What this read moved
1–25 of 45This capture is too large to show at once. Changes 1-25 of 45 are below, significant first; the rest are on the following screens.
api/beta Changed · +14 / -14 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.
Nothing in the body moved in this read. What changed is above.
api/beta/organization Changed · +14 / -14 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.
Nothing in the body moved in this read. What changed is above.
api/beta/organization/spend_limits Changed · +14 / -14 lines
from line 83
8383
8484 - `"weekly"`
8585
86 - `scope: object or object or object or 2 more`
86 - `scope: User or SeatTier or RBACGroup or 2 more`
8787
8888 What the limit applies to. A tagged union on `type`; each variant carries the identifier for its scope.
8989
from line 224
224224
225225 - `"weekly"`
226226
227 - `scope: object or object or object or 2 more`
227 - `scope: User or SeatTier or RBACGroup or 2 more`
228228
229229 What the limit applies to. A tagged union on `type`; each variant carries the identifier for its scope.
230230
from line 456
456456
457457 Tagged ID of the member the spend limit applies to.
458458
459 - `source: object or object or object or 2 more`
459 - `source: User or SeatTier or RBACGroup or 2 more`
460460
461461 - `User object`
462462
from line 588
588588
589589 - `"weekly"`
590590
591 - `scope: object or object or object or 2 more`
591 - `scope: User or SeatTier or RBACGroup or 2 more`
592592
593593 What the limit applies to. A tagged union on `type`; each variant carries the identifier for its scope.
594594
from line 714
714714
715715 Tagged ID of the member the spend limit applies to.
716716
717 - `source: object or object or object or 2 more`
717 - `source: User or SeatTier or RBACGroup or 2 more`
718718
719719 - `User object`
720720
from line 863
863863
864864 format: date-time
865865
866 - `resolved_by: object or object or null`
866 - `resolved_by: UserActor or ScopedAPIKeyActor or null`
867867
868868 - `UserActor object`
869869
from line 975
975975
976976 Tagged ID of the member the spend limit applies to.
977977
978 - `source: object or object or object or 2 more`
978 - `source: User or SeatTier or RBACGroup or 2 more`
979979
980980 - `User object`
981981
from line 1166
11661166
11671167 format: date-time
11681168
1169 - `resolved_by: object or object or null`
1169 - `resolved_by: UserActor or ScopedAPIKeyActor or null`
11701170
11711171 - `UserActor object`
11721172
from line 1278
12781278
12791279 Tagged ID of the member the spend limit applies to.
12801280
1281 - `source: object or object or object or 2 more`
1281 - `source: User or SeatTier or RBACGroup or 2 more`
12821282
12831283 - `User object`
12841284
from line 1478
14781478
14791479 format: date-time
14801480
1481- `resolved_by: object or object or null`
1481- `resolved_by: UserActor or ScopedAPIKeyActor or null`
14821482
14831483 - `UserActor object`
14841484
from line 1558
15581558
15591559 - `"weekly"`
15601560
1561 - `scope: object or object or object or 2 more`
1561 - `scope: User or SeatTier or RBACGroup or 2 more`
15621562
15631563 What the limit applies to. A tagged union on `type`; each variant carries the identifier for its scope.
15641564
from line 1682
16821682
16831683 Tagged ID of the member the spend limit applies to.
16841684
1685 - `source: object or object or object or 2 more`
1685 - `source: User or SeatTier or RBACGroup or 2 more`
16861686
16871687 - `User object`
16881688
from line 1888
18881888
18891889 format: date-time
18901890
1891 - `resolved_by: object or object or null`
1891 - `resolved_by: UserActor or ScopedAPIKeyActor or null`
18921892
18931893 - `UserActor object`
18941894
from line 2000
20002000
20012001 Tagged ID of the member the spend limit applies to.
20022002
2003 - `source: object or object or object or 2 more`
2003 - `source: User or SeatTier or RBACGroup or 2 more`
20042004
20052005 - `User object`
20062006
api/beta/organization/spend_limits/increase_requests Changed · +14 / -14 lines
from line 94
9494
9595 format: date-time
9696
97 - `resolved_by: object or object or null`
97 - `resolved_by: UserActor or ScopedAPIKeyActor or null`
9898
9999 - `UserActor object`
100100
from line 206
206206
207207 Tagged ID of the member the spend limit applies to.
208208
209 - `source: object or object or object or 2 more`
209 - `source: User or SeatTier or RBACGroup or 2 more`
210210
211211 - `User object`
212212
from line 397
397397
398398 format: date-time
399399
400 - `resolved_by: object or object or null`
400 - `resolved_by: UserActor or ScopedAPIKeyActor or null`
401401
402402 - `UserActor object`
403403
from line 509
509509
510510 Tagged ID of the member the spend limit applies to.
511511
512 - `source: object or object or object or 2 more`
512 - `source: User or SeatTier or RBACGroup or 2 more`
513513
514514 - `User object`
515515
from line 709
709709
710710 format: date-time
711711
712- `resolved_by: object or object or null`
712- `resolved_by: UserActor or ScopedAPIKeyActor or null`
713713
714714 - `UserActor object`
715715
from line 789
789789
790790 - `"weekly"`
791791
792 - `scope: object or object or object or 2 more`
792 - `scope: User or SeatTier or RBACGroup or 2 more`
793793
794794 What the limit applies to. A tagged union on `type`; each variant carries the identifier for its scope.
795795
from line 913
913913
914914 Tagged ID of the member the spend limit applies to.
915915
916 - `source: object or object or object or 2 more`
916 - `source: User or SeatTier or RBACGroup or 2 more`
917917
918918 - `User object`
919919
from line 1119
11191119
11201120 format: date-time
11211121
1122 - `resolved_by: object or object or null`
1122 - `resolved_by: UserActor or ScopedAPIKeyActor or null`
11231123
11241124 - `UserActor object`
11251125
from line 1231
12311231
12321232 Tagged ID of the member the spend limit applies to.
12331233
1234 - `source: object or object or object or 2 more`
1234 - `source: User or SeatTier or RBACGroup or 2 more`
12351235
12361236 - `User object`
12371237
from line 1404
14041404
14051405 format: date-time
14061406
1407 - `resolved_by: object or object or null`
1407 - `resolved_by: UserActor or ScopedAPIKeyActor or null`
14081408
14091409 - `UserActor object`
14101410
from line 1516
15161516
15171517 Tagged ID of the member the spend limit applies to.
15181518
1519 - `source: object or object or object or 2 more`
1519 - `source: User or SeatTier or RBACGroup or 2 more`
15201520
15211521 - `User object`
15221522
from line 1628
16281628
16291629 format: date-time
16301630
1631 - `resolved_by: object or object or null`
1631 - `resolved_by: UserActor or ScopedAPIKeyActor or null`
16321632
16331633 - `UserActor object`
16341634
from line 1708
17081708
17091709 - `"weekly"`
17101710
1711 - `scope: object or object or object or 2 more`
1711 - `scope: User or SeatTier or RBACGroup or 2 more`
17121712
17131713 What the limit applies to. A tagged union on `type`; each variant carries the identifier for its scope.
17141714
from line 1832
18321832
18331833 Tagged ID of the member the spend limit applies to.
18341834
1835 - `source: object or object or object or 2 more`
1835 - `source: User or SeatTier or RBACGroup or 2 more`
18361836
18371837 - `User object`
18381838
api/beta/organization/spend_limits/increase_requests/approve Changed · +3 / -3 lines
from line 90
9090
9191 format: date-time
9292
93- `resolved_by: object or object or null`
93- `resolved_by: UserActor or ScopedAPIKeyActor or null`
9494
9595 - `UserActor object`
9696
from line 170
170170
171171 - `"weekly"`
172172
173 - `scope: object or object or object or 2 more`
173 - `scope: User or SeatTier or RBACGroup or 2 more`
174174
175175 What the limit applies to. A tagged union on `type`; each variant carries the identifier for its scope.
176176
from line 294
294294
295295 Tagged ID of the member the spend limit applies to.
296296
297 - `source: object or object or object or 2 more`
297 - `source: User or SeatTier or RBACGroup or 2 more`
298298
299299 - `User object`
300300
api/compliance Changed · +1000 / -1000 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 4079
40794079
40804080#### Returns
40814081
4082- `data: optional array of object or object or object or 490 more`
4082- `data: optional array of AbuseDecisionReceived or AccountDeleted or AdminAPIKeyCreated or 490 more`
40834083
40844084 List of activity records. Each element's `type` field identifies which activity it is and which additional fields are present.
40854085
from line 4091
40914091
40924092 default: abuse_decision_received
40934093
4094 - `actor: object or object or object or 8 more`
4094 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
40954095
40964096 - `APIActor object`
40974097
from line 4225
42254225
42264226 default: federated_actor
42274227
4228 - `provider: object or object or object or object`
4228 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
42294229
42304230 - `FederatedActorAwsProvider object`
42314231
from line 4335
43354335
43364336 default: account_deleted
43374337
4338 - `actor: object or object or object or 8 more`
4338 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
43394339
43404340 - `APIActor object`
43414341
from line 4469
44694469
44704470 default: federated_actor
44714471
4472 - `provider: object or object or object or object`
4472 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
44734473
44744474 - `FederatedActorAwsProvider object`
44754475
from line 4567
45674567
45684568 default: admin_api_key_created
45694569
4570 - `actor: object or object or object or 8 more`
4570 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
45714571
45724572 - `APIActor object`
45734573
from line 4701
47014701
47024702 default: federated_actor
47034703
4704 - `provider: object or object or object or object`
4704 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
47054705
47064706 - `FederatedActorAwsProvider object`
47074707
from line 4807
48074807
48084808 default: admin_api_key_deleted
48094809
4810 - `actor: object or object or object or 8 more`
4810 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
48114811
48124812 - `APIActor object`
48134813
from line 4941
49414941
49424942 default: federated_actor
49434943
4944 - `provider: object or object or object or object`
4944 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
49454945
49464946 - `FederatedActorAwsProvider object`
49474947
from line 5043
50435043
50445044 default: admin_api_key_updated
50455045
5046 - `actor: object or object or object or 8 more`
5046 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
50475047
50485048 - `APIActor object`
50495049
from line 5177
51775177
51785178 default: federated_actor
51795179
5180 - `provider: object or object or object or object`
5180 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
51815181
51825182 - `FederatedActorAwsProvider object`
51835183
from line 5301
53015301
53025302 default: admin_connector_request_resolved
53035303
5304 - `actor: object or object or object or 8 more`
5304 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
53055305
53065306 - `APIActor object`
53075307
from line 5435
54355435
54365436 default: federated_actor
54375437
5438 - `provider: object or object or object or object`
5438 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
54395439
54405440 - `FederatedActorAwsProvider object`
54415441
from line 5545
55455545
55465546 default: admin_request_created
55475547
5548 - `actor: object or object or object or 8 more`
5548 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
55495549
55505550 - `APIActor object`
55515551
from line 5679
56795679
56805680 default: federated_actor
56815681
5682 - `provider: object or object or object or object`
5682 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
56835683
56845684 - `FederatedActorAwsProvider object`
56855685
from line 5779
57795779
57805780 default: admin_setup_checklist_step_delegated
57815781
5782 - `actor: object or object or object or 8 more`
5782 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
57835783
57845784 - `APIActor object`
57855785
from line 5913
59135913
59145914 default: federated_actor
59155915
5916 - `provider: object or object or object or object`
5916 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
59175917
59185918 - `FederatedActorAwsProvider object`
59195919
from line 6023
60236023
60246024 default: admin_setup_checklist_step_delegation_cancelled
60256025
6026 - `actor: object or object or object or 8 more`
6026 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
60276027
60286028 - `APIActor object`
60296029
from line 6157
61576157
61586158 default: federated_actor
61596159
6160 - `provider: object or object or object or object`
6160 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
61616161
61626162 - `FederatedActorAwsProvider object`
61636163
from line 6259
62596259
62606260 default: age_verified
62616261
6262 - `actor: object or object or object or 8 more`
6262 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
62636263
62646264 - `APIActor object`
62656265
from line 6393
63936393
63946394 default: federated_actor
63956395
6396 - `provider: object or object or object or object`
6396 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
63976397
63986398 - `FederatedActorAwsProvider object`
63996399
from line 6491
64916491
64926492 default: anonymous_mobile_login_attempted
64936493
6494 - `actor: object or object or object or 8 more`
6494 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
64956495
64966496 - `APIActor object`
64976497
from line 6625
66256625
66266626 default: federated_actor
66276627
6628 - `provider: object or object or object or object`
6628 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
66296629
66306630 - `FederatedActorAwsProvider object`
66316631
from line 6723
67236723
67246724 default: api_key_created
67256725
6726 - `actor: object or object or object or 8 more`
6726 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
67276727
67286728 - `APIActor object`
67296729
from line 6857
68576857
68586858 default: federated_actor
68596859
6860 - `provider: object or object or object or object`
6860 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
68616861
68626862 - `FederatedActorAwsProvider object`
68636863
from line 6967
69676967
69686968 default: claude_artifact_access_failed
69696969
6970 - `actor: object or object or object or 8 more`
6970 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
69716971
69726972 - `APIActor object`
69736973
from line 7101
71017101
71027102 default: federated_actor
71037103
7104 - `provider: object or object or object or object`
7104 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
71057105
71067106 - `FederatedActorAwsProvider object`
71077107
from line 7211
72117211
72127212 default: claude_artifact_commented
72137213
7214 - `actor: object or object or object or 8 more`
7214 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
72157215
72167216 - `APIActor object`
72177217
from line 7345
73457345
73467346 default: federated_actor
73477347
7348 - `provider: object or object or object or object`
7348 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
73497349
73507350 - `FederatedActorAwsProvider object`
73517351
from line 7487
74877487
74887488 default: claude_artifact_comments_viewed
74897489
7490 - `actor: object or object or object or 8 more`
7490 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
74917491
74927492 - `APIActor object`
74937493
from line 7621
76217621
76227622 default: federated_actor
76237623
7624 - `provider: object or object or object or object`
7624 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
76257625
76267626 - `FederatedActorAwsProvider object`
76277627
from line 7727
77277727
77287728 default: claude_artifact_created
77297729
7730 - `actor: object or object or object or 8 more`
7730 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
77317731
77327732 - `APIActor object`
77337733
from line 7861
78617861
78627862 default: federated_actor
78637863
7864 - `provider: object or object or object or object`
7864 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
78657865
78667866 - `FederatedActorAwsProvider object`
78677867
from line 7963
79637963
79647964 default: claude_published_artifact_deleted
79657965
7966 - `actor: object or object or object or 8 more`
7966 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
79677967
79687968 - `APIActor object`
79697969
from line 8097
80978097
80988098 default: federated_actor
80998099
8100 - `provider: object or object or object or object`
8100 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
81018101
81028102 - `FederatedActorAwsProvider object`
81038103
from line 8199
81998199
82008200 default: claude_artifact_published
82018201
8202 - `actor: object or object or object or 8 more`
8202 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
82038203
82048204 - `APIActor object`
82058205
from line 8333
83338333
83348334 default: federated_actor
83358335
8336 - `provider: object or object or object or object`
8336 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
83378337
83388338 - `FederatedActorAwsProvider object`
83398339
from line 8455
84558455
84568456 default: claude_artifact_sharing_updated
84578457
8458 - `actor: object or object or object or 8 more`
8458 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
84598459
84608460 - `APIActor object`
84618461
from line 8589
85898589
85908590 default: federated_actor
85918591
8592 - `provider: object or object or object or object`
8592 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
85938593
85948594 - `FederatedActorAwsProvider object`
85958595
from line 8661
86618661
86628662 - `user_agent: optional string or null`
86638663
8664 - `audience: array of object or object or object`
8664 - `audience: array of ArtifactSharingAudienceOrganization or ArtifactSharingAudienceUsers or ArtifactSharingAudienceAnyoneWithLink`
86658665
86668666 The artifact's sharing audience after the change. If empty, the artifact is visible only to its owner.
86678667
from line 8756
87568756
87578757 default: claude_artifact_viewed
87588758
8759 - `actor: object or object or object or 8 more`
8759 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
87608760
87618761 - `APIActor object`
87628762
from line 8890
88908890
88918891 default: federated_actor
88928892
8893 - `provider: object or object or object or object`
8893 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
88948894
88958895 - `FederatedActorAwsProvider object`
88968896
from line 8996
89968996
89978997 default: audit_log_export_accessed
89988998
8999 - `actor: object or object or object or 8 more`
8999 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
90009000
90019001 - `APIActor object`
90029002
from line 9130
91309130
91319131 default: federated_actor
91329132
9133 - `provider: object or object or object or object`
9133 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
91349134
91359135 - `FederatedActorAwsProvider object`
91369136
from line 9228
92289228
92299229 default: audit_log_export_started
92309230
9231 - `actor: object or object or object or 8 more`
9231 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
92329232
92339233 - `APIActor object`
92349234
from line 9362
93629362
93639363 default: federated_actor
93649364
9365 - `provider: object or object or object or object`
9365 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
93669366
93679367 - `FederatedActorAwsProvider object`
93689368
from line 9468
94689468
94699469 default: billing_emails_updated
94709470
9471 - `actor: object or object or object or 8 more`
9471 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
94729472
94739473 - `APIActor object`
94749474
from line 9602
96029602
96039603 default: federated_actor
96049604
9605 - `provider: object or object or object or object`
9605 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
96069606
96079607 - `FederatedActorAwsProvider object`
96089608
from line 9712
97129712
97139713 default: ccr_agent_created
97149714
9715 - `actor: object or object or object or 8 more`
9715 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
97169716
97179717 - `APIActor object`
97189718
from line 9846
98469846
98479847 default: federated_actor
98489848
9849 - `provider: object or object or object or object`
9849 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
98509850
98519851 - `FederatedActorAwsProvider object`
98529852
from line 9976
99769976
99779977 default: ccr_agent_deleted
99789978
9979 - `actor: object or object or object or 8 more`
9979 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
99809980
99819981 - `APIActor object`
99829982
9983 - `type: optional "api_actor"`
9984
9985 default: api_actor
9986
9987 - `api_key_id: string`
9988
9989 - `ip_address: string`
9990
9991 - `user_agent: string`
9992
9993 - `UserActor object`
9994
9995 - `type: optional "user_actor"`
9996
9997 default: user_actor
9998
9999 - `email_address: string`
10000
10001 format: email
10002
10003 - `ip_address: string`
10004
10005 - `user_agent: string`
10006
10007 - `user_id: string`
10008
10009 - `UnauthenticatedUserActor object`
10010
10011 - `type: optional "unauthenticated_user_actor"`
10012
10013 default: unauthenticated_user_actor
10014
10015 - `ip_address: string`
10016
10017 - `user_agent: string`
10018
10019 - `unauthenticated_email_address: optional string or null`
10020
10021 format: email
10022
10023 - `AnthropicActor object`
10024
10025 - `type: optional "anthropic_actor"`
10026
10027 default: anthropic_actor
10028
10029 - `email_address: optional string or null`
10030
10031 format: email
10032
10033 - `SystemActor object`
10034
10035 Automated background processing performed by Anthropic systems, acting
10036 without a user or customer credential.
10037
10038 - `type: optional "system_actor"`
10039
10040 default: system_actor
10041
10042 - `service: optional string or null`
10043
10044 Name of the automated process that performed the action, when known.
10045
10046 - `AdminAPIKeyActor object`
10047
10048 - `type: optional "admin_api_key_actor"`
10049
10050 default: admin_api_key_actor
10051
10052 - `admin_api_key_id: string`
10053
10054 - `ip_address: string`
10055
10056 - `user_agent: string`
10057
10058 - `ServiceAccountActor object`
10059
10060 - `type: optional "service_account_actor"`
10061
10062 default: service_account_actor
10063
10064 - `ip_address: string`
10065
10066 - `service_account_id: string`
10067
10068 - `user_agent: string`
10069
10070 - `ScimDirectorySyncActor object`
10071
10072 - `type: optional "scim_directory_sync_actor"`
10073
10074 default: scim_directory_sync_actor
10075
10076 - `directory_id: string`
10077
10078 - `workos_event_id: string`
10079
10080 - `idp_connection_type: optional string or null`
10081
10082 - `FederatedIdentityActor object`
10083
10084 A federated external workload authenticated via a verified OIDC token.
10085
10086 Carries the verified issuer, subject, and audience claims from the
10087 presented JWT.
10088
10089 - `type: optional "federated_identity_actor"`
10090
10091 default: federated_identity_actor
10092
10093 - `issuer: string`
10094
10095 - `subject: string`
10096
10097 - `audience: optional array of string`
10098
10099 - `ip_address: optional string or null`
10100
10101 - `user_agent: optional string or null`
10102
10103 - `FederatedActor object`
10104
10105 An external identi
9983
api/compliance/activities Changed · +1984 / -1984 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 4077
40774077
40784078### Returns
40794079
4080- `data: optional array of object or object or object or 490 more`
4080- `data: optional array of AbuseDecisionReceived or AccountDeleted or AdminAPIKeyCreated or 490 more`
40814081
40824082 List of activity records. Each element's `type` field identifies which activity it is and which additional fields are present.
40834083
from line 4089
40894089
40904090 default: abuse_decision_received
40914091
4092 - `actor: object or object or object or 8 more`
4092 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
40934093
40944094 - `APIActor object`
40954095
from line 4223
42234223
42244224 default: federated_actor
42254225
4226 - `provider: object or object or object or object`
4226 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
42274227
42284228 - `FederatedActorAwsProvider object`
42294229
from line 4333
43334333
43344334 default: account_deleted
43354335
4336 - `actor: object or object or object or 8 more`
4336 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
43374337
43384338 - `APIActor object`
43394339
from line 4467
44674467
44684468 default: federated_actor
44694469
4470 - `provider: object or object or object or object`
4470 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
44714471
44724472 - `FederatedActorAwsProvider object`
44734473
from line 4565
45654565
45664566 default: admin_api_key_created
45674567
4568 - `actor: object or object or object or 8 more`
4568 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
45694569
45704570 - `APIActor object`
45714571
from line 4699
46994699
47004700 default: federated_actor
47014701
4702 - `provider: object or object or object or object`
4702 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
47034703
47044704 - `FederatedActorAwsProvider object`
47054705
from line 4805
48054805
48064806 default: admin_api_key_deleted
48074807
4808 - `actor: object or object or object or 8 more`
4808 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
48094809
48104810 - `APIActor object`
48114811
from line 4939
49394939
49404940 default: federated_actor
49414941
4942 - `provider: object or object or object or object`
4942 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
49434943
49444944 - `FederatedActorAwsProvider object`
49454945
from line 5041
50415041
50425042 default: admin_api_key_updated
50435043
5044 - `actor: object or object or object or 8 more`
5044 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
50455045
50465046 - `APIActor object`
50475047
from line 5175
51755175
51765176 default: federated_actor
51775177
5178 - `provider: object or object or object or object`
5178 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
51795179
51805180 - `FederatedActorAwsProvider object`
51815181
from line 5299
52995299
53005300 default: admin_connector_request_resolved
53015301
5302 - `actor: object or object or object or 8 more`
5302 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
53035303
53045304 - `APIActor object`
53055305
from line 5433
54335433
54345434 default: federated_actor
54355435
5436 - `provider: object or object or object or object`
5436 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
54375437
54385438 - `FederatedActorAwsProvider object`
54395439
from line 5543
55435543
55445544 default: admin_request_created
55455545
5546 - `actor: object or object or object or 8 more`
5546 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
55475547
55485548 - `APIActor object`
55495549
from line 5677
56775677
56785678 default: federated_actor
56795679
5680 - `provider: object or object or object or object`
5680 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
56815681
56825682 - `FederatedActorAwsProvider object`
56835683
from line 5777
57775777
57785778 default: admin_setup_checklist_step_delegated
57795779
5780 - `actor: object or object or object or 8 more`
5780 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
57815781
57825782 - `APIActor object`
57835783
from line 5911
59115911
59125912 default: federated_actor
59135913
5914 - `provider: object or object or object or object`
5914 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
59155915
59165916 - `FederatedActorAwsProvider object`
59175917
from line 6021
60216021
60226022 default: admin_setup_checklist_step_delegation_cancelled
60236023
6024 - `actor: object or object or object or 8 more`
6024 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
60256025
60266026 - `APIActor object`
60276027
from line 6155
61556155
61566156 default: federated_actor
61576157
6158 - `provider: object or object or object or object`
6158 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
61596159
61606160 - `FederatedActorAwsProvider object`
61616161
from line 6257
62576257
62586258 default: age_verified
62596259
6260 - `actor: object or object or object or 8 more`
6260 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
62616261
62626262 - `APIActor object`
62636263
from line 6391
63916391
63926392 default: federated_actor
63936393
6394 - `provider: object or object or object or object`
6394 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
63956395
63966396 - `FederatedActorAwsProvider object`
63976397
from line 6489
64896489
64906490 default: anonymous_mobile_login_attempted
64916491
6492 - `actor: object or object or object or 8 more`
6492 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
64936493
64946494 - `APIActor object`
64956495
from line 6623
66236623
66246624 default: federated_actor
66256625
6626 - `provider: object or object or object or object`
6626 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
66276627
66286628 - `FederatedActorAwsProvider object`
66296629
from line 6721
67216721
67226722 default: api_key_created
67236723
6724 - `actor: object or object or object or 8 more`
6724 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
67256725
67266726 - `APIActor object`
67276727
from line 6855
68556855
68566856 default: federated_actor
68576857
6858 - `provider: object or object or object or object`
6858 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
68596859
68606860 - `FederatedActorAwsProvider object`
68616861
from line 6965
69656965
69666966 default: claude_artifact_access_failed
69676967
6968 - `actor: object or object or object or 8 more`
6968 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
69696969
69706970 - `APIActor object`
69716971
from line 7099
70997099
71007100 default: federated_actor
71017101
7102 - `provider: object or object or object or object`
7102 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
71037103
71047104 - `FederatedActorAwsProvider object`
71057105
from line 7209
72097209
72107210 default: claude_artifact_commented
72117211
7212 - `actor: object or object or object or 8 more`
7212 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
72137213
72147214 - `APIActor object`
72157215
from line 7343
73437343
73447344 default: federated_actor
73457345
7346 - `provider: object or object or object or object`
7346 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
73477347
73487348 - `FederatedActorAwsProvider object`
73497349
from line 7485
74857485
74867486 default: claude_artifact_comments_viewed
74877487
7488 - `actor: object or object or object or 8 more`
7488 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
74897489
74907490 - `APIActor object`
74917491
from line 7619
76197619
76207620 default: federated_actor
76217621
7622 - `provider: object or object or object or object`
7622 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
76237623
76247624 - `FederatedActorAwsProvider object`
76257625
from line 7725
77257725
77267726 default: claude_artifact_created
77277727
7728 - `actor: object or object or object or 8 more`
7728 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
77297729
77307730 - `APIActor object`
77317731
from line 7859
78597859
78607860 default: federated_actor
78617861
7862 - `provider: object or object or object or object`
7862 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
78637863
78647864 - `FederatedActorAwsProvider object`
78657865
from line 7961
79617961
79627962 default: claude_published_artifact_deleted
79637963
7964 - `actor: object or object or object or 8 more`
7964 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
79657965
79667966 - `APIActor object`
79677967
from line 8095
80958095
80968096 default: federated_actor
80978097
8098 - `provider: object or object or object or object`
8098 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
80998099
81008100 - `FederatedActorAwsProvider object`
81018101
from line 8197
81978197
81988198 default: claude_artifact_published
81998199
8200 - `actor: object or object or object or 8 more`
8200 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
82018201
82028202 - `APIActor object`
82038203
from line 8331
83318331
83328332 default: federated_actor
83338333
8334 - `provider: object or object or object or object`
8334 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
83358335
83368336 - `FederatedActorAwsProvider object`
83378337
from line 8453
84538453
84548454 default: claude_artifact_sharing_updated
84558455
8456 - `actor: object or object or object or 8 more`
8456 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
84578457
84588458 - `APIActor object`
84598459
from line 8587
85878587
85888588 default: federated_actor
85898589
8590 - `provider: object or object or object or object`
8590 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
85918591
85928592 - `FederatedActorAwsProvider object`
85938593
from line 8659
86598659
86608660 - `user_agent: optional string or null`
86618661
8662 - `audience: array of object or object or object`
8662 - `audience: array of ArtifactSharingAudienceOrganization or ArtifactSharingAudienceUsers or ArtifactSharingAudienceAnyoneWithLink`
86638663
86648664 The artifact's sharing audience after the change. If empty, the artifact is visible only to its owner.
86658665
from line 8754
87548754
87558755 default: claude_artifact_viewed
87568756
8757 - `actor: object or object or object or 8 more`
8757 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
87588758
87598759 - `APIActor object`
87608760
from line 8888
88888888
88898889 default: federated_actor
88908890
8891 - `provider: object or object or object or object`
8891 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
88928892
88938893 - `FederatedActorAwsProvider object`
88948894
from line 8994
89948994
89958995 default: audit_log_export_accessed
89968996
8997 - `actor: object or object or object or 8 more`
8997 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
89988998
89998999 - `APIActor object`
90009000
from line 9128
91289128
91299129 default: federated_actor
91309130
9131 - `provider: object or object or object or object`
9131 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
91329132
91339133 - `FederatedActorAwsProvider object`
91349134
from line 9226
92269226
92279227 default: audit_log_export_started
92289228
9229 - `actor: object or object or object or 8 more`
9229 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
92309230
92319231 - `APIActor object`
92329232
from line 9360
93609360
93619361 default: federated_actor
93629362
9363 - `provider: object or object or object or object`
9363 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
93649364
93659365 - `FederatedActorAwsProvider object`
93669366
from line 9466
94669466
94679467 default: billing_emails_updated
94689468
9469 - `actor: object or object or object or 8 more`
9469 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
94709470
94719471 - `APIActor object`
94729472
from line 9600
96009600
96019601 default: federated_actor
96029602
9603 - `provider: object or object or object or object`
9603 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
96049604
96059605 - `FederatedActorAwsProvider object`
96069606
from line 9710
97109710
97119711 default: ccr_agent_created
97129712
9713 - `actor: object or object or object or 8 more`
9713 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
97149714
97159715 - `APIActor object`
97169716
from line 9844
98449844
98459845 default: federated_actor
98469846
9847 - `provider: object or object or object or object`
9847 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
98489848
98499849 - `FederatedActorAwsProvider object`
98509850
from line 9974
99749974
99759975 default: ccr_agent_deleted
99769976
9977 - `actor: object or object or object or 8 more`
9977 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
99789978
99799979 - `APIActor object`
99809980
99819981 - `type: optional "api_actor"`
9982
9983 default: api_actor
9984
9985 - `api_key_id: string`
9986
9987 - `ip_address: string`
9988
9989 - `user_agent: string`
9990
9991 - `UserActor object`
9992
9993 - `type: optional "user_actor"`
9994
9995 default: user_actor
9996
9997 - `email_address: string`
9998
9999 format: email
10000
10001 - `ip_address: string`
10002
10003 - `user_agent: string`
10004
10005 - `user_id: string`
10006
10007 - `UnauthenticatedUserActor object`
10008
10009 - `type: optional "unauthenticated_user_actor"`
10010
10011 default: unauthenticated_user_actor
10012
10013 - `ip_address: string`
10014
10015 - `user_agent: string`
10016
10017 - `unauthenticated_email_address: optional string or null`
10018
10019 format: email
10020
10021 - `AnthropicActor object`
10022
10023 - `type: optional "anthropic_actor"`
10024
10025 default: anthropic_actor
10026
10027 - `email_address: optional string or null`
10028
10029 format: email
10030
10031 - `SystemActor object`
10032
10033 Automated background processing performed by Anthropic systems, acting
10034 without a user or customer credential.
10035
10036 - `type: optional "system_actor"`
10037
10038 default: system_actor
10039
10040 - `service: optional string or null`
10041
10042 Name of the automated process that performed the action, when known.
10043
10044 - `AdminAPIKeyActor object`
10045
10046 - `type: optional "admin_api_key_actor"`
10047
10048 default: admin_api_key_actor
10049
10050 - `admin_api_key_id: string`
10051
10052 - `ip_address: string`
10053
10054 - `user_agent: string`
10055
10056 - `ServiceAccountActor object`
10057
10058 - `type: optional "service_account_actor"`
10059
10060 default: service_account_actor
10061
10062 - `ip_address: string`
10063
10064 - `service_account_id: string`
10065
10066 - `user_agent: string`
10067
10068 - `ScimDirectorySyncActor object`
10069
10070 - `type: optional "scim_directory_sync_actor"`
10071
10072 default: scim_directory_sync_actor
10073
10074 - `directory_id: string`
10075
10076 - `workos_event_id: string`
10077
10078 - `idp_connection_type: optional string or null`
10079
10080 - `FederatedIdentityActor object`
10081
10082 A federated external workload authenticated via a verified OIDC token.
10083
10084 Carries the verified issuer, subject, and audience claims from the
10085 presented JWT.
10086
10087 - `type: optional "federated_identity_actor"`
10088
10089 default: federated_identity_actor
10090
10091 - `issuer: string`
10092
10093 - `subject: string`
10094
10095 - `audience: optional array of string`
10096
10097 - `ip_address: optional string or null`
10098
10099 - `user_agent: optional string or null`
10100
10101 - `FederatedActor object`
10102
10103 An external identity asserted by a trusted provider
api/compliance/activities/list Changed · +992 / -992 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 4075
40754075
40764076## Returns
40774077
4078- `data: optional array of object or object or object or 490 more`
4078- `data: optional array of AbuseDecisionReceived or AccountDeleted or AdminAPIKeyCreated or 490 more`
40794079
40804080 List of activity records. Each element's `type` field identifies which activity it is and which additional fields are present.
40814081
from line 4087
40874087
40884088 default: abuse_decision_received
40894089
4090 - `actor: object or object or object or 8 more`
4090 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
40914091
40924092 - `APIActor object`
40934093
from line 4221
42214221
42224222 default: federated_actor
42234223
4224 - `provider: object or object or object or object`
4224 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
42254225
42264226 - `FederatedActorAwsProvider object`
42274227
from line 4331
43314331
43324332 default: account_deleted
43334333
4334 - `actor: object or object or object or 8 more`
4334 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
43354335
43364336 - `APIActor object`
43374337
from line 4465
44654465
44664466 default: federated_actor
44674467
4468 - `provider: object or object or object or object`
4468 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
44694469
44704470 - `FederatedActorAwsProvider object`
44714471
from line 4563
45634563
45644564 default: admin_api_key_created
45654565
4566 - `actor: object or object or object or 8 more`
4566 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
45674567
45684568 - `APIActor object`
45694569
from line 4697
46974697
46984698 default: federated_actor
46994699
4700 - `provider: object or object or object or object`
4700 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
47014701
47024702 - `FederatedActorAwsProvider object`
47034703
from line 4803
48034803
48044804 default: admin_api_key_deleted
48054805
4806 - `actor: object or object or object or 8 more`
4806 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
48074807
48084808 - `APIActor object`
48094809
from line 4937
49374937
49384938 default: federated_actor
49394939
4940 - `provider: object or object or object or object`
4940 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
49414941
49424942 - `FederatedActorAwsProvider object`
49434943
from line 5039
50395039
50405040 default: admin_api_key_updated
50415041
5042 - `actor: object or object or object or 8 more`
5042 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
50435043
50445044 - `APIActor object`
50455045
from line 5173
51735173
51745174 default: federated_actor
51755175
5176 - `provider: object or object or object or object`
5176 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
51775177
51785178 - `FederatedActorAwsProvider object`
51795179
from line 5297
52975297
52985298 default: admin_connector_request_resolved
52995299
5300 - `actor: object or object or object or 8 more`
5300 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
53015301
53025302 - `APIActor object`
53035303
from line 5431
54315431
54325432 default: federated_actor
54335433
5434 - `provider: object or object or object or object`
5434 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
54355435
54365436 - `FederatedActorAwsProvider object`
54375437
from line 5541
55415541
55425542 default: admin_request_created
55435543
5544 - `actor: object or object or object or 8 more`
5544 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
55455545
55465546 - `APIActor object`
55475547
from line 5675
56755675
56765676 default: federated_actor
56775677
5678 - `provider: object or object or object or object`
5678 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
56795679
56805680 - `FederatedActorAwsProvider object`
56815681
from line 5775
57755775
57765776 default: admin_setup_checklist_step_delegated
57775777
5778 - `actor: object or object or object or 8 more`
5778 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
57795779
57805780 - `APIActor object`
57815781
from line 5909
59095909
59105910 default: federated_actor
59115911
5912 - `provider: object or object or object or object`
5912 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
59135913
59145914 - `FederatedActorAwsProvider object`
59155915
from line 6019
60196019
60206020 default: admin_setup_checklist_step_delegation_cancelled
60216021
6022 - `actor: object or object or object or 8 more`
6022 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
60236023
60246024 - `APIActor object`
60256025
from line 6153
61536153
61546154 default: federated_actor
61556155
6156 - `provider: object or object or object or object`
6156 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
61576157
61586158 - `FederatedActorAwsProvider object`
61596159
from line 6255
62556255
62566256 default: age_verified
62576257
6258 - `actor: object or object or object or 8 more`
6258 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
62596259
62606260 - `APIActor object`
62616261
from line 6389
63896389
63906390 default: federated_actor
63916391
6392 - `provider: object or object or object or object`
6392 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
63936393
63946394 - `FederatedActorAwsProvider object`
63956395
from line 6487
64876487
64886488 default: anonymous_mobile_login_attempted
64896489
6490 - `actor: object or object or object or 8 more`
6490 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
64916491
64926492 - `APIActor object`
64936493
from line 6621
66216621
66226622 default: federated_actor
66236623
6624 - `provider: object or object or object or object`
6624 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
66256625
66266626 - `FederatedActorAwsProvider object`
66276627
from line 6719
67196719
67206720 default: api_key_created
67216721
6722 - `actor: object or object or object or 8 more`
6722 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
67236723
67246724 - `APIActor object`
67256725
from line 6853
68536853
68546854 default: federated_actor
68556855
6856 - `provider: object or object or object or object`
6856 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
68576857
68586858 - `FederatedActorAwsProvider object`
68596859
from line 6963
69636963
69646964 default: claude_artifact_access_failed
69656965
6966 - `actor: object or object or object or 8 more`
6966 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
69676967
69686968 - `APIActor object`
69696969
from line 7097
70977097
70987098 default: federated_actor
70997099
7100 - `provider: object or object or object or object`
7100 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
71017101
71027102 - `FederatedActorAwsProvider object`
71037103
from line 7207
72077207
72087208 default: claude_artifact_commented
72097209
7210 - `actor: object or object or object or 8 more`
7210 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
72117211
72127212 - `APIActor object`
72137213
from line 7341
73417341
73427342 default: federated_actor
73437343
7344 - `provider: object or object or object or object`
7344 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
73457345
73467346 - `FederatedActorAwsProvider object`
73477347
from line 7483
74837483
74847484 default: claude_artifact_comments_viewed
74857485
7486 - `actor: object or object or object or 8 more`
7486 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
74877487
74887488 - `APIActor object`
74897489
from line 7617
76177617
76187618 default: federated_actor
76197619
7620 - `provider: object or object or object or object`
7620 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
76217621
76227622 - `FederatedActorAwsProvider object`
76237623
from line 7723
77237723
77247724 default: claude_artifact_created
77257725
7726 - `actor: object or object or object or 8 more`
7726 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
77277727
77287728 - `APIActor object`
77297729
from line 7857
78577857
78587858 default: federated_actor
78597859
7860 - `provider: object or object or object or object`
7860 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
78617861
78627862 - `FederatedActorAwsProvider object`
78637863
from line 7959
79597959
79607960 default: claude_published_artifact_deleted
79617961
7962 - `actor: object or object or object or 8 more`
7962 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
79637963
79647964 - `APIActor object`
79657965
from line 8093
80938093
80948094 default: federated_actor
80958095
8096 - `provider: object or object or object or object`
8096 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
80978097
80988098 - `FederatedActorAwsProvider object`
80998099
from line 8195
81958195
81968196 default: claude_artifact_published
81978197
8198 - `actor: object or object or object or 8 more`
8198 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
81998199
82008200 - `APIActor object`
82018201
from line 8329
83298329
83308330 default: federated_actor
83318331
8332 - `provider: object or object or object or object`
8332 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
83338333
83348334 - `FederatedActorAwsProvider object`
83358335
from line 8451
84518451
84528452 default: claude_artifact_sharing_updated
84538453
8454 - `actor: object or object or object or 8 more`
8454 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
84558455
84568456 - `APIActor object`
84578457
from line 8585
85858585
85868586 default: federated_actor
85878587
8588 - `provider: object or object or object or object`
8588 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
85898589
85908590 - `FederatedActorAwsProvider object`
85918591
from line 8657
86578657
86588658 - `user_agent: optional string or null`
86598659
8660 - `audience: array of object or object or object`
8660 - `audience: array of ArtifactSharingAudienceOrganization or ArtifactSharingAudienceUsers or ArtifactSharingAudienceAnyoneWithLink`
86618661
86628662 The artifact's sharing audience after the change. If empty, the artifact is visible only to its owner.
86638663
from line 8752
87528752
87538753 default: claude_artifact_viewed
87548754
8755 - `actor: object or object or object or 8 more`
8755 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
87568756
87578757 - `APIActor object`
87588758
from line 8886
88868886
88878887 default: federated_actor
88888888
8889 - `provider: object or object or object or object`
8889 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
88908890
88918891 - `FederatedActorAwsProvider object`
88928892
from line 8992
89928992
89938993 default: audit_log_export_accessed
89948994
8995 - `actor: object or object or object or 8 more`
8995 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
89968996
89978997 - `APIActor object`
89988998
from line 9126
91269126
91279127 default: federated_actor
91289128
9129 - `provider: object or object or object or object`
9129 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
91309130
91319131 - `FederatedActorAwsProvider object`
91329132
from line 9224
92249224
92259225 default: audit_log_export_started
92269226
9227 - `actor: object or object or object or 8 more`
9227 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
92289228
92299229 - `APIActor object`
92309230
from line 9358
93589358
93599359 default: federated_actor
93609360
9361 - `provider: object or object or object or object`
9361 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
93629362
93639363 - `FederatedActorAwsProvider object`
93649364
from line 9464
94649464
94659465 default: billing_emails_updated
94669466
9467 - `actor: object or object or object or 8 more`
9467 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
94689468
94699469 - `APIActor object`
94709470
from line 9598
95989598
95999599 default: federated_actor
96009600
9601 - `provider: object or object or object or object`
9601 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
96029602
96039603 - `FederatedActorAwsProvider object`
96049604
from line 9708
97089708
97099709 default: ccr_agent_created
97109710
9711 - `actor: object or object or object or 8 more`
9711 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
97129712
97139713 - `APIActor object`
97149714
from line 9842
98429842
98439843 default: federated_actor
98449844
9845 - `provider: object or object or object or object`
9845 - `provider: FederatedActorAwsProvider or FederatedActorAzureProvider or FederatedActorGcpProvider or FederatedActorOidcProvider`
98469846
98479847 - `FederatedActorAwsProvider object`
98489848
from line 9972
99729972
99739973 default: ccr_agent_deleted
99749974
9975 - `actor: object or object or object or 8 more`
9975 - `actor: APIActor or UserActor or UnauthenticatedUserActor or 8 more`
99769976
99779977 - `APIActor object`
99789978
9979 - `type: optional "api_actor"`
9980
9981 default: api_actor
9982
9983 - `api_key_id: string`
9984
9985 - `ip_address: string`
9986
9987 - `user_agent: string`
9988
9989 - `UserActor object`
9990
9991 - `type: optional "user_actor"`
9992
9993 default: user_actor
9994
9995 - `email_address: string`
9996
9997 format: email
9998
9999 - `ip_address: string`
10000
10001 - `user_agent: string`
10002
10003 - `user_id: string`
10004
10005 - `UnauthenticatedUserActor object`
10006
10007 - `type: optional "unauthenticated_user_actor"`
10008
10009 default: unauthenticated_user_actor
10010
10011 - `ip_address: string`
10012
10013 - `user_agent: string`
10014
10015 - `unauthenticated_email_address: optional string or null`
10016
10017 format: email
10018
10019 - `AnthropicActor object`
10020
10021 - `type: optional "anthropic_actor"`
10022
10023 default: anthropic_actor
10024
10025 - `email_address: optional string or null`
10026
10027 format: email
10028
10029 - `SystemActor object`
10030
10031 Automated background processing performed by Anthropic systems, acting
10032 without a user or customer credential.
10033
10034 - `type: optional "system_actor"`
10035
10036 default: system_actor
10037
10038 - `service: optional string or null`
10039
10040 Name of the automated process that performed the action, when known.
10041
10042 - `AdminAPIKeyActor object`
10043
10044 - `type: optional "admin_api_key_actor"`
10045
10046 default: admin_api_key_actor
10047
10048 - `admin_api_key_id: string`
10049
10050 - `ip_address: string`
10051
10052 - `user_agent: string`
10053
10054 - `ServiceAccountActor object`
10055
10056 - `type: optional "service_account_actor"`
10057
10058 default: service_account_actor
10059
10060 - `ip_address: string`
10061
10062 - `service_account_id: string`
10063
10064 - `user_agent: string`
10065
10066 - `ScimDirectorySyncActor object`
10067
10068 - `type: optional "scim_directory_sync_actor"`
10069
10070 default: scim_directory_sync_actor
10071
10072 - `directory_id: string`
10073
10074 - `workos_event_id: string`
10075
10076 - `idp_connection_type: optional string or null`
10077
10078 - `FederatedIdentityActor object`
10079
10080 A federated external workload authenticated via a verified OIDC token.
10081
10082 Carries the verified issuer, subject, and audience claims from the
10083 presented JWT.
10084
10085 - `type: optional "federated_identity_actor"`
10086
10087 default: federated_identity_actor
10088
10089 - `issuer: string`
10090
10091 - `subject: string`
10092
10093 - `audience: optional array of string`
10094
10095 - `ip_address: optional string or null`
10096
10097 - `user_agent: optional string or null`
10098
10099 - `FederatedActor object`
10100
10101 An external identity asserted by a trusted provid
9979 - `type: optional "api_act
api/compliance/apps Changed · +6 / -6 lines
from line 452
452452
453453 Artifact version ID e.g. 'claude_artifact_version_abc123'
454454
455 - `content: array of object or object or object`
455 - `content: array of Text or ToolUse or ToolResult`
456456
457457 Content blocks within the message
458458
from line 1486
14861486
14871487#### Returns
14881488
1489- `data: array of object or object`
1489- `data: array of ComplianceProjectFileReference or ComplianceProjectDocReference`
14901490
14911491 List of attachments sorted chronologically by created_at, tie break by id
14921492
from line 1640
16401640
16411641#### Returns
16421642
1643- `data: array of object or object or object or object`
1643- `data: array of ComplianceProjectUserCollaborator or ComplianceProjectGroupCollaborator or ComplianceProjectOrganizationCollaborator or ComplianceProjectOrganizationRoleCollaborator`
16441644
16451645 List of collaborators sorted chronologically by granted_at, tie break by the underlying role-assignment UUID
16461646
from line 2483
24832483
24842484 Message identifier, prefixed `clsm_`. Stable for as long as the message's turn is retained: identifiers of retained turns do not change as older turns age out of the organization's retention period. The `retention_elapsed` placeholder's identifier is distinct from every retained turn's and changes only when further turns age out.
24852485
2486 - `content: array of object or object or object`
2486 - `content: array of Text or ToolUse or ToolResult`
24872487
24882488 Content blocks within the message, discriminated on `type` (`text` / `tool_use` / `tool_result`: the same discriminator values as the claude.ai chat-messages endpoint; the tool variants omit `integration_name` and `mcp_server_url`, and `text` carries `truncated`). Extended-thinking content is never included. The request's `system` field is never included; a presence-only marker message is emitted when it was set. The request's `tools[]` definitions are never included as transcript messages. Project-level instructions (such as CLAUDE.md files) appear in the message stream as a user-role context block and are included. Empty when `provenance.type` is `content_unavailable`.
24892489
from line 2579
25792579
25802580 The model that served this assistant turn, as reported in the `model` field of the underlying Messages API response. Null on user messages and on any assistant message whose `provenance` is set: client-asserted history and synthetic markers were not produced by a model during this session, and for unavailable content the serving model is not known.
25812581
2582 - `provenance: object or object or object or null`
2582 - `provenance: ContentUnavailable or ClientAsserted or SyntheticMarker or null`
25832583
25842584 Where this turn's content came from, discriminated on `type`. Null (the common case) means verified content: on an assistant message, content Claude produced during this session; on a user message, content the user sent. `content_unavailable`: the turn's content cannot be returned and `content` is empty; `reason` says why. `client_asserted`: assistant content the client supplied as conversation history; `content` shows what the model received but its authorship is not verified; never on user-role messages. `synthetic_marker`: a transcript marker the endpoint generated rather than content either party sent during the session. Both `client_asserted` and `synthetic_marker` can result from normal request or client processing, not only client modification. Callers should tolerate unrecognized `type` values.
25852585
from line 3007
30073007
30083008 Unique identifier for the message, e.g. `csev_abc123`
30093009
3010 - `content: array of object or object or object`
3010 - `content: array of Text or ToolUse or ToolResult`
30113011
30123012 Content blocks within the message
30133013
api/compliance/apps/sessions Changed · +3 / -3 lines
from line 329
329329
330330 Message identifier, prefixed `clsm_`. Stable for as long as the message's turn is retained: identifiers of retained turns do not change as older turns age out of the organization's retention period. The `retention_elapsed` placeholder's identifier is distinct from every retained turn's and changes only when further turns age out.
331331
332 - `content: array of object or object or object`
332 - `content: array of Text or ToolUse or ToolResult`
333333
334334 Content blocks within the message, discriminated on `type` (`text` / `tool_use` / `tool_result`: the same discriminator values as the claude.ai chat-messages endpoint; the tool variants omit `integration_name` and `mcp_server_url`, and `text` carries `truncated`). Extended-thinking content is never included. The request's `system` field is never included; a presence-only marker message is emitted when it was set. The request's `tools[]` definitions are never included as transcript messages. Project-level instructions (such as CLAUDE.md files) appear in the message stream as a user-role context block and are included. Empty when `provenance.type` is `content_unavailable`.
335335
from line 425
425425
426426 The model that served this assistant turn, as reported in the `model` field of the underlying Messages API response. Null on user messages and on any assistant message whose `provenance` is set: client-asserted history and synthetic markers were not produced by a model during this session, and for unavailable content the serving model is not known.
427427
428 - `provenance: object or object or object or null`
428 - `provenance: ContentUnavailable or ClientAsserted or SyntheticMarker or null`
429429
430430 Where this turn's content came from, discriminated on `type`. Null (the common case) means verified content: on an assistant message, content Claude produced during this session; on a user message, content the user sent. `content_unavailable`: the turn's content cannot be returned and `content` is empty; `reason` says why. `client_asserted`: assistant content the client supplied as conversation history; `content` shows what the model received but its authorship is not verified; never on user-role messages. `synthetic_marker`: a transcript marker the endpoint generated rather than content either party sent during the session. Both `client_asserted` and `synthetic_marker` can result from normal request or client processing, not only client modification. Callers should tolerate unrecognized `type` values.
431431
from line 853
853853
854854 Unique identifier for the message, e.g. `csev_abc123`
855855
856 - `content: array of object or object or object`
856 - `content: array of Text or ToolUse or ToolResult`
857857
858858 Content blocks within the message
859859
api/compliance/apps/sessions/local/messages Changed · +4 / -4 lines
from line 81
8181
8282 Message identifier, prefixed `clsm_`. Stable for as long as the message's turn is retained: identifiers of retained turns do not change as older turns age out of the organization's retention period. The `retention_elapsed` placeholder's identifier is distinct from every retained turn's and changes only when further turns age out.
8383
84 - `content: array of object or object or object`
84 - `content: array of Text or ToolUse or ToolResult`
8585
8686 Content blocks within the message, discriminated on `type` (`text` / `tool_use` / `tool_result`: the same discriminator values as the claude.ai chat-messages endpoint; the tool variants omit `integration_name` and `mcp_server_url`, and `text` carries `truncated`). Extended-thinking content is never included. The request's `system` field is never included; a presence-only marker message is emitted when it was set. The request's `tools[]` definitions are never included as transcript messages. Project-level instructions (such as CLAUDE.md files) appear in the message stream as a user-role context block and are included. Empty when `provenance.type` is `content_unavailable`.
8787
from line 177
177177
178178 The model that served this assistant turn, as reported in the `model` field of the underlying Messages API response. Null on user messages and on any assistant message whose `provenance` is set: client-asserted history and synthetic markers were not produced by a model during this session, and for unavailable content the serving model is not known.
179179
180 - `provenance: object or object or object or null`
180 - `provenance: ContentUnavailable or ClientAsserted or SyntheticMarker or null`
181181
182182 Where this turn's content came from, discriminated on `type`. Null (the common case) means verified content: on an assistant message, content Claude produced during this session; on a user message, content the user sent. `content_unavailable`: the turn's content cannot be returned and `content` is empty; `reason` says why. `client_asserted`: assistant content the client supplied as conversation history; `content` shows what the model received but its authorship is not verified; never on user-role messages. `synthetic_marker`: a transcript marker the endpoint generated rather than content either party sent during the session. Both `client_asserted` and `synthetic_marker` can result from normal request or client processing, not only client modification. Callers should tolerate unrecognized `type` values.
183183
from line 352
352352
353353 Message identifier, prefixed `clsm_`. Stable for as long as the message's turn is retained: identifiers of retained turns do not change as older turns age out of the organization's retention period. The `retention_elapsed` placeholder's identifier is distinct from every retained turn's and changes only when further turns age out.
354354
355 - `content: array of object or object or object`
355 - `content: array of Text or ToolUse or ToolResult`
356356
357357 Content blocks within the message, discriminated on `type` (`text` / `tool_use` / `tool_result`: the same discriminator values as the claude.ai chat-messages endpoint; the tool variants omit `integration_name` and `mcp_server_url`, and `text` carries `truncated`). Extended-thinking content is never included. The request's `system` field is never included; a presence-only marker message is emitted when it was set. The request's `tools[]` definitions are never included as transcript messages. Project-level instructions (such as CLAUDE.md files) appear in the message stream as a user-role context block and are included. Empty when `provenance.type` is `content_unavailable`.
358358
from line 448
448448
449449 The model that served this assistant turn, as reported in the `model` field of the underlying Messages API response. Null on user messages and on any assistant message whose `provenance` is set: client-asserted history and synthetic markers were not produced by a model during this session, and for unavailable content the serving model is not known.
450450
451 - `provenance: object or object or object or null`
451 - `provenance: ContentUnavailable or ClientAsserted or SyntheticMarker or null`
452452
453453 Where this turn's content came from, discriminated on `type`. Null (the common case) means verified content: on an assistant message, content Claude produced during this session; on a user message, content the user sent. `content_unavailable`: the turn's content cannot be returned and `content` is empty; `reason` says why. `client_asserted`: assistant content the client supplied as conversation history; `content` shows what the model received but its authorship is not verified; never on user-role messages. `synthetic_marker`: a transcript marker the endpoint generated rather than content either party sent during the session. Both `client_asserted` and `synthetic_marker` can result from normal request or client processing, not only client modification. Callers should tolerate unrecognized `type` values.
454454
api/compliance/organizations/settings Changed · +4 / -4 lines
from line 84
8484
8585- `organization_id: string`
8686
87- `settings: array of object or object or object or 3 more`
87- `settings: array of Boolean or Integer or String or 3 more`
8888
8989 - `Boolean object`
9090
from line 303
303303
304304 default: data_retention
305305
306 - `value: map[object or object]`
306 - `value: map[Fixed or Indefinite]`
307307
308308 - `Fixed object`
309309
from line 429
429429
430430 - `organization_id: string`
431431
432 - `settings: array of object or object or object or 3 more`
432 - `settings: array of Boolean or Integer or String or 3 more`
433433
434434 - `Boolean object`
435435
from line 648
648648
649649 default: data_retention
650650
651 - `value: map[object or object]`
651 - `value: map[Fixed or Indefinite]`
652652
653653 - `Fixed object`
654654
build-with-claude/preserved-thinking Changed · +215 / -73 lines
## Who needs to change anything #### Handle the error in code ### Put changing context in the newest turn ### Libraries, proxies, and gateways
from line 11
1111
1212The model check applies to every account. The API enforces the prefix check by default for accounts created on or after August 31, 2026, 00:00 UTC. On older accounts, it enforces the prefix check only on requests that set `thinking.block_binding.prefix_mismatch_behavior`. **Later models will enforce the prefix check for all accounts**, so make your integration append-only now.
1313
14## Who needs to change anything
15
16Nothing changes for you if Claude Code, claude.ai, Claude Managed Agents, or the Claude Agent SDK builds your requests, or if your code keeps `system` and `tools` fixed for a session and only ever appends to `messages`. Claude Mythos 5.1 and models before Claude Fable 5.1 don't run the prefix check. If you never send thinking blocks back, the prefix check has nothing to reject, and the model gets none of its earlier reasoning.
17
18Check your integration if, between two requests in one conversation, it does any of the following. Each item links to what to do instead:
19
20* [Rebuilds the `system` prompt](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#new-instructions): the date, a mode flag, re-read project instructions, or a plugin or MCP server that connects after the first turn
21* [Re-renders the context in the first user message](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#changing-context)
22* [Clears or shortens old tool results, or re-encodes old images](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#server-side-trimming)
23* [Summarizes or drops old turns on the client](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#custom-compaction-on-the-client) and keeps recent turns with their thinking
24* [Adds, removes, or edits entries in `tools`](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#tool-changes)
25* [Adds a reminder to a user turn](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#per-turn-reminders) and removes or rewrites it later
26* [Drops some `thinking` blocks and keeps later ones](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#append-assistant-turns-exactly-as-returned), or removes them and [later puts them back](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#prefix-check)
27* [Rebuilds a saved session from templates](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#faq) instead of replaying what it sent
28
29On an older account, none of these produces an error unless the request sets `prefix_mismatch_behavior`, so a run with no errors on your own key doesn't show whether your code is affected. If people run your tool with their own API keys, those on newer accounts get the 400 error before you do. [Set `prefix_mismatch_behavior` in your tests](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#how-to-tell-whether-your-integration-is-impacted) to see what they see.
30
1431## Switching models mid-conversation
1532
1633Claude Fable 5.1 and Claude Mythos 5.1 read thinking blocks produced by each other and by earlier Claude models. No earlier model reads thinking blocks from Claude Fable 5.1 or Claude Mythos 5.1.
from line 67
5067
5168Request parameters outside those three fields, such as `effort`, `max_tokens`, `output_config`, `tool_choice`, and `metadata`, aren't part of the prefix check, and neither are `cache_control` markers. [What counts as an edit](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#what-counts-as-an-edit) has the full list.
5269
53Earlier thinking blocks aren't in the prefix, but each thinking block records which thinking block came before it, across turns. You can remove thinking blocks from the front of the history, oldest first. Removing one from the middle invalidates thinking blocks after it.
70Earlier thinking blocks aren't in the prefix, but each thinking block records which thinking block came before it, across turns. You can remove thinking blocks from the start of the history (oldest first), from the end, or all of them. What fails is a gap: the thinking blocks you keep must be an unbroken run of the original sequence, so removing one from the middle invalidates the thinking blocks after it. Once you remove a block, leave it out. Putting it back invalidates the thinking blocks produced while it was gone.
5471
5572Keep `system` and `tools` fixed for the session and treat `messages` as append-only. The same discipline keeps the prefix stable for [prompt caching](https://platform.claude.com/docs/en/build-with-claude/prompt-caching): the edits that invalidate thinking are the edits that restart the cache.
5673
from line 78
6178* **`"error"` (the default):** the API rejects the request with a 400 `invalid_request_error` that names the first failing block.
6279* **`"drop_block"`:** the API drops each failing block and every thinking block after it, and the request succeeds. Dropped blocks aren't billed. The model answers that turn without using reasoning from dropped blocks, and the prompt cache restarts at the edit. The response lists each dropped block in `input_transformations` (on the `message_start` event when streaming) with `reason: "prefix_binding_mismatch"`.
6380
81`"drop_block"` keeps requests succeeding but doesn't fix the edit. Count the responses in each session whose `input_transformations` has a `prefix_binding_mismatch` entry, and alert on them. In the Message Batches API, an item that leaves the field unset drops failing blocks instead of erroring, so set `"error"` explicitly there if you want batch items to fail.
82
6483Both the field and the `input_transformations` array require the `thinking-binding-controls-2026-08-01` [beta header](https://platform.claude.com/docs/en/api/beta-headers). [Set the mismatch behavior and read `input_transformations`](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#preserved-thinking-controls) shows the request in each SDK.
6584
6685The 400 message begins:
from line 94
7594That setting requires the `thinking-binding-controls-2026-08-01` value in the `anthropic-beta` header.
7695```
7796
78It usually ends with a sentence naming what changed, for example that the `system` prompt or the `tools` list differs from when the block was created. See [Troubleshooting thinking](https://platform.claude.com/docs/en/build-with-claude/thinking-troubleshooting#error-thinking-block-signature) for every variant of this error.
97It usually ends with a sentence naming what changed, for example that the `system` prompt or the `tools` list differs from when the block was created. [Troubleshooting thinking](https://platform.claude.com/docs/en/build-with-claude/thinking-troubleshooting#error-thinking-block-signature) describes what that sentence can name.
7998
80If you hit this 400 in production, retrying the same body fails the same way. Retry with the beta header and `prefix_mismatch_behavior: "drop_block"` and keep sending it for the rest of the session, or strip every `thinking` and `redacted_thinking` block from the history yourself and retry once. Then fix the edit that caused the mismatch. In the Message Batches API, an item that leaves the field unset drops failing blocks instead of erroring, so set `"error"` explicitly there if you want batch items to fail.
81
8299A tampered or undecryptable signature is a different failure. It always returns a 400 (``Invalid `signature` in `thinking` block`` with no sentence about the conversation), and `prefix_mismatch_behavior` doesn't apply to it.
83100
101#### Handle the error in code
102
103This is the 400 `invalid_request_error` shown earlier in this section. Don't resend the same body: it fails the same way every time. Retry once with the beta header and `prefix_mismatch_behavior: "drop_block"`, and store that choice with the session so every later request sends it too, including after a restart. If you can't send the beta header, remove every `thinking` and `redacted_thinking` block from the history once, leave them out, and continue. Then fix the edit that caused the mismatch.
104
84105### Set the mismatch behavior and read `input_transformations`
85106
86107The `thinking-binding-controls-2026-08-01` [beta header](https://platform.claude.com/docs/en/api/beta-headers) adds:
from line 109
88109* A top-level `input_transformations` array on every response
89110* A `block_binding` object on the `thinking` configuration, whose one field is `prefix_mismatch_behavior`
90111
91`block_binding` is accepted alongside `thinking.type: "adaptive"` and `thinking.type: "enabled"`. Sending it without the beta header returns a 400 error. Models that don't run the prefix check accept the object and report only model-check drops, so one request body works across models.
112`block_binding` is accepted alongside `thinking.type: "adaptive"` and `thinking.type: "enabled"`. Sending it without the beta header returns a 400 error whose message ends `block_binding: Extra inputs are not permitted`. Models that don't run the prefix check accept the object and report only model-check drops, so one request body works across models. The API reference calls the prefix check the conversation check.
92113
93114The following request opts into dropping rather than rejecting. On a first turn there's nothing to replay, so `input_transformations` comes back empty:
94115
from line 380
359380
360381To find out which group your account is in, take a Claude Fable 5.1 conversation that contains a thinking block, change something before that block, and send it to Claude Fable 5.1 without the beta header or the `block_binding` field. A 400 response that names the header means your account is enforced by default.
361382
362<Note>
363 If you maintain a tool or framework that people run with their own API key, your users on new accounts hit the check before you do, because your own key is likely on an older account. Test with `prefix_mismatch_behavior` set so you see what they see.
364</Note>
365
366383### What counts as an edit
367384
368385Each row compares two consecutive requests:
369386
370| Change between requests | Later thinking blocks |
371| ------------------------------------------------------------------------------------------------------------------------------------------------------ | ---------------------------------------------------------------------- |
372| Append messages at the end | Valid |
373| Add a tool with `defer_loading: true` that nothing has referenced yet | Valid |
374| Remove `thinking` blocks from the start of the history | Valid |
375| Change any request parameter outside `system`, `tools`, and `messages` (`effort`, `max_tokens`, `output_config`, `tool_choice`, `metadata`, and so on) | Valid |
376| Add, move, or remove `cache_control` markers | Valid |
377| A rotating signed URL that returns the same bytes | Valid |
378| Server-side compaction or context editing removes or replaces content | Valid (the check compares what you sent, not the server's edited copy) |
379| A cleared [turn-scoped system message](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#per-turn-reminders) left in place | Valid |
380| Edit, reorder, or delete any earlier `user`, `assistant`, or `system` message | Invalid |
381| Add a text block to an earlier user turn, or remove one you added last time | Invalid |
382| Change the top-level `system` string or blocks | Invalid |
383| Add, remove, rename, or edit a tool in `tools` | Invalid |
384| Remove a `thinking` block from the middle of the history and keep later ones | Invalid for every later thinking block |
385| An image or document URL that returns different bytes on the next request | Invalid |
386| The same turn-scoped message deleted or reworded on a later request | Invalid |
387| Change between requests | Later thinking blocks |
388| -------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------- |
389| Append messages at the end | Valid |
390| Add a tool with `defer_loading: true` that nothing has referenced yet | Valid |
391| Remove `thinking` blocks from the start of the history, from the end, or all of them | Valid (the model loses that reasoning) |
392| Change any request parameter outside `system`, `tools`, and `messages` (`effort`, `max_tokens`, `output_config`, `tool_choice`, `metadata`, `thinking.display`, and so on) | Valid |
393| Add, move, or remove `cache_control` markers | Valid |
394| A rotating signed URL that returns the same bytes | Valid |
395| Server-side compaction or context editing removes or replaces content | Valid (the check compares what you sent, not the server's edited copy) |
396| A cleared [turn-scoped system message](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#per-turn-reminders) left in place | Valid |
397| Edit, reorder, or delete any earlier `user`, `assistant`, or `system` message | Invalid |
398| Re-render the context you put in the first user message with a changed value | Invalid for every thinking block |
399| Clear or shorten an earlier `tool_result`, re-encode an earlier image, or change an earlier `tool_use` input | Invalid for every later thinking block |
400| Add a text block to an earlier user turn, or remove one you added last time | Invalid |
401| Change the top-level `system` string or blocks | Invalid |
402| Add, remove, rename, or edit a tool in `tools` | Invalid |
403| Remove a `thinking` block from the middle of the history and keep later ones | Invalid for every later thinking block |
404| Put back a `thinking` block you removed on an earlier request | Invalid for thinking blocks produced while it was gone |
405| An image or document URL that returns different bytes on the next request | Invalid |
406| The same turn-scoped message deleted or reworded on a later request | Invalid |
387407
388408### Check whether your code edits the prefix
389409
390410First, diff what you send. Capture the request bodies your integration sends over a few normal turns, including a compaction or a tool change. For each pair of consecutive requests, compare `system`, `tools`, and the `messages` they share. They should be identical up to the newly appended turns.
391411
392Then confirm against the API. Add the `thinking-binding-controls-2026-08-01` beta header, set `prefix_mismatch_behavior` to `"drop_block"`, and run a normal multi-turn session through your integration on claude-fable-5-1. The following example runs two turns the way your integration should: `messages` only grows, each assistant turn goes back exactly as the API returned it, `thinking` blocks included, and `block_binding` is set on every request. It prints the number of dropped blocks after each turn:
412Then confirm against the API. Add the `thinking-binding-controls-2026-08-01` beta header, set `prefix_mismatch_behavior` to `"drop_block"`, and run a normal multi-turn session through your integration on claude-fable-5-1. The following example runs two turns the way your integration should: `messages` only grows, each assistant turn goes back exactly as the API returned it, `thinking` blocks included, and `block_binding` is set on every request. After each turn it prints the number of `thinking` blocks in the response and the number of dropped blocks:
393413
394414<CodeGroup>
395415 ```bash cURL
416 # Counts the thinking blocks in a response and the blocks the API dropped
417 COUNTS='"thinking blocks: \([.content[] | select(.type == "thinking")] | length), " +
418 "dropped: \(.input_transformations | length)"'
419
396420 FIRST=$(curl -s https://api.anthropic.com/v1/messages \
397421 -H "content-type: application/json" \
398422 -H "x-api-key: $ANTHROPIC_API_KEY" \
from line 429
405429 "type": "adaptive",
406430 "block_binding": { "prefix_mismatch_behavior": "drop_block" }
407431 },
408 "messages": [{ "role": "user", "content": "What is 27 * 453?" }]
432 "messages": [
433 {
434 "role": "user",
435 "content": "How many positive integers below 500 have exactly 6 positive divisors?"
436 }
437 ]
409438 }')
410 echo "$FIRST" | jq '.input_transformations | length'
439 echo "$FIRST" | jq -r "$COUNTS"
411440
412441 # Turn 2: the assistant turn goes back exactly as returned, then the next user message
413442 MESSAGES=$(jq -n --argjson first "$FIRST" '[
414 { role: "user", content: "What is 27 * 453?" },
443 {
444 role: "user",
445 content: "How many positive integers below 500 have exactly 6 positive divisors?"
446 },
415447 { role: "assistant", content: $first.content },
416 { role: "user", content: "Now divide that result by 3." }
448 { role: "user", content: "How many of those are odd?" }
417449 ]')
418450
419451 jq -n --argjson messages "$MESSAGES" '{
from line 461
429461 -H "x-api-key: $ANTHROPIC_API_KEY" \
430462 -H "anthropic-version: 2023-06-01" \
431463 -H "anthropic-beta: thinking-binding-controls-2026-08-01" \
432 -d @- | jq '.input_transformations | length'
464 -d @- | jq -r "$COUNTS"
433465 ```
434466
435467 ```bash CLI
468 # Counts the thinking blocks in a response and the blocks the API dropped
469 COUNTS='"thinking blocks: \([.content[] | select(.type == "thinking")] | length), " +
470 "dropped: \(.input_transformations | length)"'
471
436472 FIRST=$(ant beta:messages create --beta thinking-binding-controls-2026-08-01 \
437 --transform content --format json <<'YAML'
473 --format json <<'YAML'
438474 model: claude-fable-5-1
439475 max_tokens: 16000
440476 thinking:
from line 479
443479 prefix_mismatch_behavior: drop_block
444480 messages:
445481 - role: user
446 content: What is 27 * 453?
482 content: How many positive integers below 500 have exactly 6 positive divisors?
447483 YAML
448484 )
485 echo "$FIRST" | jq -r "$COUNTS"
449486
450487 # Turn 2: the assistant turn goes back exactly as returned, then the next user message
451488 ant beta:messages create --beta thinking-binding-controls-2026-08-01 \
452 --transform input_transformations --format json <<YAML
489 --format json <<YAML | jq -r "$COUNTS"
453490 model: claude-fable-5-1
454491 max_tokens: 16000
455492 thinking:
from line 495
458495 prefix_mismatch_behavior: drop_block
459496 messages:
460497 - role: user
461 content: What is 27 * 453?
498 content: How many positive integers below 500 have exactly 6 positive divisors?
462499 - role: assistant
463 content: $(echo "$FIRST" | jq -c .)
500 content: $(echo "$FIRST" | jq -c .content)
464501 - role: user
465 content: Now divide that result by 3.
502 content: How many of those are odd?
466503 YAML
467504 ```
468505
from line 506
469506 ```python Python
470507 client = anthropic.Anthropic()
471508
509 user_turns = [
510 "How many positive integers below 500 have exactly 6 positive divisors?",
511 "How many of those are odd?",
512 ]
513
472514 # messages grows across turns: each assistant turn goes back exactly as returned
473515 messages = []
474 for user_turn in ["What is 27 * 453?", "Now divide that result by 3."]:
516 for user_turn in user_turns:
475517 messages.append({"role": "user", "content": user_turn})
476518 response = client.beta.messages.create(
477519 model="claude-fable-5-1",
from line 526
484526 betas=["thinking-binding-controls-2026-08-01"],
485527 )
486528 messages.append({"role": "assistant", "content": response.content})
487 print(len(response.input_transformations or []))
529 thinking_blocks = sum(block.type == "thinking" for block in response.content)
530 dropped = len(response.input_transformations or [])
531 print(f"thinking blocks: {thinking_blocks}, dropped: {dropped}")
488532 ```
489533
490534 ```typescript TypeScript
491535 const client = new Anthropic();
492536
537 const userTurns = [
538 "How many positive integers below 500 have exactly 6 positive divisors?",
539 "How many of those are odd?"
540 ];
541
493542 // messages grows across turns: each assistant turn goes back exactly as returned
494543 const messages: Anthropic.Beta.BetaMessageParam[] = [];
495 for (const userTurn of ["What is 27 * 453?", "Now divide that result by 3."]) {
544 for (const userTurn of userTurns) {
496545 messages.push({ role: "user", content: userTurn });
497546 const response = await client.beta.messages.create({
498547 model: "claude-fable-5-1",
from line 554
505554 betas: ["thinking-binding-controls-2026-08-01"]
506555 });
507556 messages.push({ role: "assistant", content: response.content });
508 console.log(response.input_transformations?.length ?? 0);
557 const thinkingBlocks = response.content.filter((block) => block.type === "thinking");
558 const dropped = response.input_transformations ?? [];
559 console.log(`thinking blocks: ${thinkingBlocks.length}, dropped: ${dropped.length}`);
509560 }
510561 ```
511562
from line 563
512563 ```csharp C#
513564 AnthropicClient client = new();
514565
566 string[] userTurns =
567 [
568 "How many positive integers below 500 have exactly 6 positive divisors?",
569 "How many of those are odd?",
570 ];
571
515572 // messages grows across turns: each assistant turn goes back exactly as returned
516573 List<BetaMessageParam> messages = [];
517 foreach (var userTurn in new[] { "What is 27 * 453?", "Now divide that result by 3." })
574 foreach (var userTurn in userTurns)
518575 {
519576 messages.Add(new() { Role = Role.User, Content = userTurn });
520577 var response = await client.Beta.Messages.Create(
from line 595
538595 Role = Role.Assistant,
539596 Content = response.Content.Select(block => new BetaContentBlockParam(block.Json)).ToList(),
540597 });
541 Console.WriteLine(response.InputTransformations?.Count ?? 0);
598 var thinkingBlocks = response.Content.Count(block => block.TryPickThinking(out _));
599 var dropped = response.InputTransformations?.Count ?? 0;
600 Console.WriteLine($"thinking blocks: {thinkingBlocks}, dropped: {dropped}");
542601 }
543602 ```
544603
from line 604
545604 ```go Go
546605 client := anthropic.NewClient()
547606
607 userTurns := []string{
608 "How many positive integers below 500 have exactly 6 positive divisors?",
609 "How many of those are odd?",
610 }
611
548612 // messages grows across turns: each assistant turn goes back exactly as returned
549613 messages := []anthropic.BetaMessageParam{}
550 for _, userTurn := range []string{"What is 27 * 453?", "Now divide that result by 3."} {
614 for _, userTurn := range userTurns {
551615 messages = append(messages, anthropic.NewBetaUserMessage(anthropic.NewBetaTextBlock(userTurn)))
552616 response, err := client.Beta.Messages.New(context.TODO(), anthropic.BetaMessageNewParams{
553617 Model: "claude-fable-5-1",
from line 630
566630 log.Fatal(err)
567631 }
568632 messages = append(messages, response.ToParam())
569 fmt.Println(len(response.InputTransformations))
633 thinkingBlocks := 0
634 for _, block := range response.Content {
635 if block.Type == "thinking" {
636 thinkingBlocks++
637 }
638 }
639 fmt.Printf("thinking blocks: %d, dropped: %d\n", thinkingBlocks, len(response.InputTransformations))
570640 }
571641 ```
572642
573643 ```java Java
574644 import com.anthropic.models.beta.AnthropicBeta;
645 import com.anthropic.models.beta.messages.BetaContentBlock;
575646 import com.anthropic.models.beta.messages.BetaMessage;
576647 import com.anthropic.models.beta.messages.BetaThinkingBlockBinding;
577648 import com.anthropic.models.beta.messages.BetaThinkingConfigAdaptive;
from line 652
581652 void main() {
582653 AnthropicClient client = AnthropicOkHttpClient.fromEnv();
583654
655 List<String> userTurns = List.of(
656 "How many positive integers below 500 have exactly 6 positive divisors?",
657 "How many of those are odd?");
658
584659 // The builder's message list grows across turns: each assistant turn goes back exactly as returned
585660 MessageCreateParams.Builder conversation = MessageCreateParams.builder()
586661 .model("claude-fable-5-1")
from line 667
592667 .build())
593668 .addBeta(AnthropicBeta.THINKING_BINDING_CONTROLS_2026_08_01);
594669
595 for (String userTurn : List.of("What is 27 * 453?", "Now divide that result by 3.")) {
670 for (String userTurn : userTurns) {
596671 conversation.addUserMessage(userTurn);
597672 BetaMessage response = client.beta().messages().create(conversation.build());
598673 conversation.addMessage(response);
599 IO.println(response.inputTransformations().map(List::size).orElse(0));
674 long thinkingBlocks = response.content().stream()
675 .filter(BetaContentBlock::isThinking)
676 .count();
677 int dropped = response.inputTransformations().map(List::size).orElse(0);
678 IO.println("thinking blocks: " + thinkingBlocks + ", dropped: " + dropped);
600679 }
601680 }
602681 ```
from line 689
610689
611690 $client = new Client();
612691
692 $userTurns = [
693 'How many positive integers below 500 have exactly 6 positive divisors?',
694 'How many of those are odd?',
695 ];
696
613697 // $messages grows across turns: each assistant turn goes back exactly as returned
614698 $messages = [];
615 foreach (['What is 27 * 453?', 'Now divide that result by 3.'] as $userTurn) {
699 foreach ($userTurns as $userTurn) {
616700 $messages[] = ['role' => 'user', 'content' => $userTurn];
617701 $response = $client->beta->messages->create(
618702 model: 'claude-fable-5-1',
from line 710
626710 betas: [AnthropicBeta::THINKING_BINDING_CONTROLS_2026_08_01],
627711 );
628712 $messages[] = ['role' => 'assistant', 'content' => $response->content];
629 echo count($response->inputTransformations ?? []), PHP_EOL;
713 $thinkingBlocks = array_filter($response->content, fn ($block) => $block->type === 'thinking');
714 $dropped = $response->inputTransformations ?? [];
715 echo 'thinking blocks: ', count($thinkingBlocks), ', dropped: ', count($dropped), PHP_EOL;
630716 }
631717 ```
632718
from line 719
633719 ```ruby Ruby
634720 client = Anthropic::Client.new
635721
722 user_turns = [
723 "How many positive integers below 500 have exactly 6 positive divisors?",
724 "How many of those are odd?"
725 ]
726
636727 # messages grows across turns: each assistant turn goes back exactly as returned
637728 messages = []
638 ["What is 27 * 453?", "Now divide that result by 3."].each do |user_turn|
729 user_turns.each do |user_turn|
639730 messages << {role: "user", content: user_turn}
640731 response = client.beta.messages.create(
641732 model: "claude-fable-5-1",
from line 739
648739 betas: [Anthropic::AnthropicBeta::THINKING_BINDING_CONTROLS_2026_08_01]
649740 )
650741 messages << {role: "assistant", content: response.content}
651 puts (response.input_transformations || []).length
742 thinking_blocks = response.content.count { |block| block.type == :thinking }
743 dropped = (response.input_transformations || []).length
744 puts "thinking blocks: #{thinking_blocks}, dropped: #{dropped}"
652745 end
653746 ```
654747</CodeGroup>
655748
656749```text Output wrap
6570
6580
750thinking blocks: 1, dropped: 0
751thinking blocks: 1, dropped: 0
659752```
660753
661Both turns print `0` because nothing earlier changed. Log `input_transformations` on every turn of your own integration. When the API drops a block, the entry looks like the following:
754Neither turn drops a block because nothing earlier changed. Check that the first response contains a `thinking` block. With adaptive thinking, some responses have none. If no response in the session has one, there is nothing to check and the dropped count is 0 whatever you change, so run the example again.
662755
756Log `input_transformations` on every turn of your own integration. When the API drops a block, the entry looks like the following:
757
663758```json
664759{
665760 "input_transformations": [
from line 767
672767}
673768```
674769
675* **Empty on every turn:** your integration keeps the prefix intact.
676* **`reason: "prefix_binding_mismatch"`:** something before the block at `path` changed since the previous request. Diff `system`, `tools`, and `messages` up to that turn to find it, then find the matching replacement in [Make changes without editing the prefix](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#replace-prefix-edits).
770* **Empty on every turn of a session that contains `thinking` blocks:** your integration keeps the prefix intact.
771* **`reason: "prefix_binding_mismatch"`:** something before the block at `path` changed since the previous request. Diff `system`, `tools`, and `messages` up to that turn to find it, or resend the request with `"error"`: the 400 usually ends with a sentence naming what changed. Then find the matching replacement in [Make changes without editing the prefix](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#replace-prefix-edits).
677772* **`reason: "model_binding_mismatch"`:** the conversation moved to a model that can't read the earlier model's blocks. This isn't a prefix edit. See [Switching models mid-conversation](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#switching-models).
678773
679To fail loudly in CI instead, set `"error"` and treat the 400 described in [What the API does with an invalid block](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#mismatch-behavior) as a test failure.
774To see a failure on purpose, send a third turn from the earlier example and add a `system` prompt to that request only, so that it differs from the first two requests, which had none. With `"drop_block"`, the dropped count is no longer 0: the response has one entry for each thinking block in the history, each with `reason: "prefix_binding_mismatch"`. With `"error"`, the request returns the 400 described in [What the API does with an invalid block](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#mismatch-behavior), and its last sentence names the `system` prompt. In the cURL and CLI tabs, remove the `jq` filter to see the error body. If the count is still 0, there was nothing to check: confirm that the model is claude-fable-5-1, that the request sets `block_binding`, that the history you sent contains `thinking` blocks, and that the first two requests had no `system` prompt.
680775
776Two plain turns rarely show the problem. Run a session through each of the following, with `"error"` set so that a regression fails your CI:
777
778* The first client-side compaction or trim
779* A tool, plugin, or MCP server that connects after the first turn
780* A mode or instruction change
781* A long tool loop, if you add reminders or shorten old tool results
782* A switch to another model and back
783* A save, a restart, and a resume on a later date
784
681785## Make changes without editing the prefix
682786
683787Each common prefix edit has a replacement that gives the model the same information and leaves earlier bytes unchanged, so later thinking stays valid. Find the edit your code makes today in the first column:
684788
685| Instead of | Use | Beta header |
686| --------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------- |
687| Rebuilding the top-level `system` prompt | A [mid-conversation system message](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#new-instructions) | None |
688| Injecting a reminder and deleting it on the next request | A [turn-scoped system message](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#per-turn-reminders) (`clear_at: "next_user_message"`) | `mid-conversation-system-clear-at-2026-08-21` |
689| Adding or removing entries in `tools` | [`tool_addition` and `tool_removal` blocks](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#tool-changes) | `mid-conversation-tool-changes-2026-07-01` |
690| Changing top-level `output_config.effort` (restarts the cache, doesn't affect thinking) | A [per-message `output_config`](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#effort-changes) | `mid-conversation-output-config-2026-07-01` |
691| Dropping or summarizing old turns on the client | Server-side [compaction or context editing](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#server-side-trimming), or [client-side compaction](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#custom-compaction-on-the-client) that keeps no stale thinking | `compact-2026-01-12` or `context-management-2025-06-27` |
692| An image or document URL whose bytes change between requests | A [`file_id` from the Files API](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#files-by-id), or base64 | None |
789| Instead of | Use | Beta header |
790| --------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------- |
791| Rebuilding the top-level `system` prompt | A [mid-conversation system message](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#new-instructions) | None |
792| Re-rendering the context in your first user message (environment, date, memory, project instructions) on each request | Render it once and resend it unchanged. When something changes, [put the new version in the newest turn](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#changing-context) | None |
793| Clearing or shortening old `tool_result` content, or re-encoding old images, in place | Shorten a tool result or downscale an image before the first time you send it, not after. To clear old results later, [trim context on the server](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#server-side-trimming) with `clear_tool_uses_20250919` | `context-management-2025-06-27` |
794| Injecting a reminder and deleting it on the next request | A [turn-scoped system message](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#per-turn-reminders) (`clear_at: "next_user_message"`) | `mid-conversation-system-clear-at-2026-08-21` |
795| Adding or removing entries in `tools` | [`tool_addition` and `tool_removal` blocks](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#tool-changes) | `mid-conversation-tool-changes-2026-07-01` |
796| Changing top-level `output_config.effort` (restarts the cache, doesn't affect thinking) | A [per-message `output_config`](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#effort-changes) | `mid-conversation-output-config-2026-07-01` |
797| Dropping or summarizing old turns on the client | Server-side [compaction or context editing](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#server-side-trimming), or [client-side compaction](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#custom-compaction-on-the-client) that keeps no stale thinking | `compact-2026-01-12` or `context-management-2025-06-27` |
798| An image or document URL whose bytes change between requests | A [`file_id` from the Files API](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#files-by-id), or base64 | None |
693799
694All of these assume you [send assistant turns back exactly as returned](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#append-assistant-turns-exactly-as-returned). To use several betas in one request, combine the values in one `anthropic-beta` header. The same names apply on Amazon Bedrock and Google Cloud (see [Beta headers](https://platform.claude.com/docs/en/api/beta-headers)):
800All of these assume you [send assistant turns back exactly as returned](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#append-assistant-turns-exactly-as-returned). Mid-conversation system messages, turn-scoped system messages, and tool changes aren't available on every model: [Mid-conversation system messages and tool changes](https://platform.claude.com/docs/en/build-with-claude/mid-conversation-system-messages) lists the models that accept them. If your code serves several models, keep editing the top-level `system` prompt for the models that don't accept them.
695801
802To use several betas in one request, combine the values in one `anthropic-beta` header. The same names apply on Amazon Bedrock and Google Cloud (see [Beta headers](https://platform.claude.com/docs/en/api/beta-headers)):
803
696804```text wrap
697805anthropic-beta: thinking-binding-controls-2026-08-01,mid-conversation-system-clear-at-2026-08-21,mid-conversation-tool-changes-2026-07-01
698806```
from line 809
701809
702810Store the `content` array from each response and send it back unchanged as the assistant turn: every block type, in the order received, including `thinking` blocks whose `thinking` field is empty. A serializer that drops unknown block types, drops empty fields, or reorders blocks edits the prefix for every later turn.
703811
812On Claude Fable 5.1, the `thinking` field is empty by default and the `signature` carries the reasoning, so a serializer that skips empty blocks removes thinking. If it removes all of them, nothing fails and the model loses its earlier reasoning on every turn. If you parse the stream yourself, keep the block even when no thinking text arrives: it opens, receives its `signature` in a `signature_delta` event, and closes. A block sent back with an empty `signature` fails.
813
704814### Add instructions with a mid-conversation system message
705815
706816Some harnesses rebuild the top-level `system` prompt on each request to carry the current time, a token budget, a mode flag, or newly discovered project context. That invalidates every thinking block in the conversation. Instead, freeze `system` at session start. When something changes, append a [`role: "system"` message](https://platform.claude.com/docs/en/build-with-claude/mid-conversation-system-messages) at the point in `messages` where the change becomes true:
from line 824
714824
715825The model treats this message with system-prompt authority, and everything before it stays unchanged. In a tool loop, place the message after the `tool_result` user message, never between an assistant `tool_use` and its `tool_result` (see [Limitations](https://platform.claude.com/docs/en/build-with-claude/mid-conversation-system-messages#limitations)). Once sent, the message is part of the prefix for later thinking: leave it in place on later requests.
716826
827### Put changing context in the newest turn
828
829Some harnesses put an environment block in the first user message (working directory, branch, date, memory, project instructions) and render it again on every request. When any value changes, `messages[0]` changes, and every thinking block in the conversation is invalid. Render that block once and resend it as it was. When a value changes, say so in the newest turn: add a text block to the user message you are about to send, or append a [mid-conversation system message](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#new-instructions) if the change comes from you as the operator.
830
831```json
832{
833 "role": "user",
834 "content": [
835 {
836 "type": "text",
837 "text": "Environment update: the current branch is now release-2."
838 },
839 { "type": "text", "text": "Run the tests again." }
840 ]
841}
842```
843
844Once sent, that text block is part of the prefix for later thinking: leave it in place on later requests.
845
717846### Send per-turn reminders as turn-scoped system messages
718847
719The most common prefix edit is the per-turn nudge: a line such as "request independent reads together" or "you haven't updated the user in a while" that your code appends after each batch of tool results. To keep reminders from piling up, send each nudge as a [mid-conversation system message](https://platform.claude.com/docs/en/build-with-claude/mid-conversation-system-messages) with `clear_at: "next_user_message"`, placed after the `tool_result` user message. `clear_at` requires the beta header `mid-conversation-system-clear-at-2026-08-21`. The following `messages` array is the request after two tool calls and their results. `messages[3]` is the previous request's nudge, left in place, and `messages[6]` is this request's copy:
848A common prefix edit is the per-turn nudge: a line such as "request independent reads together" or "you haven't updated the user in a while" that your code appends after each batch of tool results. To keep reminders from piling up, send each nudge as a [mid-conversation system message](https://platform.claude.com/docs/en/build-with-claude/mid-conversation-system-messages) with `clear_at: "next_user_message"`, placed after the `tool_result` user message. `clear_at` requires the beta header `mid-conversation-system-clear-at-2026-08-21`. The following `messages` array is the request after two tool calls and their results. `messages[3]` is the previous request's nudge, left in place, and `messages[6]` is this request's copy:
720849
721850```json
722851[
from line 939
810939
811940### Trim context on the server
812941
813The second most common prefix edit is client-side trimming: dropping or summarizing the oldest turns and keeping the recent ones verbatim. The kept turns' thinking blocks were produced while the removed history was still in place, so they fail the check. The server-side equivalents don't count as edits, because the check compares the conversation as you sent it:
942Another common prefix edit is client-side trimming: dropping or summarizing the oldest turns and keeping the recent ones verbatim. The kept turns' thinking blocks were produced while the removed history was still in place, so they fail the check. The server-side equivalents don't count as edits, because the check compares the conversation as you sent it:
814943
815944* [Compaction](https://platform.claude.com/docs/en/build-with-claude/compaction) summarizes older turns into a compaction block when the context approaches a threshold you set, and the checked prefix restarts from that block. Its [`instructions` parameter](https://platform.claude.com/docs/en/build-with-claude/compaction#custom-summarization-instructions) takes your own summarization prompt, such as "preserve every ticker, position size, and stated assumption".
816945* [Context editing](https://platform.claude.com/docs/en/build-with-claude/context-editing) clears old tool results or old thinking blocks by rule, oldest first. The strategies are `clear_tool_uses_20250919` and `clear_thinking_20251015`.
from line 1194
10651194
10661195For an `image` or `document` block with a `url` source, the check covers the fetched bytes, not the URL string. A URL whose content changes invalidates later thinking: a "latest screenshot" endpoint, or a document someone edits between turns. A rotating signed URL for the same file doesn't. For content you reference across turns, upload it once with the [Files API](https://platform.claude.com/docs/en/build-with-claude/files) and use the `file_id`, or send base64.
10671196
1197### Libraries, proxies, and gateways
1198
1199A library, proxy, or gateway sits between someone else's history and the API, so its own rewrites count as edits, and its users can't see or fix them.
1200
1201* **Pass through what you don't recognize.** Forward the caller's `anthropic-beta` values and `thinking.block_binding` unchanged, and return `input_transformations` to them. An options schema that rejects unknown keys stops your users from choosing `"drop_block"`.
1202* **Leave a `role: "system"` message where the caller put it.** Moving it into the top-level `system` field changes `system` on that request and invalidates every thinking block in the conversation.
1203* **To turn tool use off for a request, send `tool_choice: {"type": "none"}`.** Don't remove `tools`.
1204* **Don't hide the 400.** If your code catches it, strips thinking, and retries on the caller's behalf, log that it did: their history is still edited, and the model loses its earlier reasoning on every later request.
1205
10681206## FAQ
10691207
10701208<AccordionGroup>
from line 1231
10931231 </Accordion>
10941232
10951233 <Accordion title="Can I resume a saved session later, after a restart or the next day?">
1096 Yes. A resumed session is an ordinary follow-up request: `system`, `tools`, and the earlier `messages` must match what you last sent byte-for-byte. Persist exactly what you sent and received, and replay that: the rendered system prompt, the tool definitions, and each assistant turn as returned. Don't re-render from inputs that might have changed since, such as the date, an updated instruction file, or a new tool version. Anything new goes in an appended message. See [Send assistant turns back exactly as returned](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#append-assistant-turns-exactly-as-returned).
1234 Yes. A resumed session is an ordinary follow-up request: `system`, `tools`, and the earlier `messages` must have the same content as what you last sent. JSON formatting and key order don't matter; the values do. Persist exactly what you sent and received, and replay that: the rendered system prompt, the tool definitions, and each assistant turn as returned. Don't re-render from inputs that might have changed since, such as the date, an updated instruction file, or a new tool version. Anything new goes in an appended message. See [Send assistant turns back exactly as returned](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#append-assistant-turns-exactly-as-returned).
1235 </Accordion>
1236
1237 <Accordion title="A saved session now fails on every request. How do I get it working again?">
1238 The stored history has an edit in it, so replaying it can't succeed. Send that session with `prefix_mismatch_behavior: "drop_block"` from now on, or remove its `thinking` and `redacted_thinking` blocks once and continue. Thinking the model produces from that point on stays valid as long as nothing before it changes again. Then find the edit so that new sessions don't hit it. See [Handle the error in code](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#handle-the-error-in-code).
10971239 </Accordion>
10981240
10991241 <Accordion title="My harness can route a turn to a non-Claude model. Do those turns invalidate Claude's earlier thinking?">
build-with-claude/thinking Changed · +6 / -3 lines
from line 470
470470* If you pass thinking blocks back in multi-turn conversations, pass them unchanged. The server decrypts the `signature` to reconstruct the original thinking for prompt construction (see [Preserving thinking blocks](https://platform.claude.com/docs/en/build-with-claude/thinking#preserving-thinking-blocks)). Any text you place in the `thinking` field of a round-tripped omitted block is ignored.
471471* `display` is invalid with `thinking.type: "disabled"` (there is nothing to display).
472472* When using `thinking.type: "adaptive"` and the model skips thinking for a simple request, no thinking block is produced regardless of `display`.
473* When streaming with `display: "omitted"`, no `thinking_delta` events are emitted. With `display: "updates"`, only [progress-update blocks](https://platform.claude.com/docs/en/build-with-claude/thinking#progress-updates) stream `thinking_delta` events. See [Streaming thinking](https://platform.claude.com/docs/en/build-with-claude/thinking#streaming-thinking) for the event sequence.
473* When streaming with `display: "omitted"`, no thinking text is streamed. Each thinking block streams a `thinking_delta` with an empty `thinking` string, then its `signature_delta`. With `display: "updates"`, only [progress-update blocks](https://platform.claude.com/docs/en/build-with-claude/thinking#progress-updates) stream `thinking_delta` events that carry text. See [Streaming thinking](https://platform.claude.com/docs/en/build-with-claude/thinking#streaming-thinking) for the event sequence.
474474
475475<Note>
476476 The `signature` field is identical whichever `display` value you set. Switching `display` values between turns in a conversation is supported.
from line 500
500500
501501Thinking works with [streaming](https://platform.claude.com/docs/en/build-with-claude/streaming). Thinking blocks stream as `thinking_delta` events inside `content_block_delta` events, followed by a single `signature_delta` event just before the block's `content_block_stop`. Text blocks stream afterward as usual.
502502
503
503
504504
505505The following examples stream a response with adaptive thinking, printing thinking and text deltas as they arrive:
506506
from line 780
780780 ```
781781</Accordion>
782782
783When `display: "omitted"` is set, the thinking block opens, a single `signature_delta` arrives, and the block closes without any `thinking_delta` events. Text streaming begins immediately after:
783When `display: "omitted"` is set, the thinking block opens, a `thinking_delta` with an empty `thinking` string arrives, a single `signature_delta` follows, and the block closes. Text streaming begins immediately after:
784784
785785```sse Output
786786event: content_block_start
787787data: {"type":"content_block_start","index":0,"content_block":{"type":"thinking","thinking":"","signature":""}}
788
789event: content_block_delta
790data: {"type":"content_block_delta","index":0,"delta":{"type":"thinking_delta","thinking":""}}
788791
789792event: content_block_delta
790793data: {"type":"content_block_delta","index":0,"delta":{"type":"signature_delta","signature":"EosnCkYICxIMMb3LzNrMu..."}}
build-with-claude/thinking-troubleshooting Changed · +8 / -2 lines
from line 106
106106messages.{i}.content.{j}: Invalid `signature` in `thinking` block. The block is bound to a different conversation. Remove the block, or set `thinking.block_binding.prefix_mismatch_behavior` to "drop_block".
107107```
108108
109If the request didn't send the `thinking-binding-controls-2026-08-01` beta header, the message adds ``That setting requires the `thinking-binding-controls-2026-08-01` value in the `anthropic-beta` header.`` The message can also end with a sentence naming the first message that changed. If the message has no reason clause at all, the block's content was modified. See [A 400 error says thinking blocks cannot be modified](https://platform.claude.com/docs/en/build-with-claude/thinking-troubleshooting#error-thinking-blocks-modified).
109If the request didn't send the `thinking-binding-controls-2026-08-01` beta header, the message adds ``That setting requires the `thinking-binding-controls-2026-08-01` value in the `anthropic-beta` header.``
110110
111On Claude Fable 5.1, the API accepts a replayed thinking block [only while the `system` prompt, `tools`, and messages that preceded it are unchanged](https://platform.claude.com/docs/en/build-with-claude/thinking#preserved-in-conversation). The error means something earlier in the conversation changed between requests: an edited, reordered, or removed turn, a per-turn reminder that was injected and later removed, a rebuilt `system` prompt or `tools` array, or client-side compaction that kept recent turns and their thinking verbatim. The check is enforced for new accounts created on or after August 31, 2026, and for any request that sets `thinking.block_binding.prefix_mismatch_behavior`. Server-side [compaction](https://platform.claude.com/docs/en/build-with-claude/compaction) and [context editing](https://platform.claude.com/docs/en/build-with-claude/context-editing) never trigger it.
111The message usually ends with a sentence naming what changed: the `system` prompt, the `tools` list, the first message or block that differs, content that is missing or new, or an earlier thinking block that is missing or out of order. That sentence is for people and logs. Its wording can change, so don't match on it in code.
112112
113If the message stops after ``Invalid `signature` in `thinking` block``, the signature itself didn't verify: it was truncated, altered, or sent back empty, and `prefix_mismatch_behavior` doesn't apply. Edited thinking text returns a different error. See [A 400 error says thinking blocks cannot be modified](https://platform.claude.com/docs/en/build-with-claude/thinking-troubleshooting#error-thinking-blocks-modified).
114
115On Claude Fable 5.1, the API accepts a replayed thinking block only while the `system` prompt, `tools`, and messages that preceded it are unchanged. See [Keeping the prefix unchanged](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#prefix-check). The error means something earlier in the conversation changed between requests: an edited, reordered, or removed turn, a per-turn reminder that was injected and later removed, a rebuilt `system` prompt or `tools` array, or client-side compaction that kept recent turns and their thinking verbatim. The check is enforced for new accounts created on or after August 31, 2026, and for any request that sets `thinking.block_binding.prefix_mismatch_behavior`. Server-side [compaction](https://platform.claude.com/docs/en/build-with-claude/compaction) and [context editing](https://platform.claude.com/docs/en/build-with-claude/context-editing) never trigger it.
116
113117To fix it, keep the history append-only: pass earlier turns back exactly as sent and received, add instructions with a [mid-conversation system message](https://platform.claude.com/docs/en/build-with-claude/mid-conversation-system-messages) instead of editing `system` or `tools`, and let server-side [context editing](https://platform.claude.com/docs/en/build-with-claude/context-editing) or [compaction](https://platform.claude.com/docs/en/build-with-claude/compaction) do any trimming. Retrying the same request body doesn't clear the error. To continue this request without the invalidated reasoning, send the `thinking-binding-controls-2026-08-01` beta header and set `thinking.block_binding.prefix_mismatch_behavior` to `"drop_block"`. Alternatively, strip every `thinking` and `redacted_thinking` block from the history (at minimum the named block and every one after it, in that turn and all later turns), leave each turn's other blocks in place, and retry once.
114118
115119A block from a model the target model can't read never produces this error: the API drops it and, under the beta header, reports it in `input_transformations`.
from line 125
121125This happens because `display` defaults to `"omitted"` on newer models, which returns thinking blocks without their text.
122126
123127Set `display: "summarized"` in your thinking configuration to receive the summarized thinking text. See [Controlling thinking display](https://platform.claude.com/docs/en/build-with-claude/thinking#controlling-thinking-display) for the defaults per model. If you only want the short status lines some models write between tool calls, and not the reasoning, set `display: "updates"` (beta) instead. See [Progress updates between tool calls](https://platform.claude.com/docs/en/build-with-claude/thinking#progress-updates).
128
129A block whose `thinking` field is empty is still complete: the `signature` holds the reasoning. Send it back with the turn like any other. See [Send assistant turns back exactly as returned](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#append-assistant-turns-exactly-as-returned).
124130
125131## No thinking block appears on some turns
126132
manage-claude/inference-hooks-configuration Changed · +3 / -3 lines
from line 101
101101
102102Under **Exclusions**, select roles whose members are not covered by Inference hooks: their prompts are never sent to your AI security server. Only custom roles your organization created can be excluded; the built-in roles aren't offered. Pick them in the role selector, whose placeholder reads **Select roles to exclude**, and manage who holds each role from the roles admin page (**Manage roles**); changing exclusions requires identity management permission. The list is empty by default, and with no roles excluded, every governed request is inspected.
103103
104Exclusion applies to a user's interactive sessions; traffic authenticated by machine credentials is always inspected. If Claude can't resolve a requester's role membership, the request fails closed with a retryable error rather than proceeding uninspected. Changes to the exclusion list are recorded in the audit trail.
104Exclusion applies to a user's interactive sessions; traffic authenticated by machine credentials is always inspected. Changes to the exclusion list are recorded in the audit trail.
105105
106106## Custom blocked prompt message
107107
from line 127
127127
128128To recover, fix the server, then turn **Enforce verdicts** back on to reset the breaker.
129129
130The breaker can also reset on its own. Starting 10 minutes after the trip, Anthropic tests whether your server has recovered: at most about once per minute, one request from your organization's normal traffic is sent to your server for inspection, and that request proceeds for its user whether or not your server answers. If your server responds with a valid verdict, allow or deny, the breaker resets and enforcement resumes. Any other outcome is a webhook failure: the breaker stays tripped and testing continues.
130The breaker can also reset on its own. Starting 10 minutes after the trip, Anthropic checks whether your server has recovered by sending it a test request in the background, at most about once per minute; no user request is involved. If your server responds with a valid verdict, allow or deny, the breaker resets and enforcement resumes. Otherwise the breaker stays tripped and the checks continue.
131131
132Automatic recovery runs only while your Inference hooks settings are unchanged since the trip. If you change any Inference hooks setting after a trip, including rotating the signing secret, testing stops and the breaker no longer resets on its own; turn **Enforce verdicts** back on when your server is fixed. Automatic recovery applies only to trips: if you turn **Enforce verdicts** off yourself, enforcement stays off until you turn it back on.
132Automatic recovery runs only while your Inference hooks settings are unchanged since the trip. If you change any Inference hooks setting after a trip, including rotating the signing secret, the checks stop and the breaker no longer resets on its own; turn **Enforce verdicts** back on when your server is fixed. Automatic recovery applies only to trips: if you turn **Enforce verdicts** off yourself, enforcement stays off until you turn it back on.
133133
134134## Rotate your signing secret
135135
manage-claude/inference-hooks-endpoint Changed · +5 / -7 lines
from line 176
176176| `model` | string or null | Public model identifier for this request, when available. |
177177| `metadata` | object | Reserved extension map of string keys to string values, sent empty today. Require nothing from it, and tolerate its absence, its presence, and any keys that appear. |
178178
179<Note>
180 Requests currently also carry deprecated legacy aliases of some of these fields. Read the field names documented on this page and ignore any others; the aliases exist only for earlier integrations.
181</Note>
182
183179An example request body:
184180
185181```json
from line 226
230226| `tool_result` | `content`: the tool's output as text, with parts joined by newlines; binary parts such as images are replaced by placeholder markers, and raw bytes are never sent. `is_error`: whether the tool call failed. `tool_name`: the tool's name, so a policy can condition on tool identity without cross-referencing an earlier block. `tool_use_id`: the `id` of the matching `tool_use` block. |
231227| `attachment` | `file_name`: the original file name or path. `media_type`: the attachment's media type. `size_bytes`: the size of the original file. `text`: the text content of the attachment when available, such as extracted document text, an audio transcript, or link metadata. Raw attachment bytes are never sent. |
232228
229Apart from `type`, a `text` block's `text`, and a `tool_result` block's `content` and `is_error`, any of these fields can be `null` when the value isn't known; for example, an image arrives as an `attachment` block with `file_name` and `text` set to `null`.
230
233231A block whose `type` you don't recognize is a forward-compatible addition. The only field it guarantees is `type`; your policy may inspect whatever other fields are present, but must not reject the request because of an unrecognized type.
234232
235233### What the transcript contains
from line 236
238236
239237A turn whose every block is excluded is omitted entirely, so don't assume strict user and assistant alternation.
240238
241Transcripts are sent untruncated, so a long conversation with large attachments produces a large request body, up to an upper bound of 10 MB. Raise your server's body limit to accept that ceiling. Several common defaults are much smaller, including nginx `client_max_body_size` at 1 MB and Express `express.json()` at 100 kB, and a rejected body counts as a webhook failure, so under **Allow the request** failure handling an oversized prompt would reach the model uninspected.
239Transcripts are sent untruncated, so a long conversation with large attachments produces a large request body. In practice the model's context window keeps bodies under about 10 MB, but the protocol allows up to 64 MiB, so raise your server's body limit to accept that ceiling. Several common defaults are much smaller, including nginx `client_max_body_size` at 1 MB and Express `express.json()` at 100 kB, and a rejected body counts as a webhook failure, so under **Allow the request** failure handling an oversized prompt would reach the model uninspected.
242240
243241### Source values
244242
245`source.application` is an open string, not a closed enum. Known values are `claude-ai` and `claude-code`; [connection tests](https://platform.claude.com/docs/en/manage-claude/inference-hooks-configuration) use `config-test`. New values may appear, and your server must not reject a request because of one it doesn't recognize.
243`source.application` is an open string, not a closed enum. Common values are `claude-ai`, `claude-code`, and `cowork`; [connection tests](https://platform.claude.com/docs/en/manage-claude/inference-hooks-configuration) and automatic circuit-breaker [recovery checks](https://platform.claude.com/docs/en/manage-claude/inference-hooks-endpoint#circuit-breaker) use `config-test`. New values may appear, and your server must not reject a request because of one it doesn't recognize.
246244
247245Treat `source.application` as advisory routing metadata, not a trust boundary: don't rest a security-critical policy decision on it alone.
248246
from line 687
689687
690688Sustained webhook failures attributable to your AI security server trip a circuit breaker that stops enforcement: Anthropic stops contacting your server, and failure handling applies to every request.
691689
692Starting 10 minutes after the trip, Anthropic tests whether your server has recovered: at most about once per minute, one request, carried by your organization's own traffic, is delivered to your server for inspection, signed and shaped like any other. Respond to it normally. A valid verdict, allow or deny, resets the breaker and enforcement resumes. A webhook failure leaves the breaker tripped, and testing continues. Either way, the test request itself proceeds for its user: its verdict is not enforced, and a failed test does not block it, even under **Block the request**. An administrator can also reset the breaker at any time, and administrator configuration changes stop the automatic testing; see [Circuit breaker](https://platform.claude.com/docs/en/manage-claude/inference-hooks-configuration#circuit-breaker).
690Starting 10 minutes after the trip, Anthropic checks whether your server has recovered: at most about once per minute it sends your server the same synthetic test request that **Test connection** sends (`source.application` is `config-test`), signed like any other request and carrying no user content. Respond to it normally. A valid verdict, allow or deny, resets the breaker and enforcement resumes; a webhook failure leaves the breaker tripped, and the checks continue. An administrator can also reset the breaker at any time, and administrator configuration changes stop the automatic checks; see [Circuit breaker](https://platform.claude.com/docs/en/manage-claude/inference-hooks-configuration#circuit-breaker).
693691
694692Each trip is recorded as an `inference_hooks_circuit_breaker_tripped` activity in the [Activity Feed](https://platform.claude.com/docs/en/manage-claude/compliance-activity-feed), one activity per trip. While the breaker is tripped, no per-request Inference hooks activities are recorded, so the trip activity is the feed's only record of the tripped window.
695693
api/beta/organization/rbac_roles Changed · +1 / -1 lines
from line 257
257257 `all_connectors` grants carry a tool-access action, the scope action, or
258258 an authentication-method action (`interactive` or `managed`).
259259
260 - `resource: object or object or object or 2 more`
260 - `resource: Organization or ConnectorTool or ConnectorScope or 2 more`
261261
262262 What the permission applies to.
263263
api/beta/organization/rbac_roles/permissions Changed · +2 / -2 lines
from line 64
6464 `all_connectors` grants carry a tool-access action, the scope action, or
6565 an authentication-method action (`interactive` or `managed`).
6666
67 - `resource: object or object or object or 2 more`
67 - `resource: Organization or ConnectorTool or ConnectorScope or 2 more`
6868
6969 What the permission applies to.
7070
from line 216
216216 `all_connectors` grants carry a tool-access action, the scope action, or
217217 an authentication-method action (`interactive` or `managed`).
218218
219 - `resource: object or object or object or 2 more`
219 - `resource: Organization or ConnectorTool or ConnectorScope or 2 more`
220220
221221 What the permission applies to.
222222
api/beta/organization/rbac_roles/permissions/list Changed · +1 / -1 lines
from line 62
6262 `all_connectors` grants carry a tool-access action, the scope action, or
6363 an authentication-method action (`interactive` or `managed`).
6464
65 - `resource: object or object or object or 2 more`
65 - `resource: Organization or ConnectorTool or ConnectorScope or 2 more`
6666
6767 What the permission applies to.
6868
api/beta/organization/spend_limits/create Changed · +1 / -1 lines
from line 81
8181
8282 - `"weekly"`
8383
84 - `scope: object or object or object or 2 more`
84 - `scope: User or SeatTier or RBACGroup or 2 more`
8585
8686 What the limit applies to. A tagged union on `type`; each variant carries the identifier for its scope.
8787
api/beta/organization/spend_limits/increase_requests/deny Changed · +2 / -2 lines
from line 78
7878
7979 format: date-time
8080
81 - `resolved_by: object or object or null`
81 - `resolved_by: UserActor or ScopedAPIKeyActor or null`
8282
8383 - `UserActor object`
8484
from line 190
190190
191191 Tagged ID of the member the spend limit applies to.
192192
193 - `source: object or object or object or 2 more`
193 - `source: User or SeatTier or RBACGroup or 2 more`
194194
195195 - `User object`
196196
api/beta/organization/spend_limits/increase_requests/list Changed · +2 / -2 lines
from line 92
9292
9393 format: date-time
9494
95 - `resolved_by: object or object or null`
95 - `resolved_by: UserActor or ScopedAPIKeyActor or null`
9696
9797 - `UserActor object`
9898
from line 204
204204
205205 Tagged ID of the member the spend limit applies to.
206206
207 - `source: object or object or object or 2 more`
207 - `source: User or SeatTier or RBACGroup or 2 more`
208208
209209 - `User object`
210210
api/beta/organization/spend_limits/increase_requests/retrieve Changed · +2 / -2 lines
from line 74
7474
7575 format: date-time
7676
77 - `resolved_by: object or object or null`
77 - `resolved_by: UserActor or ScopedAPIKeyActor or null`
7878
7979 - `UserActor object`
8080
from line 186
186186
187187 Tagged ID of the member the spend limit applies to.
188188
189 - `source: object or object or object or 2 more`
189 - `source: User or SeatTier or RBACGroup or 2 more`
190190
191191 - `User object`
192192
api/beta/organization/spend_limits/list_effective Changed · +1 / -1 lines
from line 113
113113
114114 Tagged ID of the member the spend limit applies to.
115115
116 - `source: object or object or object or 2 more`
116 - `source: User or SeatTier or RBACGroup or 2 more`
117117
118118 - `User object`
119119