Saltar al contenido principal
PUT
Update incoming calls
Primero lee la configuración actual con Consultar llamadas entrantes y envía la configuración completa con "confirm": true. Las claves MCP necesitan numbers:write.

Modos por franja

El horario es el de Contactabilidad de la tarea donde empiezan las llamadas. Si no hay horario definido, todas las llamadas usan in_hours.

Modos distintos según el horario

  • Toda franja distinta de ai_only necesita team_number, con + y el código de país. No puede ser el número del propio agente.
  • Define mode como lo hace el panel: ai_human con transfer_to igual a team_number cuando una franja es ai_first_team_when_asked, forward con transfer_to cuando ambas franjas son team_only, y ai en los demás casos.
  • human_first_then_ai y team_only necesitan un número que pueda llamar a un teléfono externo. Si no, la respuesta es 422 call_forwarding_provider_unsupported.
Sin slots, mode aplica todo el tiempo. off detiene las llamadas entrantes y mantiene el número asignado al agente.

Después de guardar

La respuesta vuelve con status: "pending". Las llamadas toman el cambio en uno o dos minutos; vuelve a leer hasta que status sea ready. Un estado pending o error todavía no está activo. Las respuestas de estas escrituras incluyen saved_in_agent_draft. Es false cuando el cambio se aplicó al agente en vivo (toda API key y MCP key de cliente). Es true solo cuando NexorGPT guardó el cambio en el borrador del agente; ese cambio llega a las llamadas al publicar el borrador. Un 409 group_call_settings_number_in_use indica que otro agente ya recibe llamadas en este número. Cada número contesta llamadas para un solo agente: quítalo primero del otro agente, o usa otro número.

Autorizaciones

X-API-Key
string
header
requerido

API key for authentication. Keys are prefixed with nxr_live_.

Parámetros de ruta

id
string<uuid>
requerido

Agent ID (master_workflow in the API). A malformed ID returns 400 invalid_group_id.

Cuerpo

application/json

Any other key returns 400 group_call_settings_invalid.

mode
enum<string>
requerido

off stops incoming calls and keeps the number assigned. ai: the AI answers. ai_human: the AI answers and transfers to transfer_to on request. forward: calls go to transfer_to without the AI. With slots, use ai_human when a slot is ai_first_team_when_asked, forward when both slots are team_only, and ai otherwise.

Opciones disponibles:
off,
ai,
ai_human,
forward
number_id
string<uuid>
requerido

The agent's number, from numbers in the GET response.

confirm
enum<boolean>
requerido

Must be true.

Opciones disponibles:
true
transfer_to
string | null

Required for ai_human and forward, not allowed for off and ai. When team_number is also sent, it must be the same number.

Pattern: ^\+[1-9]\d{7,14}$
slots
object | null

Who answers during business hours and after hours. Business hours are the Contactability hours of the task where calls start; with no hours set, every call uses in_hours. ai_only: the AI answers and does not transfer when asked for a person. ai_first_team_when_asked: the AI answers and transfers to team_number when the caller asks for a person. human_first_then_ai: team_number rings first, and the AI answers after rings rings if nobody takes the call. team_only: calls go to team_number without the AI.

team_number
string | null

Your team's number, with + and the country code. Required when a slot is not ai_only. It cannot be the agent's own number.

Pattern: ^\+[1-9]\d{7,14}$
rings
integer | null

Rings before the AI answers in human_first_then_ai, from 1 to 6. Default 4. Only with slots.

Rango requerido: 1 <= x <= 6

Respuesta

Saved settings. status is pending until calls use them.

agent_group_id
string<uuid>
requerido

Agent ID.

mode
enum<string>
requerido

Single incoming-call mode. When slots is set, slots decides who answers each call.

Opciones disponibles:
off,
ai,
ai_human,
forward
number_id
string<uuid> | null
requerido

The agent's number that receives the calls.

transfer_to
string | null
requerido

Destination for ai_human or forward; otherwise null.

Pattern: ^\+[1-9]\d{7,14}$
status
enum<string>
requerido

ready when calls already use these settings. pending while the change is being applied, usually a minute or two. Never treat pending or error as active.

Opciones disponibles:
ready,
pending,
error,
unconfigured
numbers
object[]
requerido

Numbers of the account that this agent can use.

linked_numbers
object[]
requerido

Other numbers whose calls follow these settings.

slots
object | null
requerido

Who answers during business hours and after hours. Business hours are the Contactability hours of the task where calls start; with no hours set, every call uses in_hours. ai_only: the AI answers and does not transfer when asked for a person. ai_first_team_when_asked: the AI answers and transfers to team_number when the caller asks for a person. human_first_then_ai: team_number rings first, and the AI answers after rings rings if nobody takes the call. team_only: calls go to team_number without the AI. null when incoming calls are off.

team_number
string | null
requerido

Your team's number, shared by both slots.

Pattern: ^\+[1-9]\d{7,14}$
rings
integer | null
requerido

Rings before the AI answers in human_first_then_ai. null means the default of 4. Each ring lasts about 6 seconds.

Rango requerido: 1 <= x <= 6
saved_in_agent_draft
boolean
requerido

false when the change was applied to the live agent (every customer API key and MCP key). true only when NexorGPT saved the change to the agent's draft; that change reaches calls when the draft is published.

Última modificación el 26 de octubre de 2026