Create a new assistant

Creates a new assistant with the provided configuration.

POST /v1/assistants

Create a new assistant

cURL

curl --request POST \
  --url https://api.devic.ai/v1/assistants \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "name": "Customer Support Assistant",
  "description": "<string>",
  "presets": "<string>",
  "model": "<string>",
  "provider": "<string>",
  "imgUrl": "<string>",
  "availableToolsGroupsUids": [
    "<string>"
  ],
  "enabledTools": [
    "<string>"
  ],
  "accessConfiguration": {
    "externalAccess": true,
    "visibilityByRole": [
      "<string>"
    ]
  },
  "widgetConfiguration": {
    "enabled": true,
    "sourcesWhiteList": [
      "<string>"
    ],
    "color": "<string>",
    "welcomeMessage": "<string>"
  },
  "memoryDocuments": [
    {}
  ],
  "structuredOutput": {},
  "guardrailsConfiguration": {
    "enabled": true,
    "guardrails": {}
  },
  "codeSnippetIds": [
    "<string>"
  ],
  "availableSkillIds": [
    "<string>"
  ],
  "subagentsIds": [
    "<string>"
  ],
  "maxChatMessages": 123,
  "maxToolResponseInputTokens": 123,
  "contextManagement": {
    "fullContextTurnDepth": 123,
    "alwaysIncludeUserMessages": true
  }
}
'

201

{
  "_id": "<string>",
  "identifier": "<string>",
  "name": "<string>",
  "description": "<string>",
  "presets": "<string>",
  "model": "<string>",
  "provider": "<string>",
  "imgUrl": "<string>",
  "state": "<string>",
  "availableToolsGroupsUids": [
    "<string>"
  ],
  "enabledTools": [
    "<string>"
  ],
  "accessConfiguration": {},
  "widgetConfiguration": {},
  "memoryDocuments": [
    {}
  ],
  "structuredOutput": {},
  "guardrailsConfiguration": {},
  "codeSnippetIds": [
    "<string>"
  ],
  "availableSkillIds": [
    "<string>"
  ],
  "subagentsIds": [
    "<string>"
  ],
  "maxChatMessages": 123,
  "maxToolResponseInputTokens": 123,
  "contextManagement": {
    "fullContextTurnDepth": 123,
    "alwaysIncludeUserMessages": true
  },
  "creationTimestampMs": 123,
  "lastEditTimestampMs": 123,
  "reasoningEffort": "none"
}

Authorizations

Authorization · string · header · required

An API key from your Devic console, sent as Authorization: Bearer <key>. An embedded front end sends a tenant-session token instead, so no API key reaches the browser, and an OAuth integration sends its access token. Missing, invalid or expired credentials get a 401. See Authentication.

Body

application/json

name · string · required

Assistant name

Example:

"Customer Support Assistant"

description · string

Assistant description

presets · string

System prompt / instructions

model · string

Default LLM model

provider · string

Default LLM provider

imgUrl · string

Image URL for the assistant

state · enum<string>

Assistant state

Available options: active, inactive, coming_soon

availableToolsGroupsUids · string[]

Tool group UIDs the assistant can use

enabledTools · string[]

Explicit subset of enabled tool names

accessConfiguration · object

Access configuration

Show child attributes

widgetConfiguration · object

Widget configuration

Show child attributes

memoryDocuments · object[]

Memory documents for RAG

structuredOutput · object

Structured output JSON schema configuration

guardrailsConfiguration · object

Guardrails configuration

Show child attributes

codeSnippetIds · string[]

Code snippet IDs

availableSkillIds · string[]

Skill IDs the assistant can use

subagentsIds · string[]

Subagent IDs the assistant can invoke

maxChatMessages · integer

Maximum chat messages in context

maxToolResponseInputTokens · integer

Maximum input tokens for tool responses

contextManagement · object

Context depth control: summarize older turns to reduce the token footprint sent to the model.

Show child attributes

reasoningEffort · enum<string> | null

Configured reasoning level. Null restores inheritance; none is an explicit level. Supported values depend on the model: GET /api/v1/models.

Available options: none, minimal, low, medium, high, xhigh, max

Response

201

application/json

Assistant created successfully

_id · string

Assistant ID

identifier · string

Unique identifier

name · string

Assistant name

description · string

Assistant description

presets · string

System prompt / instructions

model · string

Default LLM model

provider · string

Default LLM provider

imgUrl · string

Image URL

state · string

Assistant state

availableToolsGroupsUids · string[]

enabledTools · string[]

accessConfiguration · object

widgetConfiguration · object

memoryDocuments · object[]

structuredOutput · object

guardrailsConfiguration · object

codeSnippetIds · string[]

availableSkillIds · string[]

subagentsIds · string[]

maxChatMessages · integer

maxToolResponseInputTokens · integer

contextManagement · object

Context depth control: summarize older turns to reduce the token footprint sent to the model.

Show child attributes

creationTimestampMs · integer

Creation timestamp in milliseconds

lastEditTimestampMs · integer

Last edit timestamp in milliseconds

reasoningEffort · enum<string> | null

Configured reasoning level. Null restores inheritance; none is an explicit level. Supported values depend on the model: GET /api/v1/models.

Available options: none, minimal, low, medium, high, xhigh, max