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

Create Agent changedapi/beta/agents/create

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+764added
Lines−755removed
From line 127 where the diff opens
First seen 14 Aug 2026 this site's first read of the page
Recorded edits11to this page, all time

The whole hunk

from line 127, old and new numbered
/
lines
from line 127
127127 
128128 Model identifier. Accepts the [model string](https://platform.claude.com/docs/en/about-claude/models/overview#latest-models-comparison), e.g. `claude-opus-5`, or a `model_config` object for additional configuration control
129129 
130 - `BetaManagedAgentsModel = "claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more or string`
130 - `BetaManagedAgentsModel = string or "claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more`
131131 
132132 The model that will power your agent.
133133 
134134 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
135135 
136 - `string`
137 
136138 - `"claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more`
137139 
138140 The model that will power your agent.
from line 201
199201 
200202 High-performance model for agents and coding
201203 
202 - `string`
203 
204204 - `BetaManagedAgentsModelConfigParams object`
205205 
206206 An object that defines additional configuration control over model use
from line 275
275275 
276276 - `speed: optional "standard" or "fast" or null`
277277 
278 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.
278 Inference speed mode. Defaults to `standard`.
279279 
280280 - `"standard"`
281281 
from line 317
317317 
318318- `multiagent: optional BetaManagedAgentsMultiagentParams or null`
319319 
320 A coordinator topology: the session's primary thread orchestrates work by spawning session threads, each running an agent drawn from the `agents` roster.
320 Multiagent orchestration configuration. Currently supports the `coordinator` topology with a roster of 1-20 agents.
321321 
322322 - `type: "coordinator"`
323323 
from line 439
439439 
440440 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
441441 
442 Permission policy for tool execution.
442 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
443443 
444444 - `BetaManagedAgentsAlwaysAllowPolicy object`
445445 
from line 475
475475 
476476 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
477477 
478 Permission policy for tool execution.
478 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
479479 
480480 - `BetaManagedAgentsAlwaysAllowPolicy object`
481481 
from line 505
505505 
506506 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
507507 
508 Permission policy for tool execution.
508 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
509509 
510510 - `BetaManagedAgentsAlwaysAllowPolicy object`
511511 
from line 535
535535 
536536 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
537537 
538 Permission policy for tool execution.
538 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
539539 
540540 - `BetaManagedAgentsAlwaysAllowPolicy object`
541541 
from line 565
565565 
566566 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
567567 
568 Permission policy for tool execution.
568 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
569569 
570570 - `BetaManagedAgentsAlwaysAllowPolicy object`
571571 
from line 595
595595 
596596 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
597597 
598 Permission policy for tool execution.
598 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
599599 
600600 - `BetaManagedAgentsAlwaysAllowPolicy object`
601601 
from line 639
639639 
640640 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
641641 
642 Permission policy for tool execution.
642 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
643643 
644644 - `BetaManagedAgentsAlwaysAllowPolicy object`
645645 
from line 677
677677 
678678 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
679679 
680 Permission policy for tool execution.
680 Permission policy for this tool. Controls whether tool calls are auto-approved or require confirmation.
681681 
682682 - `BetaManagedAgentsAlwaysAllowPolicy object`
683683 
from line 723
723723 
724724 - `default_config: optional BetaManagedAgentsAgentToolsetDefaultConfigParams or null`
725725 
726 Default configuration for all tools in a toolset.
726 Default configuration applied to all tools in this set.
727727 
728728 - `enabled: optional boolean or null`
729729 
from line 731
731731 
732732 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
733733 
734 Permission policy for tool execution.
734 Default permission policy for tools. Controls whether tool calls are auto-approved or require confirmation.
735735 
736736 - `BetaManagedAgentsAlwaysAllowPolicy object`
737737 
from line 773
773773 
774774 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
775775 
776 Permission policy for tool execution.
776 Permission policy for this tool. Overrides the `default_config` setting.
777777 
778778 - `BetaManagedAgentsAlwaysAllowPolicy object`
779779 
from line 789
789789 
790790 - `default_config: optional BetaManagedAgentsMCPToolsetDefaultConfigParams or null`
791791 
792 Default configuration for all tools from an MCP server.
792 Default configuration for all tools from this server.
793793 
794794 - `enabled: optional boolean or null`
795795 
from line 797
797797 
798798 - `permission_policy: optional BetaManagedAgentsAlwaysAllowPolicy or BetaManagedAgentsAlwaysAskPolicy or BetaManagedAgentsAutoPolicy or null`
799799 
800 Permission policy for tool execution.
800 Default permission policy for tools from this server.
801801 
802802 - `BetaManagedAgentsAlwaysAllowPolicy object`
803803 
from line 825
825825 
826826 - `input_schema: BetaManagedAgentsCustomToolInputSchema`
827827 
828 JSON Schema for custom tool input parameters.
828 JSON Schema defining the expected input parameters for the tool.
829829 
830830 - `type: "object"`
831831 
from line 851
851851 
852852 - `archived_at: string or null`
853853 
854 A timestamp in RFC 3339 format
854 When the agent was archived. Null if not archived.
855855 
856856 format: date-time
857857 
from line 883
883883 
884884 See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.
885885 
886 - `string`
887 
886888 - `"claude-opus-5-5" or "claude-fable-5-1" or "claude-sonnet-5" or 12 more`
887889 
888890 The model that will power your agent.
from line 951
949951 
950952 High-performance model for agents and coding
951953 
952 - `string`
953 
954954 - `effort: optional BetaManagedAgentsEffortLow or BetaManagedAgentsEffortMedium or BetaManagedAgentsEffortHigh or 2 more`
955955 
956 How hard Claude works on each turn. Sets `output_config.effort` on every Messages call the session makes.
956 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.
957957 
958958 - `BetaManagedAgentsEffortLow object`
959959 
from line 991
991991 
992992 - `speed: optional "standard" or "fast"`
993993 
994 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.
994 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.
995995 
996996 - `"standard"`
997997 
from line 999
999999 
10001000 - `multiagent: BetaManagedAgentsMultiagent or null`
10011001 
1002 Resolved coordinator topology with a concrete agent roster.
1002 Multiagent orchestration configuration. Null when the agent is single-threaded.
10031003 
10041004 - `type: "coordinator"`
10051005 
from line 1437
14371437 "description": "A general-purpose starter agent.",
14381438 "metadata": {
14391439 "foo": "bar"
1440 },
1441 "multiagent": {
1442 "agents": [
1443 "agent_011CZkYqphY8vELVzwCUpqiQ",
1444 {
1445 "type": "self"
1446 }
1447 ],
1448 "type": "coordinator"
14401449 },
14411450 "system": "You are a general-purpose agent that can research, write code, run commands, and use connected tools to complete the user'\''s task end to end.",
14421451 "tools": [
Feedback