Follow Discord
Sweep 25 Sep 2026 · 19:33Z Build v2.1.283 504 read Stable v2.1.274 Latest v2.1.283 Next v2.1.283 Feeds RSS JSON llms.txt llms-full.txt Unofficial
One change · api

Sessions changedapi/beta/sessions

Nearest release: v2.1.283, published 3 hours after this site recorded the change. Shown because the two are within 24 hours of each other. Nothing here says the release caused the edit.

Recorded here
Lines+10,002added
Lines−10,010removed
From line 183 where the diff opens
First seen 14 Aug 2026 this site's first read of the page
Recorded edits14to this page, all time

The whole hunk

from line 183, old and new numbered
/
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 183
183183 
184184 Replacement model. Accepts the model string, e.g. `claude-opus-5`, or a `model_config` object. Omit to use the agent's model.
185185 
186 - `BetaManagedAgentsModel = "claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more or string`
186 - `BetaManagedAgentsModel = string or "claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more`
187187 
188188 The model that will power your agent.
189189 
190190 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
191191 
192 - `string`
193 
192194 - `"claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more`
193195 
194196 The model that will power your agent.
from line 257
255257 
256258 High-performance model for agents and coding
257259 
258 - `string`
259 
260260 - `BetaManagedAgentsModelConfigParams object`
261261 
262262 An object that defines additional configuration control over model use
from line 331
331331 
332332 - `speed: optional "standard" or "fast" or null`
333333 
334 Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
334 Inference speed mode. Defaults to `standard`.
335335 
336336 - `"standard"`
337337 
from line 413
413413 
414414 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
415415 
416 Permission policy for tool execution.
416 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
417417 
418418 - `BetaManagedAgentsAlwaysAllowPolicy object`
419419 
from line 449
449449 
450450 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
451451 
452 Permission policy for tool execution.
452 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
453453 
454454 - `BetaManagedAgentsAlwaysAllowPolicy object`
455455 
from line 479
479479 
480480 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
481481 
482 Permission policy for tool execution.
482 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
483483 
484484 - `BetaManagedAgentsAlwaysAllowPolicy object`
485485 
from line 509
509509 
510510 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
511511 
512 Permission policy for tool execution.
512 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
513513 
514514 - `BetaManagedAgentsAlwaysAllowPolicy object`
515515 
from line 539
539539 
540540 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
541541 
542 Permission policy for tool execution.
542 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
543543 
544544 - `BetaManagedAgentsAlwaysAllowPolicy object`
545545 
from line 569
569569 
570570 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
571571 
572 Permission policy for tool execution.
572 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
573573 
574574 - `BetaManagedAgentsAlwaysAllowPolicy object`
575575 
from line 613
613613 
614614 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
615615 
616 Permission policy for tool execution.
616 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
617617 
618618 - `BetaManagedAgentsAlwaysAllowPolicy object`
619619 
from line 651
651651 
652652 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
653653 
654 Permission policy for tool execution.
654 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
655655 
656656 - `BetaManagedAgentsAlwaysAllowPolicy object`
657657 
from line 697
697697 
698698 - `default_config: optional BetaManagedAgentsAgentToolsetDefaultConfigParams or null`
699699 
700 Default configuration for all tools in a toolset.
700 Default configuration applied to all tools in this set.
701701 
702702 - `enabled: optional boolean or null`
703703 
from line 705
705705 
706706 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
707707 
708 Permission policy for tool execution.
708 Default permission policy for tools. Controls whether tool calls are auto-approved or require confirmation.
709709 
710710 - `BetaManagedAgentsAlwaysAllowPolicy object`
711711 
from line 747
747747 
748748 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
749749 
750 Permission policy for tool execution.
750 Permission policy for this tool. Overrides the `default_config` setting.
751751 
752752 - `BetaManagedAgentsAlwaysAllowPolicy object`
753753 
from line 763
763763 
764764 - `default_config: optional BetaManagedAgentsMCPToolsetDefaultConfigParams or null`
765765 
766 Default configuration for all tools from an MCP server.
766 Default configuration for all tools from this server.
767767 
768768 - `enabled: optional boolean or null`
769769 
from line 771
771771 
772772 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
773773 
774 Permission policy for tool execution.
774 Default permission policy for tools from this server.
775775 
776776 - `BetaManagedAgentsAlwaysAllowPolicy object`
777777 
from line 799
799799 
800800 - `input_schema: BetaManagedAgentsCustomToolInputSchema`
801801 
802 JSON Schema for custom tool input parameters.
802 JSON Schema defining the expected input parameters for the tool.
803803 
804804 - `type: "object"`
805805 
from line 827
827827 
828828- `budget: optional BetaManagedAgentsBudgetLimit`
829829 
830 A hard spend ceiling. The session stops issuing new model requests once the tracked list cost reaches `max_list_cost`.
830 Enforced spend ceiling for the session. Omit to create an uncapped session. Every model the session can run — the agent's model and each callable agent's model — must have a public list price, or the request is rejected with reason `model_not_budgetable`.
831831 
832832 - `type: "limit"`
833833 
834834 - `max_list_cost: BetaMonetaryAmount`
835835 
836 A monetary amount in a specific currency.
836 Maximum list cost the session may accrue. List price is used regardless of any negotiated discount, so the cap fires at or before the actual charge.
837837 
838838 - `amount: string`
839839 
from line 877
877877 
878878 - `source: BetaManagedAgentsBase64ImageSource or BetaManagedAgentsURLImageSource or BetaManagedAgentsFileImageSource`
879879 
880 Union type for image source variants.
880 The source of the image data.
881881 
882882 - `BetaManagedAgentsBase64ImageSource object`
883883 
from line 929
929929 
930930 - `source: BetaManagedAgentsBase64DocumentSource or BetaManagedAgentsPlainTextDocumentSource or BetaManagedAgentsURLDocumentSource or BetaManagedAgentsFileDocumentSource`
931931 
932 Union type for document source variants.
932 The source of the document data.
933933 
934934 - `BetaManagedAgentsBase64DocumentSource object`
935935 
from line 1015
10151015 
10161016 - `rubric: BetaManagedAgentsFileRubricParams or BetaManagedAgentsTextRubricParams`
10171017 
1018 Rubric for grading the quality of an outcome.
1018 How to grade the outcome. Text or file reference.
10191019 
10201020 - `BetaManagedAgentsFileRubricParams object`
10211021 
from line 1131
11311131 
11321132 - `access: optional "read_write" or "read_only" or null`
11331133 
1134 Access mode for an attached memory store.
1134 Access mode for the mounted store. Defaults to read_write. read_only mounts the store as a read-only filesystem.
11351135 
11361136 - `"read_write"`
11371137 
from line 1191
11911191 
11921192 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
11931193 
1194 - `string`
1195 
11941196 - `"claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more`
11951197 
11961198 The model that will power your agent.
from line 1259
12571259 
12581260 High-performance model for agents and coding
12591261 
1260 - `string`
1261 
12621262 - `effort: optional BetaManagedAgentsEffortLow or BetaManagedAgentsEffortMedium or BetaManagedAgentsEffortHigh or 2 more`
12631263 
1264 How hard Claude works on each turn. Sets `output_config.effort` on every Messages call the session makes.
1264 How hard Claude works on each inference call. One of `low`, `medium`, `high`, `xhigh`, `max`. Always present; resolved to the per-model default at save time when not supplied.
12651265 
12661266 - `BetaManagedAgentsEffortLow object`
12671267 
from line 1299
12991299 
13001300 - `speed: optional "standard" or "fast"`
13011301 
1302 Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
1302 Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Defaults to `standard`. Not all models support `fast`; invalid combinations are rejected at create time.
13031303 
13041304 - `"standard"`
13051305 
from line 1307
13071307 
13081308 - `multiagent: BetaManagedAgentsSessionMultiagentCoordinator or null`
13091309 
1310 Resolved coordinator topology with full agent definitions for each roster member.
1310 Resolved multiagent orchestration configuration. Null when the agent is single-threaded.
13111311 
13121312 - `type: "coordinator"`
13131313 
from line 1763
17631763 
17641764 - `archived_at: string or null`
17651765 
1766 A timestamp in RFC 3339 format
1766 When the session was archived. Null if not archived.
17671767 
17681768 format: date-time
17691769 
17701770 - `budget: BetaManagedAgentsBudgetLimit or null`
17711771 
1772 A hard spend ceiling. The session stops issuing new model requests once the tracked list cost reaches `max_list_cost`.
1772 The session's enforced spend ceiling, or null when no budget is set.
17731773 
17741774 - `type: "limit"`
17751775 
17761776 - `max_list_cost: BetaMonetaryAmount`
17771777 
1778 A monetary amount in a specific currency.
1778 Maximum list cost the session may accrue. List price is used regardless of any negotiated discount, so the cap fires at or before the actual charge.
17791779 
17801780 - `amount: string`
17811781 
from line 1803
18031803 
18041804 - `completed_at: string or null`
18051805 
1806 A timestamp in RFC 3339 format
1806 When the outcome reached a terminal result. Null while `pending`/`running`/`evaluating`.
18071807 
18081808 format: date-time
18091809 
from line 1909
19091909 
19101910 - `access: optional "read_write" or "read_only" or null`
19111911 
1912 Access mode for an attached memory store.
1912 Access mode for the mounted store. Defaults to `read_write`. `read_only` mounts the store as a read-only filesystem.
19131913 
19141914 - `"read_write"`
19151915 
from line 1935
19351935 
19361936 - `stats: BetaManagedAgentsSessionStats`
19371937 
1938 Timing statistics for a session.
1938 Timing statistics for the session.
19391939 
19401940 - `active_seconds: optional number`
19411941 
from line 1979
19791979 
19801980 - `usage: BetaManagedAgentsSessionUsage`
19811981 
1982 Cumulative token usage for a session across all turns.
1982 Cumulative token usage for the session.
19831983 
19841984 - `active_seconds: optional number`
19851985 
from line 1989
19891989 
19901990 - `cache_creation: optional BetaManagedAgentsCacheCreationUsage`
19911991 
1992 Prompt-cache creation token usage broken down by cache lifetime.
1992 Tokens used to create prompt cache entries, broken down by cache TTL.
19931993 
19941994 - `ephemeral_1h_input_tokens: optional number`
19951995 
from line 2017
20172017 
20182018 - `list_cost: optional BetaMonetaryAmount or null`
20192019 
2020 A monetary amount in a specific currency.
2020 Cumulative list cost of the session across all turns, priced at public list rates. Absent until cost tracking is available for the session.
20212021 
20222022 - `output_tokens: optional number`
20232023 
from line 2027
20272027 
20282028 - `server_tool_use: optional BetaManagedAgentsServerToolUsage or null`
20292029 
2030 Cumulative count of server-executed tool invocations, broken down by tool.
2030 Cumulative server-executed tool usage across all turns. Absent until server-tool tracking is available for the session.
20312031 
20322032 - `web_fetch_requests: optional number`
20332033 
from line 2500
25002500 
25012501 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
25022502 
2503 - `string`
2504 
25032505 - `"claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more`
25042506 
25052507 The model that will power your agent.
from line 2568
25662568 
25672569 High-performance model for agents and coding
25682570 
2569 - `string`
2570 
25712571 - `effort: optional BetaManagedAgentsEffortLow or BetaManagedAgentsEffortMedium or BetaManagedAgentsEffortHigh or 2 more`
25722572 
2573 How hard Claude works on each turn. Sets `output_config.effort` on every Messages call the session makes.
2573 How hard Claude works on each inference call. One of `low`, `medium`, `high`, `xhigh`, `max`. Always present; resolved to the per-model default at save time when not supplied.
25742574 
25752575 - `BetaManagedAgentsEffortLow object`
25762576 
from line 2608
26082608 
26092609 - `speed: optional "standard" or "fast"`
26102610 
2611 Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
2611 Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Defaults to `standard`. Not all models support `fast`; invalid combinations are rejected at create time.
26122612 
26132613 - `"standard"`
26142614 
from line 2616
26162616 
26172617 - `multiagent: BetaManagedAgentsSessionMultiagentCoordinator or null`
26182618 
2619 Resolved coordinator topology with full agent definitions for each roster member.
2619 Resolved multiagent orchestration configuration. Null when the agent is single-threaded.
26202620 
26212621 - `type: "coordinator"`
26222622 
from line 3072
30723072 
30733073 - `archived_at: string or null`
30743074 
3075 A timestamp in RFC 3339 format
3075 When the session was archived. Null if not archived.
30763076 
30773077 format: date-time
30783078 
30793079 - `budget: BetaManagedAgentsBudgetLimit or null`
30803080 
3081 A hard spend ceiling. The session stops issuing new model requests once the tracked list cost reaches `max_list_cost`.
3081 The session's enforced spend ceiling, or null when no budget is set.
30823082 
30833083 - `type: "limit"`
30843084 
30853085 - `max_list_cost: BetaMonetaryAmount`
30863086 
3087 A monetary amount in a specific currency.
3087 Maximum list cost the session may accrue. List price is used regardless of any negotiated discount, so the cap fires at or before the actual charge.
30883088 
30893089 - `amount: string`
30903090 
from line 3112
31123112 
31133113 - `completed_at: string or null`
31143114 
3115 A timestamp in RFC 3339 format
3115 When the outcome reached a terminal result. Null while `pending`/`running`/`evaluating`.
31163116 
31173117 format: date-time
31183118 
from line 3218
32183218 
32193219 - `access: optional "read_write" or "read_only" or null`
32203220 
3221 Access mode for an attached memory store.
3221 Access mode for the mounted store. Defaults to `read_write`. `read_only` mounts the store as a read-only filesystem.
32223222 
32233223 - `"read_write"`
32243224 
from line 3244
32443244 
32453245 - `stats: BetaManagedAgentsSessionStats`
32463246 
3247 Timing statistics for a session.
3247 Timing statistics for the session.
32483248 
32493249 - `active_seconds: optional number`
32503250 
from line 3288
32883288 
32893289 - `usage: BetaManagedAgentsSessionUsage`
32903290 
3291 Cumulative token usage for a session across all turns.
3291 Cumulative token usage for the session.
32923292 
32933293 - `active_seconds: optional number`
32943294 
from line 3298
32983298 
32993299 - `cache_creation: optional BetaManagedAgentsCacheCreationUsage`
33003300 
3301 Prompt-cache creation token usage broken down by cache lifetime.
3301 Tokens used to create prompt cache entries, broken down by cache TTL.
33023302 
33033303 - `ephemeral_1h_input_tokens: optional number`
33043304 
from line 3326
33263326 
33273327 - `list_cost: optional BetaMonetaryAmount or null`
33283328 
3329 A monetary amount in a specific currency.
3329 Cumulative list cost of the session across all turns, priced at public list rates. Absent until cost tracking is available for the session.
33303330 
33313331 - `output_tokens: optional number`
33323332 
from line 3336
33363336 
33373337 - `server_tool_use: optional BetaManagedAgentsServerToolUsage or null`
33383338 
3339 Cumulative count of server-executed tool invocations, broken down by tool.
3339 Cumulative server-executed tool usage across all turns. Absent until server-tool tracking is available for the session.
33403340 
33413341 - `web_fetch_requests: optional number`
33423342 
from line 3735
37353735 
37363736 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
37373737 
3738 - `string`
3739 
37383740 - `"claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more`
37393741 
37403742 The model that will power your agent.
from line 3803
38013803 
38023804 High-performance model for agents and coding
38033805 
3804 - `string`
3805 
38063806 - `effort: optional BetaManagedAgentsEffortLow or BetaManagedAgentsEffortMedium or BetaManagedAgentsEffortHigh or 2 more`
38073807 
3808 How hard Claude works on each turn. Sets `output_config.effort` on every Messages call the session makes.
3808 How hard Claude works on each inference call. One of `low`, `medium`, `high`, `xhigh`, `max`. Always present; resolved to the per-model default at save time when not supplied.
38093809 
38103810 - `BetaManagedAgentsEffortLow object`
38113811 
from line 3843
38433843 
38443844 - `speed: optional "standard" or "fast"`
38453845 
3846 Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
3846 Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Defaults to `standard`. Not all models support `fast`; invalid combinations are rejected at create time.
38473847 
38483848 - `"standard"`
38493849 
from line 3851
38513851 
38523852 - `multiagent: BetaManagedAgentsSessionMultiagentCoordinator or null`
38533853 
3854 Resolved coordinator topology with full agent definitions for each roster member.
3854 Resolved multiagent orchestration configuration. Null when the agent is single-threaded.
38553855 
38563856 - `type: "coordinator"`
38573857 
from line 4307
43074307 
43084308 - `archived_at: string or null`
43094309 
4310 A timestamp in RFC 3339 format
4310 When the session was archived. Null if not archived.
43114311 
43124312 format: date-time
43134313 
43144314 - `budget: BetaManagedAgentsBudgetLimit or null`
43154315 
4316 A hard spend ceiling. The session stops issuing new model requests once the tracked list cost reaches `max_list_cost`.
4316 The session's enforced spend ceiling, or null when no budget is set.
43174317 
43184318 - `type: "limit"`
43194319 
43204320 - `max_list_cost: BetaMonetaryAmount`
43214321 
4322 A monetary amount in a specific currency.
4322 Maximum list cost the session may accrue. List price is used regardless of any negotiated discount, so the cap fires at or before the actual charge.
43234323 
43244324 - `amount: string`
43254325 
from line 4347
43474347 
43484348 - `completed_at: string or null`
43494349 
4350 A timestamp in RFC 3339 format
4350 When the outcome reached a terminal result. Null while `pending`/`running`/`evaluating`.
43514351 
43524352 format: date-time
43534353 
from line 4453
44534453 
44544454 - `access: optional "read_write" or "read_only" or null`
44554455 
4456 Access mode for an attached memory store.
4456 Access mode for the mounted store. Defaults to `read_write`. `read_only` mounts the store as a read-only filesystem.
44574457 
44584458 - `"read_write"`
44594459 
from line 4479
44794479 
44804480 - `stats: BetaManagedAgentsSessionStats`
44814481 
4482 Timing statistics for a session.
4482 Timing statistics for the session.
44834483 
44844484 - `active_seconds: optional number`
44854485 
from line 4523
45234523 
45244524 - `usage: BetaManagedAgentsSessionUsage`
45254525 
4526 Cumulative token usage for a session across all turns.
4526 Cumulative token usage for the session.
45274527 
45284528 - `active_seconds: optional number`
45294529 
from line 4533
45334533 
45344534 - `cache_creation: optional BetaManagedAgentsCacheCreationUsage`
45354535 
4536 Prompt-cache creation token usage broken down by cache lifetime.
4536 Tokens used to create prompt cache entries, broken down by cache TTL.
45374537 
45384538 - `ephemeral_1h_input_tokens: optional number`
45394539 
from line 4561
45614561 
45624562 - `list_cost: optional BetaMonetaryAmount or null`
45634563 
4564 A monetary amount in a specific currency.
4564 Cumulative list cost of the session across all turns, priced at public list rates. Absent until cost tracking is available for the session.
45654565 
45664566 - `output_tokens: optional number`
45674567 
from line 4571
45714571 
45724572 - `server_tool_use: optional BetaManagedAgentsServerToolUsage or null`
45734573 
4574 Cumulative count of server-executed tool invocations, broken down by tool.
4574 Cumulative server-executed tool usage across all turns. Absent until server-tool tracking is available for the session.
45754575 
45764576 - `web_fetch_requests: optional number`
45774577 
from line 4922
49224922 
49234923- `agent: optional BetaManagedAgentsSessionAgentUpdate`
49244924 
4925 Mid-session agent configuration update. Only `tools` and `mcp_servers` are updatable. Full replacement: the provided array becomes the new value. To preserve existing entries, GET the session, modify the array, and POST it back.
4925 Agent configuration update. Only `tools` and `mcp_servers` are updatable mid-session. Only valid for sessions created from an agent or deployment reference. The session must not be running.
49264926 
49274927 - `mcp_servers: optional array of BetaManagedAgentsURLMCPServerParams`
49284928 
from line 4972
49724972 
49734973 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
49744974 
4975 Permission policy for tool execution.
4975 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
49764976 
49774977 - `BetaManagedAgentsAlwaysAllowPolicy object`
49784978 
from line 5008
50085008 
50095009 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
50105010 
5011 Permission policy for tool execution.
5011 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
50125012 
50135013 - `BetaManagedAgentsAlwaysAllowPolicy object`
50145014 
from line 5038
50385038 
50395039 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
50405040 
5041 Permission policy for tool execution.
5041 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
50425042 
50435043 - `BetaManagedAgentsAlwaysAllowPolicy object`
50445044 
from line 5068
50685068 
50695069 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
50705070 
5071 Permission policy for tool execution.
5071 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
50725072 
50735073 - `BetaManagedAgentsAlwaysAllowPolicy object`
50745074 
from line 5098
50985098 
50995099 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
51005100 
5101 Permission policy for tool execution.
5101 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
51025102 
51035103 - `BetaManagedAgentsAlwaysAllowPolicy object`
51045104 
from line 5128
51285128 
51295129 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
51305130 
5131 Permission policy for tool execution.
5131 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
51325132 
51335133 - `BetaManagedAgentsAlwaysAllowPolicy object`
51345134 
from line 5172
51725172 
51735173 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
51745174 
5175 Permission policy for tool execution.
5175 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
51765176 
51775177 - `BetaManagedAgentsAlwaysAllowPolicy object`
51785178 
from line 5210
52105210 
52115211 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
52125212 
5213 Permission policy for tool execution.
5213 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
52145214 
52155215 - `BetaManagedAgentsAlwaysAllowPolicy object`
52165216 
from line 5256
52565256 
52575257 - `default_config: optional BetaManagedAgentsAgentToolsetDefaultConfigParams or null`
52585258 
5259 Default configuration for all tools in a toolset.
5259 Default configuration applied to all tools in this set.
52605260 
52615261 - `enabled: optional boolean or null`
52625262 
from line 5264
52645264 
52655265 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
52665266 
5267 Permission policy for tool execution.
5267 Default permission policy for tools. Controls whether tool calls are auto-approved or require confirmation.
52685268 
52695269 - `BetaManagedAgentsAlwaysAllowPolicy object`
52705270 
from line 5306
53065306 
53075307 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
53085308 
5309 Permission policy for tool execution.
5309 Permission policy for this tool. Overrides the `default_config` setting.
53105310 
53115311 - `BetaManagedAgentsAlwaysAllowPolicy object`
53125312 
from line 5322
53225322 
53235323 - `default_config: optional BetaManagedAgentsMCPToolsetDefaultConfigParams or null`
53245324 
5325 Default configuration for all tools from an MCP server.
5325 Default configuration for all tools from this server.
53265326 
53275327 - `enabled: optional boolean or null`
53285328 
from line 5330
53305330 
53315331 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
53325332 
5333 Permission policy for tool execution.
5333 Default permission policy for tools from this server.
53345334 
53355335 - `BetaManagedAgentsAlwaysAllowPolicy object`
53365336 
from line 5358
53585358 
53595359 - `input_schema: BetaManagedAgentsCustomToolInputSchema`
53605360 
5361 JSON Schema for custom tool input parameters.
5361 JSON Schema defining the expected input parameters for the tool.
53625362 
53635363 - `type: "object"`
53645364 
from line 5374
53745374 
53755375- `budget: optional BetaManagedAgentsBudgetLimit or null`
53765376 
5377 A hard spend ceiling. The session stops issuing new model requests once the tracked list cost reaches `max_list_cost`.
5377 Enforced spend ceiling for the session. Set an object to replace the budget of a session that was created with one, or `null` to remove it; omit to preserve. A budget cannot be added to a session created without one (rejected with reason `budget_create_only`), and a removed budget cannot be re-added. Allowed in any non-terminated status. Lowering `max_list_cost` to at or below the session's consumed list cost is rejected with reason `budget_not_raised`, and every model the session can run must have a public list price or the request is rejected with reason `model_not_budgetable`.
53785378 
53795379 - `type: "limit"`
53805380 
53815381 - `max_list_cost: BetaMonetaryAmount`
53825382 
5383 A monetary amount in a specific currency.
5383 Maximum list cost the session may accrue. List price is used regardless of any negotiated discount, so the cap fires at or before the actual charge.
53845384 
53855385 - `amount: string`
53865386 
from line 5442
54425442 
54435443 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
54445444 
5445 - `string`
5446 
54455447 - `"claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more`
54465448 
54475449 The model that will power your agent.
from line 5510
55085510 
55095511 High-performance model for agents and coding
55105512 
5511 - `string`
5512 
55135513 - `effort: optional BetaManagedAgentsEffortLow or BetaManagedAgentsEffortMedium or BetaManagedAgentsEffortHigh or 2 more`
55145514 
5515 How hard Claude works on each turn. Sets `output_config.effort` on every Messages call the session makes.
5515 How hard Claude works on each inference call. One of `low`, `medium`, `high`, `xhigh`, `max`. Always present; resolved to the per-model default at save time when not supplied.
55165516 
55175517 - `BetaManagedAgentsEffortLow object`
55185518 
from line 5550
55505550 
55515551 - `speed: optional "standard" or "fast"`
55525552 
5553 Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
5553 Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Defaults to `standard`. Not all models support `fast`; invalid combinations are rejected at create time.
55545554 
55555555 - `"standard"`
55565556 
from line 5558
55585558 
55595559 - `multiagent: BetaManagedAgentsSessionMultiagentCoordinator or null`
55605560 
5561 Resolved coordinator topology with full agent definitions for each roster member.
5561 Resolved multiagent orchestration configuration. Null when the agent is single-threaded.
55625562 
55635563 - `type: "coordinator"`
55645564 
from line 6014
60146014 
60156015 - `archived_at: string or null`
60166016 
6017 A timestamp in RFC 3339 format
6017 When the session was archived. Null if not archived.
60186018 
60196019 format: date-time
60206020 
60216021 - `budget: BetaManagedAgentsBudgetLimit or null`
60226022 
6023 A hard spend ceiling. The session stops issuing new model requests once the tracked list cost reaches `max_list_cost`.
6023 The session's enforced spend ceiling, or null when no budget is set.
60246024 
60256025 - `type: "limit"`
60266026 
60276027 - `max_list_cost: BetaMonetaryAmount`
60286028 
6029 A monetary amount in a specific currency.
6029 Maximum list cost the session may accrue. List price is used regardless of any negotiated discount, so the cap fires at or before the actual charge.
60306030 
60316031 - `amount: string`
60326032 
from line 6054
60546054 
60556055 - `completed_at: string or null`
60566056 
6057 A timestamp in RFC 3339 format
6057 When the outcome reached a terminal result. Null while `pending`/`running`/`evaluating`.
60586058 
60596059 format: date-time
60606060 
from line 6160
61606160 
61616161 - `access: optional "read_write" or "read_only" or null`
61626162 
6163 Access mode for an attached memory store.
6163 Access mode for the mounted store. Defaults to `read_write`. `read_only` mounts the store as a read-only filesystem.
61646164 
61656165 - `"read_write"`
61666166 
from line 6186
61866186 
61876187 - `stats: BetaManagedAgentsSessionStats`
61886188 
6189 Timing statistics for a session.
6189 Timing statistics for the session.
61906190 
61916191 - `active_seconds: optional number`
61926192 
from line 6230
62306230 
62316231 - `usage: BetaManagedAgentsSessionUsage`
62326232 
6233 Cumulative token usage for a session across all turns.
6233 Cumulative token usage for the session.
62346234 
62356235 - `active_seconds: optional number`
62366236 
from line 6240
62406240 
62416241 - `cache_creation: optional BetaManagedAgentsCacheCreationUsage`
62426242 
6243 Prompt-cache creation token usage broken down by cache lifetime.
6243 Tokens used to create prompt cache entries, broken down by cache TTL.
62446244 
62456245 - `ephemeral_1h_input_tokens: optional number`
62466246 
from line 6268
62686268 
62696269 - `list_cost: optional BetaMonetaryAmount or null`
62706270 
6271 A monetary amount in a specific currency.
6271 Cumulative list cost of the session across all turns, priced at public list rates. Absent until cost tracking is available for the session.
62726272 
62736273 - `output_tokens: optional number`
62746274 
from line 6278
62786278 
62796279 - `server_tool_use: optional BetaManagedAgentsServerToolUsage or null`
62806280 
6281 Cumulative count of server-executed tool invocations, broken down by tool.
6281 Cumulative server-executed tool usage across all turns. Absent until server-tool tracking is available for the session.
62826282 
62836283 - `web_fetch_requests: optional number`
62846284 
from line 6818
68186818 
68196819 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
68206820 
6821 - `string`
6822 
68216823 - `"claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more`
68226824 
68236825 The model that will power your agent.
from line 6886
68846886 
68856887 High-performance model for agents and coding
68866888 
6887 - `string`
6888 
68896889 - `effort: optional BetaManagedAgentsEffortLow or BetaManagedAgentsEffortMedium or BetaManagedAgentsEffortHigh or 2 more`
68906890 
6891 How hard Claude works on each turn. Sets `output_config.effort` on every Messages call the session makes.
6891 How hard Claude works on each inference call. One of `low`, `medium`, `high`, `xhigh`, `max`. Always present; resolved to the per-model default at save time when not supplied.
68926892 
68936893 - `BetaManagedAgentsEffortLow object`
68946894 
from line 6926
69266926 
69276927 - `speed: optional "standard" or "fast"`
69286928 
6929 Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
6929 Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Defaults to `standard`. Not all models support `fast`; invalid combinations are rejected at create time.
69306930 
69316931 - `"standard"`
69326932 
from line 6934
69346934 
69356935 - `multiagent: BetaManagedAgentsSessionMultiagentCoordinator or null`
69366936 
6937 Resolved coordinator topology with full agent definitions for each roster member.
6937 Resolved multiagent orchestration configuration. Null when the agent is single-threaded.
69386938 
69396939 - `type: "coordinator"`
69406940 
from line 7390
73907390 
73917391 - `archived_at: string or null`
73927392 
7393 A timestamp in RFC 3339 format
7393 When the session was archived. Null if not archived.
73947394 
73957395 format: date-time
73967396 
73977397 - `budget: BetaManagedAgentsBudgetLimit or null`
73987398 
7399 A hard spend ceiling. The session stops issuing new model requests once the tracked list cost reaches `max_list_cost`.
7399 The session's enforced spend ceiling, or null when no budget is set.
74007400 
74017401 - `type: "limit"`
74027402 
74037403 - `max_list_cost: BetaMonetaryAmount`
74047404 
7405 A monetary amount in a specific currency.
7405 Maximum list cost the session may accrue. List price is used regardless of any negotiated discount, so the cap fires at or before the actual charge.
74067406 
74077407 - `amount: string`
74087408 
from line 7430
74307430 
74317431 - `completed_at: string or null`
74327432 
7433 A timestamp in RFC 3339 format
7433 When the outcome reached a terminal result. Null while `pending`/`running`/`evaluating`.
74347434 
74357435 format: date-time
74367436 
from line 7536
75367536 
75377537 - `access: optional "read_write" or "read_only" or null`
75387538 
7539 Access mode for an attached memory store.
7539 Access mode for the mounted store. Defaults to `read_write`. `read_only` mounts the store as a read-only filesystem.
75407540 
75417541 - `"read_write"`
75427542 
from line 7562
75627562 
75637563 - `stats: BetaManagedAgentsSessionStats`
75647564 
7565 Timing statistics for a session.
7565 Timing statistics for the session.
75667566 
75677567 - `active_seconds: optional number`
75687568 
from line 7606
76067606 
76077607 - `usage: BetaManagedAgentsSessionUsage`
76087608 
7609 Cumulative token usage for a session across all turns.
7609 Cumulative token usage for the session.
76107610 
76117611 - `active_seconds: optional number`
76127612 
from line 7616
76167616 
76177617 - `cache_creation: optional BetaManagedAgentsCacheCreationUsage`
76187618 
7619 Prompt-cache creation token usage broken down by cache lifetime.
7619 Tokens used to create prompt cache entries, broken down by cache TTL.
76207620 
76217621 - `ephemeral_1h_input_tokens: optional number`
76227622 
from line 7644
76447644 
76457645 - `list_cost: optional BetaMonetaryAmount or null`
76467646 
7647 A monetary amount in a specific currency.
7647 Cumulative list cost of the session across all turns, priced at public list rates. Absent until cost tracking is available for the session.
76487648 
76497649 - `output_tokens: optional number`
76507650 
from line 7654
76547654 
76557655 - `server_tool_use: optional BetaManagedAgentsServerToolUsage or null`
76567656 
7657 Cumulative count of server-executed tool invocations, broken down by tool.
7657 Cumulative server-executed tool usage across all turns. Absent until server-tool tracking is available for the session.
76587658 
76597659 - `web_fetch_requests: optional number`
76607660 
from line 7972
79727972 
79737973 Replacement model. Accepts the model string, e.g. `claude-opus-5`, or a `model_config` object. Omit to use the agent's model.
79747974 
7975 - `BetaManagedAgentsModel = "claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more or string`
7975 - `BetaManagedAgentsModel = string or "claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more`
79767976 
79777977 The model that will power your agent.
79787978 
79797979 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
79807980 
7981 - `string`
7982 
79817983 - `"claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more`
79827984 
79837985 The model that will power your agent.
from line 8046
80448046 
80458047 High-performance model for agents and coding
80468048 
8047 - `string`
8048 
80498049 - `BetaManagedAgentsModelConfigParams object`
80508050 
80518051 An object that defines additional configuration control over model use
from line 8120
81208120 
81218121 - `speed: optional "standard" or "fast" or null`
81228122 
8123 Inference speed mode. `fast` provides significantly faster output token generation at premium pricing. Not all models support `fast`; invalid combinations are rejected at create time.
8123 Inference speed mode. Defaults to `standard`.
81248124 
81258125 - `"standard"`
81268126 
from line 8202
82028202 
82038203 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
82048204 
8205 Permission policy for tool execution.
8205 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
82068206 
82078207 - `BetaManagedAgentsAlwaysAllowPolicy object`
82088208 
from line 8238
82388238 
82398239 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
82408240 
8241 Permission policy for tool execution.
8241 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
82428242 
82438243 - `BetaManagedAgentsAlwaysAllowPolicy object`
82448244 
from line 8268
82688268 
82698269 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
82708270 
8271 Permission policy for tool execution.
8271 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
82728272 
82738273 - `BetaManagedAgentsAlwaysAllowPolicy object`
82748274 
from line 8298
82988298 
82998299 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
83008300 
8301 Permission policy for tool execution.
8301 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
83028302 
83038303 - `BetaManagedAgentsAlwaysAllowPolicy object`
83048304 
from line 8328
83288328 
83298329 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
83308330 
8331 Permission policy for tool execution.
8331 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
83328332 
83338333 - `BetaManagedAgentsAlwaysAllowPolicy object`
83348334 
from line 8358
83588358 
83598359 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
83608360 
8361 Permission policy for tool execution.
8362 
8363 - `BetaManagedAgentsAlwaysAllowPolicy object`
8364 
8365 Tool calls are automatically approved without user confirmation.
8366 
8367 - `BetaManagedAgentsAlwaysAskPolicy object`
8368 
8369 Tool calls require user confirmation before execution.
8370 
8371 - `BetaManagedAgentsAutoPolicy object`
8372 
8373 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
8374 
8375 - `BetaManagedAgentsWebFetchToolConfigParams object`
8376 
8377 Configuration override for the web_fetch tool.
8378 
8379 - `type: optional "web_fetch"`
8380 
8381 - `name: "web_fetch"`
8382 
8383 Must be "web_fetch".
8384 
8385 - `allowed_domains: optional array of string`
8386 
8387 Only fetch URLs whose host is one of these domains or a subdomain of one. Each entry is a plain hostname like "docs.example.com" (no scheme, port, or path). At most 64 entries; an empty list is rejected (omit the field instead). Cannot be combined with blocked_domains.
8388 
8389 - `blocked_domains: optional array of string`
8390 
8391 Never fetch URLs whose host is one of these domains or a subdomain of one. Each entry is a plain hostname like "ads.example.com" (no scheme, port, or path). At most 64 entries; an empty list is rejected (omit the field instead). Cannot be combined with allowed_domains.
8392 
8393 - `enabled: optional boolean or null`
8394 
8395 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
8396 
8397 - `max_content_tokens: optional number or null`
8398 
8399 Maximum number of tokens of fetched text content to include in context per call. Does not apply to binary content such as PDFs.
8400 
8401 format: int32
8402 
8403 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
8404 
8405 Permission policy for tool execution.
8406 
8407 - `BetaManagedAgentsAlwaysAllowPolicy object`
8408 
8409 Tool calls are automatically approved without user confirmation.
8410 
8411 - `BetaManagedAgentsAlwaysAskPolicy object`
8412 
8413 Tool calls require user confirmation before execution.
8414 
8415 - `BetaManagedAgentsAutoPolicy object`
8416 
8417 The server decides each tool call individually: it judges, from the tool, its input, and the session content so far, whether the call is safe to execute or high-risk, and evaluates it to allow when judged safe and to deny when judged high-risk. A call the server cannot reach a judgement on evaluates to ask.
8418 
8419 - `BetaManagedAgentsWebSearchToolConfigParams object`
8420 
8421 Configuration override for the web_search tool.
8422 
8423 - `type: optional "web_search"`
8424 
8425 - `name: "web_search"`
8426 
8427 Must be "web_search".
8428 
8429 - `allowed_domains: optional array of string`
8430 
8431 Only return search results whose host is one of these domains or a subdomain of one. Each entry is a plain hostname like "docs.example.com" (no scheme or port; an optional path suffix is accepted). At most 64 entries; an empty list is rejected (omit the field instead). Cannot be combined with blocked_domains.
8432 
8433 - `blocked_domains: optional array of string`
8434 
8435 Never return search results whose host is one of these domains or a subdomain of one. Each entry is a plain hostname like "ads.example.com" (no scheme or port; an optional path suffix is accepted). At most 64 entries; an empty list is rejected (omit the field instead). Cannot be combined with allowed_domains.
8436 
8437 - `enabled: optional boolean or null`
8438 
8439 Whether this tool is enabled and available to Claude. Overrides the default_config setting.
8440 
8441 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
8442 
8443 Permission policy for tool execution.
8444 
8445 - `BetaManagedAgentsAlwaysAllowPolicy object`
8446 
8447 Tool calls are automatically approved without user confirmation.
8448 
8449 - `BetaManagedAgentsAlwaysAskPolicy object`
8450 
8451 Tool calls require user confirmation before execution.
8452 
8453 - `BetaManagedAgentsAutoPolicy object`
8454 
8455
8361 Permission p
Feedback