Create an agent

POST

agents

object

description

string

display_description

string

display_name

string

greeting_message

string

initial_turn

enum<string>

model

enum<string>

name

string

openai

object

frequency_penalty

number

presence_penalty

number

temperature

number

top_p

number

system_message

string

Body

application/json

The data with which an agent is created.

model

enum<string>

required

The underlying model that powers the agent.

Available options:

anthropic.claude-v1,

anthropic.claude-v1-100k,

anthropic.claude-instant-v1,

anthropic.claude-instant-v1-100k,

openai.gpt-4,

openai.gpt-3.5-turbo,

openai.gpt-3.5-turbo-16k

description

string | null

The description of the agent, intended for developers' reference.

display_description

string | null

The agent's public-facing description, meant to be visible to your users.

display_name

string | null

The agent's display name, intended to serve as the public-facing name visible to your users.

greeting_message

string | null

The message sent by the agent to initiate an interaction. This property supports <a href="https://handlebarsjs.com/guide/#what-is-handlebars">Handlebars</a> template format where you have access to the <code>user</code> and <code>agent</code> parameters. You can use these parameters to add context and personalize the message.

initial_turn

enum<string> | null

Indicates who sends the first message in each interaction. Defaults to "user".

Available options:

user,

agent

name

string | null

The name of the agent, designed to be visible to developers for internal reference purposes.

openai

object | null

The configuration that is applied to OpenAI models.

openai.frequency_penalty

number | null

Can take values between -2.0 and 2.0, with positive values penalizing the model for generating new tokens that have already appeared frequently in the generated text so far. This reduces the likelihood of the model repeating the same line or phrase verbatim. In essence, a higher positive value for <code>frequency_penalty</code> encourages the model to use less frequently used tokens in the generated text. See original <a href="https://platform.openai.com/docs/api-reference/completions/create#completions/create-frequency_penalty">definition</a>.

openai.presence_penalty

number | null

Can take values between -2.0 and 2.0, with positive values penalizing the model for generating new tokens that have already appeared in the generated text. This encourages the model to generate text about new topics and increases the likelihood of generating more diverse content. In essence, a higher positive value for <code>presence_penalty</code> will lead to more avoidance of repeating previously used tokens in the generated text. See original <a href="https://platform.openai.com/docs/api-reference/completions/create#completions/create-presence_penalty">definition</a>.

openai.temperature

number | null

Determines the level of randomness in the generated output. Can take values between 0 and 2.0, with higher values like 0.8 resulting in more randomness, and lower values like 0.2 making the output more predictable. See original <a href="https://platform.openai.com/docs/api-reference/completions/create#completions/create-temperature">definition</a>.

openai.top_p

number | null

An alternative to the <code>temperature</code> option. Uses a method called nucleus sampling, which involves considering only the tokens with the highest probability mass. It specifies the proportion of the probability mass to consider, with a value of 0.1 indicating that only the tokens comprising the top 10% probability mass will be taken into account. In essence, <code>top_p</code> determines the size of the set of tokens to choose from, based on their probabilities. See original <a href="https://platform.openai.com/docs/api-reference/completions/create#completions/create-top_p">definition</a>.

system_message

string | null

The system prompt that will be injected at the beginning of each interaction. This property supports <a href="https://handlebarsjs.com/guide/#what-is-handlebars">Handlebars</a> template format where you have access to the <code>user</code> and <code>agent</code> parameters. You can use these parameters to provide rich context to the agent and steer it in a specific direction.

Response

200 - application/json

An _agent_ is an AI entity that exists to serve your project's users. Agents are intelligent, configurable and context-aware and can interact with your users via a messaging system enabled by the Proficient API.

string

required

The unique identifier of the agent

object

enum<string>

required

The type of the Agent object

Available options:

agent

active

boolean

required

Whether the agent is currently active. If not, users will not be able to interact with the agent.

description

string

required

The description of the agent, intended for developers' reference.

display_description

string

required

The agent's public-facing description, meant to be visible to your users.

display_name

string

required

The agent's display name, intended to serve as the public-facing name visible to your users.

name

string

required

The name of the agent, designed to be visible to developers for internal reference purposes.

created_at

integer

required

The time at which the object was created, measured in milliseconds since the Unix epoch

updated_at

integer

required

The time at which the object was last updated, measured in milliseconds since the Unix epoch

List all agents Retrieve an agent

Admin API

Agents

Users

Body

Response