Skip to main content
POST
Create or update an AI agent

Authorizations

X-Api-Key
string
header
required

Your LinkedCamp API key.

Body

application/json
name
string
required

Internal name.

agentId
string

Omit to create a new agent.

type
enum<string>

CHAT for inbox replies, CONTENT for post drafting.

Available options:
CHAT,
VOICE,
CONTENT,
CAMPAIGN,
REPORTING
businessName
string

Company the agent represents; injected into its prompt.

provider
enum<string>

Primary model provider. Must be connected.

Available options:
OPENAI,
ANTHROPIC,
GEMINI,
GROK,
DEEPSEEK
fallbackProvider
enum<string>

Used when the primary provider errors.

Available options:
OPENAI,
ANTHROPIC,
GEMINI,
GROK,
DEEPSEEK
language
string

Language code, e.g. "en", "fr", "de", "it", "pt", "ja".

channels
string[]

Where the agent operates, e.g. ["linkedin","email"].

personality
string

Who the agent is and how it carries itself.

intent
string

The outcome it is steering the conversation toward.

instructions
string

Concrete writing rules: tone, length, what to avoid.

prompt
string

Raw system prompt override. Usually leave unset.

topics
string[]

Subjects the agent is allowed to discuss.

faqs
object[]

Canned answers the agent should use verbatim.

followups
object[]

Follow-up messages sent when a lead goes quiet.

prohabitedWords
string[]

Words the agent must never use. Field name is spelled this way in the schema.

waitTime
integer

Minutes to wait before replying, so it reads as human.

Required range: x >= 0
maxMessages
integer

Cap on agent messages per conversation.

Required range: x >= 1
pauseBotOnManualMessage
boolean

Stand down as soon as a human replies in the thread.

stopScenarios
object

Situations in which the agent stops replying.

action
string

Enable calendar booking.

Allowed value: "APPOINTMENT"
actionSettings
object

Booking settings. Used when action is "APPOINTMENT".

userIds
string[]

Share this agent with these sub-accounts.

notify
boolean

Notify you when the agent replies.

isActive
boolean

Whether the agent runs. Keep false while drafting.

Response

The result, as JSON. Writes answer { "ok": true, "message": ..., "data": ... }.

The response is of type object.