-connect). This event reports the outcomes of attempted token exchanges. Repeated failures with the same cause may be reported once until the cause changes, and requests refused before a token exchange is attempted are not reported, except the "connector_scope_not_granted" and "identity_assertion_refused" failures described under error_type.
12181218
12191219 - `"mcp_server_managed_auth_updated"`
12201220
1221 An MCP server's enterprise managed authorization settings were set, changed, or cleared, including when they were supplied while the server was being added or edited. Fields without a "previous_" prefix describe the settings after the change and are null when the server has no managed authorization settings afterwards; "previous_" fields describe the settings before the change and are null when the server had none before (always the case for a newly added server).
1221 An MCP server's enterprise managed authorization (auto-connect) settings were set, changed, or cleared, including when they were supplied while the server was being added or edited. Fields without a "previous_" prefix describe the settings after the change and are null when the server has no managed authorization settings afterwards; "previous_" fields describe the settings before the change and are null when the server had none before (always the case for a newly added server).
12221222
12231223 - `"mcp_server_updated"`
12241224
from line 2165
21652165
21662166 - `"rbac_role_permission_added"`
21672167
2168 Admin added a permission to an RBAC custom role.
2168 A permission was added to an RBAC custom role.
21692169
2170 Emitted once per requested permission, including permissions the role
2171 already had, so a retried request still produces a complete audit record.
2170 A request records one activity per permission it asks to add, including any
2171 the role already had, so a retry still leaves a complete audit record.
21722172
21732173 - `"rbac_role_permission_removed"`
21742174
2175 Admin removed a permission from an RBAC custom role.
2175 A permission was removed from an RBAC custom role.
21762176
2177 Emitted once per requested permission, including permissions the role
2178 already lacked, so a retried request still produces a complete audit
2177 A request records one activity per permission it asks to remove, including
2178 any the role already lacked, so a retry still leaves a complete audit
21792179 record.
21802180
21812181 - `"rbac_role_unassigned"`
from line 3644
36443644
36453645 - `"mcp_server_managed_auth_token_exchanged"`
36463646
3647 A user attempted to obtain an access token for an MCP server via enterprise managed authorization. This event reports the outcomes of attempted token exchanges. Repeated failures with the same cause may be reported once until the cause changes, and requests refused before a token exchange is attempted are not reported, except the "connector_scope_not_granted" and "identity_assertion_refused" failures described under error_type.
3647 A user attempted to obtain an access token for an MCP server via enterprise managed authorization (auto-connect). This event reports the outcomes of attempted token exchanges. Repeated failures with the same cause may be reported once until the cause changes, and requests refused before a token exchange is attempted are not reported, except the "connector_scope_not_granted" and "identity_assertion_refused" failures described under error_type.
36483648
36493649 - `"mcp_server_managed_auth_updated"`
36503650
3651 An MCP server's enterprise managed authorization settings were set, changed, or cleared, including when they were supplied while the server was being added or edited. Fields without a "previous_" prefix describe the settings after the change and are null when the server has no managed authorization settings afterwards; "previous_" fields describe the settings before the change and are null when the server had none before (always the case for a newly added server).
3651 An MCP server's enterprise managed authorization (auto-connect) settings were set, changed, or cleared, including when they were supplied while the server was being added or edited. Fields without a "previous_" prefix describe the settings after the change and are null when the server has no managed authorization settings afterwards; "previous_" fields describe the settings before the change and are null when the server had none before (always the case for a newly added server).
36523652
36533653 - `"mcp_server_updated"`
36543654
from line 4595
45954595
45964596 - `"rbac_role_permission_added"`
45974597
4598 Admin added a permission to an RBAC custom role.
4598 A permission was added to an RBAC custom role.
45994599
4600 Emitted once per requested permission, including permissions the role
4601 already had, so a retried request still produces a complete audit record.
4600 A request records one activity per permission it asks to add, including any
4601 the role already had, so a retry still leaves a complete audit record.
46024602
46034603 - `"rbac_role_permission_removed"`
46044604
4605 Admin removed a permission from an RBAC custom role.
4605 A permission was removed from an RBAC custom role.
46064606
4607 Emitted once per requested permission, including permissions the role
4608 already lacked, so a retried request still produces a complete audit
4607 A request records one activity per permission it asks to remove, including
4608 any the role already lacked, so a retry still leaves a complete audit
46094609 record.
46104610
46114611 - `"rbac_role_unassigned"`
from line 4865
48654865
48664866### Headers
48674867
4868- `"anthropic-version": optional string`
4868- `"anthropic-version": string`
48694869
4870 The version of the Claude API you want to use.
4870 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
48714871
4872 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
4872 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
48734873
48744874- `"x-api-key": optional string`
48754875
from line 8976
89768976
89778977 - `external_client_id: string`
89788978
8979 - `kid_hash: string`
8980
8981 - `ip_address: optional string or null`
8982
8983 - `user_agent: optional string or null`
8984
8985 - `claude_artifact_id: string`
8986
8987 The artifact's identifier.
8988
8989 - `comment_action: "activate_thread" or "create_thread" or "deactivate_thread" or 10 more
8979
api/compliance/activities/list Changed · +97 / -39 lines
This page is larger than the 256 KiB this site keeps, so one side of the diff below stops where the stored text does.
from line 1212
12121212
12131213 - `"mcp_server_managed_auth_token_exchanged"`
12141214
1215 A user attempted to obtain an access token for an MCP server via enterprise managed authorization. This event reports the outcomes of attempted token exchanges. Repeated failures with the same cause may be reported once until the cause changes, and requests refused before a token exchange is attempted are not reported, except the "connector_scope_not_granted" and "identity_assertion_refused" failures described under error_type.
1215 A user attempted to obtain an access token for an MCP server via enterprise managed authorization (auto-connect). This event reports the outcomes of attempted token exchanges. Repeated failures with the same cause may be reported once until the cause changes, and requests refused before a token exchange is attempted are not reported, except the "connector_scope_not_granted" and "identity_assertion_refused" failures described under error_type.
12161216
12171217 - `"mcp_server_managed_auth_updated"`
12181218
1219 An MCP server's enterprise managed authorization settings were set, changed, or cleared, including when they were supplied while the server was being added or edited. Fields without a "previous_" prefix describe the settings after the change and are null when the server has no managed authorization settings afterwards; "previous_" fields describe the settings before the change and are null when the server had none before (always the case for a newly added server).
1219 An MCP server's enterprise managed authorization (auto-connect) settings were set, changed, or cleared, including when they were supplied while the server was being added or edited. Fields without a "previous_" prefix describe the settings after the change and are null when the server has no managed authorization settings afterwards; "previous_" fields describe the settings before the change and are null when the server had none before (always the case for a newly added server).
12201220
12211221 - `"mcp_server_updated"`
12221222
from line 2163
21632163
21642164 - `"rbac_role_permission_added"`
21652165
2166 Admin added a permission to an RBAC custom role.
2166 A permission was added to an RBAC custom role.
21672167
2168 Emitted once per requested permission, including permissions the role
2169 already had, so a retried request still produces a complete audit record.
2168 A request records one activity per permission it asks to add, including any
2169 the role already had, so a retry still leaves a complete audit record.
21702170
21712171 - `"rbac_role_permission_removed"`
21722172
2173 Admin removed a permission from an RBAC custom role.
2173 A permission was removed from an RBAC custom role.
21742174
2175 Emitted once per requested permission, including permissions the role
2176 already lacked, so a retried request still produces a complete audit
2175 A request records one activity per permission it asks to remove, including
2176 any the role already lacked, so a retry still leaves a complete audit
21772177 record.
21782178
21792179 - `"rbac_role_unassigned"`
from line 3642
36423642
36433643 - `"mcp_server_managed_auth_token_exchanged"`
36443644
3645 A user attempted to obtain an access token for an MCP server via enterprise managed authorization. This event reports the outcomes of attempted token exchanges. Repeated failures with the same cause may be reported once until the cause changes, and requests refused before a token exchange is attempted are not reported, except the "connector_scope_not_granted" and "identity_assertion_refused" failures described under error_type.
3645 A user attempted to obtain an access token for an MCP server via enterprise managed authorization (auto-connect). This event reports the outcomes of attempted token exchanges. Repeated failures with the same cause may be reported once until the cause changes, and requests refused before a token exchange is attempted are not reported, except the "connector_scope_not_granted" and "identity_assertion_refused" failures described under error_type.
36463646
36473647 - `"mcp_server_managed_auth_updated"`
36483648
3649 An MCP server's enterprise managed authorization settings were set, changed, or cleared, including when they were supplied while the server was being added or edited. Fields without a "previous_" prefix describe the settings after the change and are null when the server has no managed authorization settings afterwards; "previous_" fields describe the settings before the change and are null when the server had none before (always the case for a newly added server).
3649 An MCP server's enterprise managed authorization (auto-connect) settings were set, changed, or cleared, including when they were supplied while the server was being added or edited. Fields without a "previous_" prefix describe the settings after the change and are null when the server has no managed authorization settings afterwards; "previous_" fields describe the settings before the change and are null when the server had none before (always the case for a newly added server).
36503650
36513651 - `"mcp_server_updated"`
36523652
from line 4593
45934593
45944594 - `"rbac_role_permission_added"`
45954595
4596 Admin added a permission to an RBAC custom role.
4596 A permission was added to an RBAC custom role.
45974597
4598 Emitted once per requested permission, including permissions the role
4599 already had, so a retried request still produces a complete audit record.
4598 A request records one activity per permission it asks to add, including any
4599 the role already had, so a retry still leaves a complete audit record.
46004600
46014601 - `"rbac_role_permission_removed"`
46024602
4603 Admin removed a permission from an RBAC custom role.
4603 A permission was removed from an RBAC custom role.
46044604
4605 Emitted once per requested permission, including permissions the role
4606 already lacked, so a retried request still produces a complete audit
4605 A request records one activity per permission it asks to remove, including
4606 any the role already lacked, so a retry still leaves a complete audit
46074607 record.
46084608
46094609 - `"rbac_role_unassigned"`
from line 4863
48634863
48644864## Headers
48654865
4866- `"anthropic-version": optional string`
4866- `"anthropic-version": string`
48674867
4868 The version of the Claude API you want to use.
4868 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
48694869
4870 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
4870 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
48714871
48724872- `"x-api-key": optional string`
48734873
from line 8973
89738973 default: attested_device_actor
89748974
89758975 - `external_client_id: string`
8976
8977 - `kid_hash: string`
8978
8979 - `ip_address: optional string or null`
8980
8981 - `user_agent: optional string or null`
8982
8983 - `claude_artifact_id: string`
8984
8985 The artifact's identifier.
8986
8987 - `comment_action: "activate_thread" or "create_thread" or "deactivate_thread" or 10
api/compliance/apps Changed · +180 / -176 lines
from line 123
123123
124124#### Headers
125125
126- `"anthropic-version": optional string`
126- `"anthropic-version": string`
127127
128 The version of the Claude API you want to use.
128 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
129129
130 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
130 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
131131
132132- `"x-api-key": optional string`
133133
from line 213
213213
214214```bash
215215curl https://api.anthropic.com/v1/compliance/apps/chats \
216 -H 'anthropic-version: 2023-06-01' \
216 -H 'anthropic-version: 2026-10-21' \
217217 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
218218```
219219
from line 264
264264
265265#### Headers
266266
267- `"anthropic-version": optional string`
267- `"anthropic-version": string`
268268
269 The version of the Claude API you want to use.
269 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
270270
271 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
271 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
272272
273273- `"x-api-key": optional string`
274274
from line 289
289289```bash
290290curl https://api.anthropic.com/v1/compliance/apps/chats/$CLAUDE_CHAT_ID \
291291 -X DELETE \
292 -H 'anthropic-version: 2023-06-01' \
292 -H 'anthropic-version: 2026-10-21' \
293293 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
294294```
295295
from line 408
408408
409409#### Headers
410410
411- `"anthropic-version": optional string`
411- `"anthropic-version": string`
412412
413 The version of the Claude API you want to use.
413 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
414414
415 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
415 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
416416
417417- `"x-api-key": optional string`
418418
from line 706
706706
707707```bash
708708curl https://api.anthropic.com/v1/compliance/apps/chats/$CLAUDE_CHAT_ID/messages \
709 -H 'anthropic-version: 2023-06-01' \
709 -H 'anthropic-version: 2026-10-21' \
710710 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
711711```
712712
from line 851
851851
852852#### Headers
853853
854- `"anthropic-version": optional string`
854- `"anthropic-version": string`
855855
856 The version of the Claude API you want to use.
856 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
857857
858 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
858 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
859859
860860- `"x-api-key": optional string`
861861
from line 899
899899
900900```bash
901901curl https://api.anthropic.com/v1/compliance/apps/chats/files/$CLAUDE_FILE_ID \
902 -H 'anthropic-version: 2023-06-01' \
902 -H 'anthropic-version: 2026-10-21' \
903903 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
904904```
905905
from line 937
937937
938938#### Headers
939939
940- `"anthropic-version": optional string`
940- `"anthropic-version": string`
941941
942 The version of the Claude API you want to use.
942 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
943943
944 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
944 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
945945
946946- `"x-api-key": optional string`
947947
from line 962
962962```bash
963963curl https://api.anthropic.com/v1/compliance/apps/chats/files/$CLAUDE_FILE_ID \
964964 -X DELETE \
965 -H 'anthropic-version: 2023-06-01' \
965 -H 'anthropic-version: 2026-10-21' \
966966 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
967967```
968968
from line 989
989989
990990#### Headers
991991
992- `"anthropic-version": optional string`
992- `"anthropic-version": string`
993993
994 The version of the Claude API you want to use.
994 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
995995
996 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
996 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
997997
998998- `"x-api-key": optional string`
999999
from line 1001
10011001
10021002```bash
10031003curl https://api.anthropic.com/v1/compliance/apps/chats/files/$CLAUDE_FILE_ID/content \
1004 -H 'anthropic-version: 2023-06-01' \
1004 -H 'anthropic-version: 2026-10-21' \
10051005 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
10061006```
10071007
from line 1023
10231023
10241024#### Headers
10251025
1026- `"anthropic-version": optional string`
1026- `"anthropic-version": string`
10271027
1028 The version of the Claude API you want to use.
1028 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
10291029
1030 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
1030 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
10311031
10321032- `"x-api-key": optional string`
10331033
from line 1067
10671067
10681068```bash
10691069curl https://api.anthropic.com/v1/compliance/apps/chats/generated-files/$CLAUDE_GEN_FILE_ID \
1070 -H 'anthropic-version: 2023-06-01' \
1070 -H 'anthropic-version: 2026-10-21' \
10711071 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
10721072```
10731073
from line 1099
10991099
11001100#### Headers
11011101
1102- `"anthropic-version": optional string`
1102- `"anthropic-version": string`
11031103
1104 The version of the Claude API you want to use.
1104 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
11051105
1106 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
1106 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
11071107
11081108- `"x-api-key": optional string`
11091109
from line 1111
11111111
11121112```bash
11131113curl https://api.anthropic.com/v1/compliance/apps/chats/generated-files/$CLAUDE_GEN_FILE_ID/content \
1114 -H 'anthropic-version: 2023-06-01' \
1114 -H 'anthropic-version: 2026-10-21' \
11151115 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
11161116```
11171117
from line 1198
11981198
11991199#### Headers
12001200
1201- `"anthropic-version": optional string`
1201- `"anthropic-version": string`
12021202
1203 The version of the Claude API you want to use.
1203 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
12041204
1205 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
1205 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
12061206
12071207- `"x-api-key": optional string`
12081208
from line 1280
12801280
12811281```bash
12821282curl https://api.anthropic.com/v1/compliance/apps/projects \
1283 -H 'anthropic-version: 2023-06-01' \
1283 -H 'anthropic-version: 2026-10-21' \
12841284 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
12851285```
12861286
from line 1323
13231323
13241324#### Headers
13251325
1326- `"anthropic-version": optional string`
1326- `"anthropic-version": string`
13271327
1328 The version of the Claude API you want to use.
1328 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
13291329
1330 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
1330 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
13311331
13321332- `"x-api-key": optional string`
13331333
from line 1409
14091409
14101410```bash
14111411curl https://api.anthropic.com/v1/compliance/apps/projects/$PROJECT_ID \
1412 -H 'anthropic-version: 2023-06-01' \
1412 -H 'anthropic-version: 2026-10-21' \
14131413 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
14141414```
14151415
from line 1460
14601460
14611461#### Headers
14621462
1463- `"anthropic-version": optional string`
1463- `"anthropic-version": string`
14641464
1465 The version of the Claude API you want to use.
1465 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
14661466
1467 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
1467 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
14681468
14691469- `"x-api-key": optional string`
14701470
from line 1485
14851485```bash
14861486curl https://api.anthropic.com/v1/compliance/apps/projects/$PROJECT_ID \
14871487 -X DELETE \
1488 -H 'anthropic-version: 2023-06-01' \
1488 -H 'anthropic-version: 2026-10-21' \
14891489 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
14901490```
14911491
from line 1535
15351535
15361536#### Headers
15371537
1538- `"anthropic-version": optional string`
1538- `"anthropic-version": string`
15391539
1540 The version of the Claude API you want to use.
1540 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
15411541
1542 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
1542 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
15431543
15441544- `"x-api-key": optional string`
15451545
from line 1633
16331633
16341634```bash
16351635curl https://api.anthropic.com/v1/compliance/apps/projects/$PROJECT_ID/attachments \
1636 -H 'anthropic-version: 2023-06-01' \
1636 -H 'anthropic-version: 2026-10-21' \
16371637 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
16381638```
16391639
from line 1690
16901690
16911691#### Headers
16921692
1693- `"anthropic-version": optional string`
1693- `"anthropic-version": string`
16941694
1695 The version of the Claude API you want to use.
1695 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
16961696
1697 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
1697 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
16981698
16991699- `"x-api-key": optional string`
17001700
from line 1844
18441844
18451845```bash
18461846curl https://api.anthropic.com/v1/compliance/apps/projects/$PROJECT_ID/collaborators \
1847 -H 'anthropic-version: 2023-06-01' \
1847 -H 'anthropic-version: 2026-10-21' \
18481848 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
18491849```
18501850
from line 1881
18811881
18821882#### Headers
18831883
1884- `"anthropic-version": optional string`
1884- `"anthropic-version": string`
18851885
1886 The version of the Claude API you want to use.
1886 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
18871887
1888 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
1888 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
18891889
18901890- `"x-api-key": optional string`
18911891
from line 1925
19251925
19261926```bash
19271927curl https://api.anthropic.com/v1/compliance/apps/projects/documents/$DOCUMENT_ID \
1928 -H 'anthropic-version: 2023-06-01' \
1928 -H 'anthropic-version: 2026-10-21' \
19291929 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
19301930```
19311931
from line 1963
19631963
19641964#### Headers
19651965
1966- `"anthropic-version": optional string`
1966- `"anthropic-version": string`
19671967
1968 The version of the Claude API you want to use.
1968 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
19691969
1970 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
1970 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
19711971
19721972- `"x-api-key": optional string`
19731973
from line 2021
20212021
20222022```bash
20232023curl https://api.anthropic.com/v1/compliance/apps/projects/documents/$DOCUMENT_ID/metadata \
2024 -H 'anthropic-version: 2023-06-01' \
2024 -H 'anthropic-version: 2026-10-21' \
20252025 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
20262026```
20272027
from line 2059
20592059
20602060#### Headers
20612061
2062- `"anthropic-version": optional string`
2062- `"anthropic-version": string`
20632063
2064 The version of the Claude API you want to use.
2064 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
20652065
2066 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
2066 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
20672067
20682068- `"x-api-key": optional string`
20692069
from line 2084
20842084```bash
20852085curl https://api.anthropic.com/v1/compliance/apps/projects/documents/$DOCUMENT_ID \
20862086 -X DELETE \
2087 -H 'anthropic-version: 2023-06-01' \
2087 -H 'anthropic-version: 2026-10-21' \
20882088 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
20892089```
20902090
from line 2118
21182118
21192119#### Headers
21202120
2121- `"anthropic-version": optional string`
2121- `"anthropic-version": string`
21222122
2123 The version of the Claude API you want to use.
2123 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
21242124
2125 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
2125 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
21262126
21272127- `"x-api-key": optional string`
21282128
from line 2166
21662166
21672167```bash
21682168curl https://api.anthropic.com/v1/compliance/apps/artifacts/$ARTIFACT_VERSION_ID \
2169 -H 'anthropic-version: 2023-06-01' \
2169 -H 'anthropic-version: 2026-10-21' \
21702170 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
21712171```
21722172
from line 2201
22012201
22022202#### Headers
22032203
2204- `"anthropic-version": optional string`
2204- `"anthropic-version": string`
22052205
2206 The version of the Claude API you want to use.
2206 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
22072207
2208 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
2208 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
22092209
22102210- `"x-api-key": optional string`
22112211
from line 2213
22132213
22142214```bash
22152215curl https://api.anthropic.com/v1/compliance/apps/artifacts/$ARTIFACT_VERSION_ID/content \
2216 -H 'anthropic-version: 2023-06-01' \
2216 -H 'anthropic-version: 2026-10-21' \
22172217 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
22182218```
22192219
from line 2276
22762276
22772277#### Headers
22782278
2279- `"anthropic-version": optional string`
2279- `"anthropic-version": string`
22802280
2281 The version of the Claude API you want to use.
2281 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
22822282
2283 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
2283 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
22842284
22852285- `"x-api-key": optional string`
22862286
from line 2348
23482348
23492349```bash
23502350curl https://api.anthropic.com/v1/compliance/apps/sessions/local \
2351 -H 'anthropic-version: 2023-06-01' \
2351 -H 'anthropic-version: 2026-10-21' \
23522352 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
23532353```
23542354
from line 2391
23912391
23922392#### Headers
23932393
2394- `"anthropic-version": optional string`
2394- `"anthropic-version": string`
23952395
2396 The version of the Claude API you want to use.
2396 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
23972397
2398 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
2398 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
23992399
24002400- `"x-api-key": optional string`
24012401
from line 2455
24552455
24562456```bash
24572457curl https://api.anthropic.com/v1/compliance/apps/sessions/local/$LOCAL_SESSION_ID \
2458 -H 'anthropic-version: 2023-06-01' \
2458 -H 'anthropic-version: 2026-10-21' \
24592459 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
24602460```
24612461
from line 2540
25402540
25412541#### Headers
25422542
2543- `"anthropic-version": optional string`
2543- `"anthropic-version": string`
25442544
2545 The version of the Claude API you want to use.
2545 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
25462546
2547 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
2547 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
25482548
25492549- `"x-api-key": optional string`
25502550
from line 2774
27742774
27752775```bash
27762776curl https://api.anthropic.com/v1/compliance/apps/sessions/local/$LOCAL_SESSION_ID/messages \
2777 -H 'anthropic-version: 2023-06-01' \
2777 -H 'anthropic-version: 2026-10-21' \
27782778 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
27792779```
27802780
from line 2902
29022902
29032903#### Headers
29042904
2905- `"anthropic-version": optional string`
2905- `"anthropic-version": string`
29062906
2907 The version of the Claude API you want to use.
2907 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
29082908
2909 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
2909 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
29102910
29112911- `"x-api-key": optional string`
29122912
from line 2982
29822982
29832983```bash
29842984curl https://api.anthropic.com/v1/compliance/apps/sessions/remote \
2985 -H 'anthropic-version: 2023-06-01' \
2985 -H 'anthropic-version: 2026-10-21' \
29862986 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
29872987```
29882988
from line 3084
30843084
30853085#### Headers
30863086
3087- `"anthropic-version": optional string`
3087- `"anthropic-version": string`
30883088
3089 The version of the Claude API you want to use.
3089 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
30903090
3091 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
3091 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
30923092
30933093- `"x-api-key": optional string`
30943094
from line 3284
32843284
32853285```bash
32863286curl https://api.anthropic.com/v1/compliance/apps/sessions/remote/$CLAUDE_REMOTE_SESSION_ID/messages \
3287 -H 'anthropic-version: 2023-06-01' \
3287 -H 'anthropic-version: 2026-10-21' \
32883288 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
32893289```
32903290
from line 3401
34013401
34023402#### Headers
34033403
3404- `"anthropic-version": optional string`
3404- `"anthropic-version": string`
34053405
3406 The version of the Claude API you want to use.
3406 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
34073407
3408 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
3408 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
34093409
34103410- `"x-api-key": optional string`
34113411
from line 3419
34193419
34203420 Artifact identifier (tagged ID)
34213421
3422 - `artifact_type: "claude_dashboards" or "claude_design" or "claude_design_systems" or 5 more`
3422 - `artifact_type: "claude_dashboards" or "claude_design" or "claude_design_systems" or 6 more`
34233423
3424 Which kind of Artifact this is: `code` for a site published from Claude Code, or the built-in Artifact type it was made from — `claude_docs` (Claude Docs), `claude_slides` (Slides), `claude_design` (Design), `claude_design_systems` (a design system), `claude_motion` (Motion) or `claude_dashboards` (Dashboards). `other` is an Artifact made from a built-in type this list does not name yet.
3424 Which kind of Artifact this is: `code` for a site published from Claude Code, or the built-in Artifact type it was made from — `claude_docs` (Claude Docs), `claude_slides` (Slides), `claude_design` (Design), `claude_design_systems` (a design system), `claude_motion` (Motion), `claude_dashboards` (Dashboards) or `claude_sheets` (Claude Sheets). `other` is an Artifact made from a built-in type this list does not name yet.
34253425
34263426 - `"claude_dashboards"`
34273427
from line 3433
34333433
34343434 - `"claude_motion"`
34353435
3436 - `"claude_sheets"`
3437
34363438 - `"claude_slides"`
34373439
34383440 - `"code"`
from line 3451
34493451
34503452 - `published_version_id: string or null`
34513453
3452 Identifier of the version a non-owner viewer would render when `read_mode` permits them — the version the owner has pinned for non-owner readers if one is pinned, otherwise the owner's latest. When `read_mode` is `owner` no non-owner renders any version; the field still reports which version would be served were read_mode widened. Null for a Claude Doc; see `versions`.
3454 Identifier of the version a non-owner viewer would render when `read_mode` permits them — the version the owner has pinned for non-owner readers if one is pinned, otherwise the owner's latest. When `read_mode` is `owner` no non-owner renders any version; the field still reports which version would be served were read_mode widened. Null for a Claude Doc or a Claude Sheet; see `versions`.
34533455
34543456 - `read_mode: "org" or "owner" or "public" or "users"`
34553457
from line 3485
34833485
34843486 - `versions: array of object`
34853487
3486 Up to roughly 20 most-recently-published versions of this Artifact (older versions are not retained). Metadata only — use `GET /v1/compliance/apps/code/artifacts/{artifact_id}/versions/{version_id}` to download a version's content. Empty for a Claude Doc, because a Doc's versions hold only the web page that displays it, not its text. Download a Doc's text with `GET /v1/compliance/apps/code/artifacts/{artifact_id}/content?multi_file_format=zip`.
3488 Up to roughly 20 most-recently-published versions of this Artifact (older versions are not retained). Metadata only — use `GET /v1/compliance/apps/code/artifacts/{artifact_id}/versions/{version_id}` to download a version's content. Empty for a Claude Doc or a Claude Sheet, because their versions hold only the web page that displays them, not the Doc's text or the Sheet's data. Download a Doc's text or a Sheet's data with `GET /v1/compliance/apps/code/artifacts/{artifact_id}/content?multi_file_format=zip`.
34873489
34883490 - `id: string`
34893491
from line 3513
35113513
35123514```bash
35133515curl https://api.anthropic.com/v1/compliance/apps/code/artifacts \
3514 -H 'anthropic-version: 2023-06-01' \
3516 -H 'anthropic-version: 2026-10-21' \
35153517 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
35163518```
35173519
from line 3552
35503552
35513553**GET** `/v1/compliance/apps/code/artifacts/{artifact_id}/versions/{version_id}`
35523554
3553Streams the content of one version of an Artifact as the response body, in chunks with no `Content-Length`. With `multi_file_format=zip`, a version with more than one file, other than a Claude Doc's (an Artifact whose `artifact_type` is `claude_docs`), comes back as a ZIP archive of its files.
3555Streams the content of one version of an Artifact as the response body, in chunks with no `Content-Length`. With `multi_file_format=zip`, a version with more than one file, other than a Claude Doc's or a Claude Sheet's (an Artifact whose `artifact_type` is `claude_docs` or `claude_sheets`), comes back as a ZIP archive of its files.
35543556
35553557Headers: `Content-Type` (`application/zip` for a ZIP archive, otherwise the stored MIME type), `Content-Disposition: attachment; filename*=utf-8''<name>` (RFC 5987), `Cache-Control: no-store, private` and, when available, `Content-MD5` (base64). `<name>` is `{artifact_id}-{version_id}.zip` for a ZIP archive, otherwise `{artifact_id}-{version_id}.txt`, whatever the content is. `Content-MD5` is only on a single file stored uncompressed, so it is never on a ZIP archive. Check the body against it when present.
35563558
from line 3560
35583560
35593561- `404`: the Artifact doesn't exist or the API key cannot read it, or the version is no longer kept. Older versions are dropped as new ones are published, so list the Artifact's versions again. You also get this when the organization's customer-managed encryption key can't be used. Retry once the key can be used again.
35603562- `503`: the content can't be served right now, for example because its upload is still in progress or was abandoned. Retry with backoff.
3561- `422` with `error.details.error_code` `multi_file_unavailable`: the version has more than one file. Add `multi_file_format=zip`. A Claude Doc's version with more than one file gets this error either way, because it holds only the web page that displays the Doc. Download the Doc's text with `GET /v1/compliance/apps/code/artifacts/{artifact_id}/content?multi_file_format=zip`.
3563- `422` with `error.details.error_code` `multi_file_unavailable`: the version has more than one file. Add `multi_file_format=zip`. A Claude Doc's or a Claude Sheet's version with more than one file gets this error either way, because it holds only the web page that displays the Doc or the Sheet. Download the Doc's text or the Sheet's data with `GET /v1/compliance/apps/code/artifacts/{artifact_id}/content?multi_file_format=zip`.
35623564- `422` with `file_set_unavailable`: the request sets `multi_file_format=zip`, the version has more than one file, and one of them has a path the API cannot export. Do not retry.
35633565
35643566An error after the download has started stops the body early, with the `200` status already sent. For example, this happens to content stored compressed when it is over 16 MiB once decompressed. For a single file with no `Content-MD5`, the only sign is a transfer that aborts before its last chunk, so treat a download that aborts this way as failed.
from line 3592
35903592
35913593- `multi_file_format: optional string`
35923594
3593 Set to `zip`, its only value, to receive a version that has more than one file as a ZIP archive of its files with an index, `files.json`. A version with one file is returned as that file. A Claude Doc's version never comes back as a ZIP.
3595 Set to `zip`, its only value, to receive a version that has more than one file as a ZIP archive of its files with an index, `files.json`. A version with one file is returned as that file. A Claude Doc's or a Claude Sheet's version never comes back as a ZIP.
35943596
35953597#### Headers
35963598
3597- `"anthropic-version": optional string`
3599- `"anthropic-version": string`
35983600
3599 The version of the Claude API you want to use.
3601 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
36003602
3601 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
3603 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
36023604
36033605- `"x-api-key": optional string`
36043606
from line 3608
36063608
36073609```bash
36083610curl https://api.anthropic.com/v1/compliance/apps/code/artifacts/$ARTIFACT_ID/versions/$VERSION_ID \
3609 -H 'anthropic-version: 2023-06-01' \
3611 -H 'anthropic-version: 2026-10-21' \
36103612 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
36113613```
36123614
from line 3620
36183620
36193621For a Claude Doc (an Artifact whose `artifact_type` is `claude_docs`), the request must set `multi_file_format=zip`, and the content is the Doc's text as a ZIP archive of Word files. That ZIP has no `files.json`. It holds one Word file per text tab, in tab order, named `NN-<tab id>.docx`, where `NN` counts from `01`. A Doc with one text tab gets one file, `01-document.docx`.
36203622
3623For a Claude Sheet (an Artifact whose `artifact_type` is `claude_sheets`), the request must set `multi_file_format=zip` too, and the content is the Sheet's data as a ZIP archive that holds one .xlsx file, named `01-<id>.xlsx`, where `<id>` is an opaque identifier. The file has all of the Sheet's tabs and, when the Sheet has dropdowns, one more hidden tab that lists their options. It leaves out charts, conditional formatting and comments, among other things. A formula that an .xlsx file cannot hold is written as its result. If the Sheet's formulas cannot be calculated in time, every formula comes with `#VALUE!` in place of its result. That ZIP has no `files.json` either.
3624
36213625For any other Artifact, the content is the latest version. With `multi_file_format=zip`, when the latest version has more than one file, it comes back as a ZIP archive of its files.
36223626
3623Headers: `Content-Type` (`application/zip` for a ZIP archive, otherwise the stored MIME type), `Content-Disposition: attachment; filename*=utf-8''<name>` (RFC 5987), `Cache-Control: no-store, private` and, when available, `Content-MD5` (base64). `<name>` is `{artifact_id}.zip` for a ZIP archive, otherwise `{artifact_id}.txt`, whatever the content is. `Content-MD5` is always on a Claude Doc's ZIP, only on a single file stored uncompressed, and never on a ZIP archive of a version's files. Check the body against it when present.
3627Headers: `Content-Type` (`application/zip` for a ZIP archive, otherwise the stored MIME type), `Content-Disposition: attachment; filename*=utf-8''<name>` (RFC 5987), `Cache-Control: no-store, private` and, when available, `Content-MD5` (base64). `<name>` is `{artifact_id}.zip` for a ZIP archive, otherwise `{artifact_id}.txt`, whatever the content is. `Content-MD5` is always on a Claude Doc's or a Claude Sheet's ZIP, only on a single file stored uncompressed, and never on a ZIP archive of a version's files. Check the body against it when present.
36243628
36253629Errors specific to this download:
36263630
from line 3631
36273631- `404`: the Artifact doesn't exist, or the API key cannot read it.
36283632- `400` with `error.details.error_code` `cmek_key_disabled`: the organization's customer-managed encryption key can't be used, for example because it is disabled or revoked. Retry once the key can be used again.
36293633- `503`: the content can't be produced right now. Wait the number of seconds in the `Retry-After` header, or retry with backoff when it is absent.
3630- `422` with `multi_file_unavailable`: the request does not set `multi_file_format=zip`, and the Artifact is a Claude Doc or its latest version has more than one file. Add the parameter.
3631- `422` with `file_set_unavailable`: the request sets `multi_file_format=zip`, the Artifact is not a Claude Doc, and one of the latest version's files has a path the API cannot export. Do not retry.
3632- `422` with `document_unavailable`: Claude Docs has no document it can export for the Doc, for example because it has no text tab or is in the trash. Retrying will not help unless that changes.
3633- `422` with `document_too_large`: the Doc's export is over 16 MiB. Do not retry.
3634- `422` with `multi_file_unavailable`: the request does not set `multi_file_format=zip`, and the Artifact is a Claude Doc or a Claude Sheet, or its latest version has more than one file. Add the parameter.
3635- `422` with `file_set_unavailable`: the request sets `multi_file_format=zip`, the Artifact is neither a Claude Doc nor a Claude Sheet, and one of the latest version's files has a path the API cannot export. Do not retry.
3636- `422` with `document_unavailable`: Claude Docs has no document it can export for the Doc, for example because it has no text tab or is in the trash. For a Claude Sheet, nothing can be returned as an .xlsx file, for example because the Sheet is in the trash. Retrying will not help unless that changes.
3637- `422` with `document_too_large`: the Doc's or the Sheet's export is over 16 MiB. Do not retry.
36343638- `422` with `cmek_org_unsupported`: the organization uses customer-managed encryption keys and the export can't be served for it right now. Try again later, and contact Anthropic support if this continues.
36353639
36363640An error after the download has started stops the body early, with the `200` status already sent. For example, this happens to content stored compressed when it is over 16 MiB once decompressed. For a single file with no `Content-MD5`, the only sign is a transfer that aborts before its last chunk, so treat a download that aborts this way as failed.
from line 3662
36583662
36593663- `multi_file_format: optional string`
36603664
3661 Set to `zip`, its only value. A Claude Doc downloads only this way, as a ZIP archive of Word files, one per text tab. For any other Artifact, when the latest version has more than one file, it comes back as a ZIP archive of its files with an index, `files.json`.
3665 Set to `zip`, its only value. A Claude Doc downloads only this way, as a ZIP archive of Word files, one per text tab. So does a Claude Sheet, as a ZIP archive that holds one .xlsx file. For any other Artifact, when the latest version has more than one file, it comes back as a ZIP archive of its files with an index, `files.json`.
36623666
36633667#### Headers
36643668
3665- `"anthropic-version": optional string`
3669- `"anthropic-version": string`
36663670
3667 The version of the Claude API you want to use.
3671 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
36683672
3669 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
3673 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
36703674
36713675- `"x-api-key": optional string`
36723676
from line 3678
36743678
36753679```bash
36763680curl https://api.anthropic.com/v1/compliance/apps/code/artifacts/$ARTIFACT_ID/content \
3677 -H 'anthropic-version: 2023-06-01' \
3681 -H 'anthropic-version: 2026-10-21' \
36783682 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
36793683```
36803684
from line 3703
36993703
37003704#### Headers
37013705
3702- `"anthropic-version": optional string`
3706- `"anthropic-version": string`
37033707
3704 The version of the Claude API you want to use.
3708 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
37053709
3706 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
3710 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
37073711
37083712- `"x-api-key": optional string`
37093713
from line 3728
37243728```bash
37253729curl https://api.anthropic.com/v1/compliance/apps/code/artifacts/$ARTIFACT_ID \
37263730 -X DELETE \
3727 -H 'anthropic-version: 2023-06-01' \
3731 -H 'anthropic-version: 2026-10-21' \
37283732 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
37293733```
37303734
api/compliance/apps/artifacts Changed · +8 / -8 lines
from line 24
2424
2525### Headers
2626
27- `"anthropic-version": optional string`
27- `"anthropic-version": string`
2828
29 The version of the Claude API you want to use.
29 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
3030
31 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
31 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
3232
3333- `"x-api-key": optional string`
3434
from line 72
7272
7373```bash
7474curl https://api.anthropic.com/v1/compliance/apps/artifacts/$ARTIFACT_VERSION_ID \
75 -H 'anthropic-version: 2023-06-01' \
75 -H 'anthropic-version: 2026-10-21' \
7676 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
7777```
7878
from line 107
107107
108108### Headers
109109
110- `"anthropic-version": optional string`
110- `"anthropic-version": string`
111111
112 The version of the Claude API you want to use.
112 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
113113
114 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
114 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
115115
116116- `"x-api-key": optional string`
117117
from line 119
119119
120120```bash
121121curl https://api.anthropic.com/v1/compliance/apps/artifacts/$ARTIFACT_VERSION_ID/content \
122 -H 'anthropic-version: 2023-06-01' \
122 -H 'anthropic-version: 2026-10-21' \
123123 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
124124```
125125
api/compliance/apps/artifacts/download Changed · +4 / -4 lines
from line 19
1919
2020## Headers
2121
22- `"anthropic-version": optional string`
22- `"anthropic-version": string`
2323
24 The version of the Claude API you want to use.
24 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
2525
26 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
26 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
2727
2828- `"x-api-key": optional string`
2929
from line 31
3131
3232```bash
3333curl https://api.anthropic.com/v1/compliance/apps/artifacts/$ARTIFACT_VERSION_ID/content \
34 -H 'anthropic-version: 2023-06-01' \
34 -H 'anthropic-version: 2026-10-21' \
3535 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
3636```
3737
api/compliance/apps/artifacts/retrieve Changed · +4 / -4 lines
from line 22
2222
2323## Headers
2424
25- `"anthropic-version": optional string`
25- `"anthropic-version": string`
2626
27 The version of the Claude API you want to use.
27 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
2828
29 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
29 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
3030
3131- `"x-api-key": optional string`
3232
from line 70
7070
7171```bash
7272curl https://api.anthropic.com/v1/compliance/apps/artifacts/$ARTIFACT_VERSION_ID \
73 -H 'anthropic-version: 2023-06-01' \
73 -H 'anthropic-version: 2026-10-21' \
7474 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
7575```
7676
api/compliance/apps/chats Changed · +48 / -48 lines
from line 121
121121
122122### Headers
123123
124- `"anthropic-version": optional string`
124- `"anthropic-version": string`
125125
126 The version of the Claude API you want to use.
126 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
127127
128 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
128 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
129129
130130- `"x-api-key": optional string`
131131
from line 211
211211
212212```bash
213213curl https://api.anthropic.com/v1/compliance/apps/chats \
214 -H 'anthropic-version: 2023-06-01' \
214 -H 'anthropic-version: 2026-10-21' \
215215 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
216216```
217217
from line 262
262262
263263### Headers
264264
265- `"anthropic-version": optional string`
265- `"anthropic-version": string`
266266
267 The version of the Claude API you want to use.
267 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
268268
269 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
269 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
270270
271271- `"x-api-key": optional string`
272272
from line 287
287287```bash
288288curl https://api.anthropic.com/v1/compliance/apps/chats/$CLAUDE_CHAT_ID \
289289 -X DELETE \
290 -H 'anthropic-version: 2023-06-01' \
290 -H 'anthropic-version: 2026-10-21' \
291291 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
292292```
293293
from line 490
490490
491491#### Headers
492492
493- `"anthropic-version": optional string`
493- `"anthropic-version": string`
494494
495 The version of the Claude API you want to use.
495 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
496496
497 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
497 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
498498
499499- `"x-api-key": optional string`
500500
from line 788
788788
789789```bash
790790curl https://api.anthropic.com/v1/compliance/apps/chats/$CLAUDE_CHAT_ID/messages \
791 -H 'anthropic-version: 2023-06-01' \
791 -H 'anthropic-version: 2026-10-21' \
792792 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
793793```
794794
from line 933
933933
934934#### Headers
935935
936- `"anthropic-version": optional string`
936- `"anthropic-version": string`
937937
938 The version of the Claude API you want to use.
938 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
939939
940 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
940 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
941941
942942- `"x-api-key": optional string`
943943
from line 981
981981
982982```bash
983983curl https://api.anthropic.com/v1/compliance/apps/chats/files/$CLAUDE_FILE_ID \
984 -H 'anthropic-version: 2023-06-01' \
984 -H 'anthropic-version: 2026-10-21' \
985985 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
986986```
987987
from line 1019
10191019
10201020#### Headers
10211021
1022- `"anthropic-version": optional string`
1022- `"anthropic-version": string`
10231023
1024 The version of the Claude API you want to use.
1024 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
10251025
1026 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
1026 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
10271027
10281028- `"x-api-key": optional string`
10291029
from line 1044
10441044```bash
10451045curl https://api.anthropic.com/v1/compliance/apps/chats/files/$CLAUDE_FILE_ID \
10461046 -X DELETE \
1047 -H 'anthropic-version: 2023-06-01' \
1047 -H 'anthropic-version: 2026-10-21' \
10481048 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
10491049```
10501050
from line 1071
10711071
10721072#### Headers
10731073
1074- `"anthropic-version": optional string`
1074- `"anthropic-version": string`
10751075
1076 The version of the Claude API you want to use.
1076 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
10771077
1078 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
1078 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
10791079
10801080- `"x-api-key": optional string`
10811081
from line 1083
10831083
10841084```bash
10851085curl https://api.anthropic.com/v1/compliance/apps/chats/files/$CLAUDE_FILE_ID/content \
1086 -H 'anthropic-version: 2023-06-01' \
1086 -H 'anthropic-version: 2026-10-21' \
10871087 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
10881088```
10891089
from line 1105
11051105
11061106#### Headers
11071107
1108- `"anthropic-version": optional string`
1108- `"anthropic-version": string`
11091109
1110 The version of the Claude API you want to use.
1110 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
11111111
1112 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
1112 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
11131113
11141114- `"x-api-key": optional string`
11151115
from line 1149
11491149
11501150```bash
11511151curl https://api.anthropic.com/v1/compliance/apps/chats/generated-files/$CLAUDE_GEN_FILE_ID \
1152 -H 'anthropic-version: 2023-06-01' \
1152 -H 'anthropic-version: 2026-10-21' \
11531153 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
11541154```
11551155
from line 1181
11811181
11821182#### Headers
11831183
1184- `"anthropic-version": optional string`
1184- `"anthropic-version": string`
11851185
1186 The version of the Claude API you want to use.
1186 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
11871187
1188 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
1188 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
11891189
11901190- `"x-api-key": optional string`
11911191
from line 1193
11931193
11941194```bash
11951195curl https://api.anthropic.com/v1/compliance/apps/chats/generated-files/$CLAUDE_GEN_FILE_ID/content \
1196 -H 'anthropic-version: 2023-06-01' \
1196 -H 'anthropic-version: 2026-10-21' \
11971197 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
11981198```
11991199
api/compliance/apps/chats/delete Changed · +4 / -4 lines
from line 23
2323
2424## Headers
2525
26- `"anthropic-version": optional string`
26- `"anthropic-version": string`
2727
28 The version of the Claude API you want to use.
28 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
2929
30 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
30 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
3131
3232- `"x-api-key": optional string`
3333
from line 48
4848```bash
4949curl https://api.anthropic.com/v1/compliance/apps/chats/$CLAUDE_CHAT_ID \
5050 -X DELETE \
51 -H 'anthropic-version: 2023-06-01' \
51 -H 'anthropic-version: 2026-10-21' \
5252 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
5353```
5454
api/compliance/apps/chats/files Changed · +18 / -18 lines
from line 21
2121
2222### Headers
2323
24- `"anthropic-version": optional string`
24- `"anthropic-version": string`
2525
26 The version of the Claude API you want to use.
26 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
2727
28 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
28 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
2929
3030- `"x-api-key": optional string`
3131
from line 69
6969
7070```bash
7171curl https://api.anthropic.com/v1/compliance/apps/chats/files/$CLAUDE_FILE_ID \
72 -H 'anthropic-version: 2023-06-01' \
72 -H 'anthropic-version: 2026-10-21' \
7373 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
7474```
7575
from line 107
107107
108108### Headers
109109
110- `"anthropic-version": optional string`
110- `"anthropic-version": string`
111111
112 The version of the Claude API you want to use.
112 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
113113
114 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
114 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
115115
116116- `"x-api-key": optional string`
117117
from line 132
132132```bash
133133curl https://api.anthropic.com/v1/compliance/apps/chats/files/$CLAUDE_FILE_ID \
134134 -X DELETE \
135 -H 'anthropic-version: 2023-06-01' \
135 -H 'anthropic-version: 2026-10-21' \
136136 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
137137```
138138
from line 159
159159
160160### Headers
161161
162- `"anthropic-version": optional string`
162- `"anthropic-version": string`
163163
164 The version of the Claude API you want to use.
164 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
165165
166 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
166 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
167167
168168- `"x-api-key": optional string`
169169
from line 171
171171
172172```bash
173173curl https://api.anthropic.com/v1/compliance/apps/chats/files/$CLAUDE_FILE_ID/content \
174 -H 'anthropic-version: 2023-06-01' \
174 -H 'anthropic-version: 2026-10-21' \
175175 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
176176```
177177
api/compliance/apps/chats/files/delete Changed · +4 / -4 lines
from line 18
1818
1919## Headers
2020
21- `"anthropic-version": optional string`
21- `"anthropic-version": string`
2222
23 The version of the Claude API you want to use.
23 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
2424
25 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
25 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
2626
2727- `"x-api-key": optional string`
2828
from line 43
4343```bash
4444curl https://api.anthropic.com/v1/compliance/apps/chats/files/$CLAUDE_FILE_ID \
4545 -X DELETE \
46 -H 'anthropic-version: 2023-06-01' \
46 -H 'anthropic-version: 2026-10-21' \
4747 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
4848```
4949
api/compliance/apps/chats/files/download Changed · +4 / -4 lines
from line 17
1717
1818## Headers
1919
20- `"anthropic-version": optional string`
20- `"anthropic-version": string`
2121
22 The version of the Claude API you want to use.
22 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
2323
24 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
24 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
2525
2626- `"x-api-key": optional string`
2727
from line 29
2929
3030```bash
3131curl https://api.anthropic.com/v1/compliance/apps/chats/files/$CLAUDE_FILE_ID/content \
32 -H 'anthropic-version: 2023-06-01' \
32 -H 'anthropic-version: 2026-10-21' \
3333 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
3434```
3535
api/compliance/apps/chats/files/retrieve Changed · +4 / -4 lines
from line 19
1919
2020## Headers
2121
22- `"anthropic-version": optional string`
22- `"anthropic-version": string`
2323
24 The version of the Claude API you want to use.
24 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
2525
26 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
26 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
2727
2828- `"x-api-key": optional string`
2929
from line 67
6767
6868```bash
6969curl https://api.anthropic.com/v1/compliance/apps/chats/files/$CLAUDE_FILE_ID \
70 -H 'anthropic-version: 2023-06-01' \
70 -H 'anthropic-version: 2026-10-21' \
7171 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
7272```
7373
api/compliance/apps/chats/generated_files Changed · +8 / -8 lines
from line 21
2121
2222### Headers
2323
24- `"anthropic-version": optional string`
24- `"anthropic-version": string`
2525
26 The version of the Claude API you want to use.
26 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
2727
28 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
28 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
2929
3030- `"x-api-key": optional string`
3131
from line 65
6565
6666```bash
6767curl https://api.anthropic.com/v1/compliance/apps/chats/generated-files/$CLAUDE_GEN_FILE_ID \
68 -H 'anthropic-version: 2023-06-01' \
68 -H 'anthropic-version: 2026-10-21' \
6969 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
7070```
7171
from line 97
9797
9898### Headers
9999
100- `"anthropic-version": optional string`
100- `"anthropic-version": string`
101101
102 The version of the Claude API you want to use.
102 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
103103
104 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
104 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
105105
106106- `"x-api-key": optional string`
107107
from line 109
109109
110110```bash
111111curl https://api.anthropic.com/v1/compliance/apps/chats/generated-files/$CLAUDE_GEN_FILE_ID/content \
112 -H 'anthropic-version: 2023-06-01' \
112 -H 'anthropic-version: 2026-10-21' \
113113 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
114114```
115115
api/compliance/apps/chats/generated_files/download Changed · +4 / -4 lines
from line 17
1717
1818## Headers
1919
20- `"anthropic-version": optional string`
20- `"anthropic-version": string`
2121
22 The version of the Claude API you want to use.
22 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
2323
24 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
24 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
2525
2626- `"x-api-key": optional string`
2727
from line 29
2929
3030```bash
3131curl https://api.anthropic.com/v1/compliance/apps/chats/generated-files/$CLAUDE_GEN_FILE_ID/content \
32 -H 'anthropic-version: 2023-06-01' \
32 -H 'anthropic-version: 2026-10-21' \
3333 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
3434```
3535
api/compliance/apps/chats/generated_files/retrieve Changed · +4 / -4 lines
from line 19
1919
2020## Headers
2121
22- `"anthropic-version": optional string`
22- `"anthropic-version": string`
2323
24 The version of the Claude API you want to use.
24 The version of the Claude API to use. Send an available version on every request. Some other Compliance API endpoints return a 400 error for a request that omits this header, or sends a value that is not an available version. This endpoint treats such a request as version `2023-06-01`.
2525
26 Read more about versioning and our version history [here](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http).
26 See [API versions](https://platform.claude.com/docs/en/api/versioning?sdk_languages=http) for the available versions.
2727
2828- `"x-api-key": optional string`
2929
from line 63
6363
6464```bash
6565curl https://api.anthropic.com/v1/compliance/apps/chats/generated-files/$CLAUDE_GEN_FILE_ID \
66 -H 'anthropic-version: 2023-06-01' \
66 -H 'anthropic-version: 2026-10-21' \
6767 -H "Authorization: Bearer $ANTHROPIC_COMPLIANCE_API_KEY"
6868```
6969