import type { Inputs } from "@scoopika/types";
{
  "message": "Hello",
  "images": [ "scoopika.com/path-to-image" ],
  "audio": [
    {
      "type": "remote",
      "path": "scoopika.com/path-to-audio"
    }
  ],
  "context": [
    {
      "scope": "run",
      "description": "current date",
      "value": "1 May 2024"
    }
  ]
}

Inputs is the primary input for running both agents and boxes through the .run method. It allows you to specify the message, history session ID, run ID, and inject additional data.

Props

message
string

The primary message to be sent to the agent for processing. This message becomes the latest chat message within the agent’s context and the chat session.

images
string[]

A list of images urls (links or data urls). only pass if the LLM powering your agent supports vision.

audio
AudioInput[]

A list of audio files (as links or base64).

Reference to AudioInput

context
ContextInput[]

You can define extra context about anything you want the agent to be aware of. a context object has a scope, run or session. refer to this link for more info:

Reference to ContextInput

import type { Inputs } from "@scoopika/types";
{
  "message": "Hello",
  "images": [ "scoopika.com/path-to-image" ],
  "audio": [
    {
      "type": "remote",
      "path": "scoopika.com/path-to-audio"
    }
  ],
  "context": [
    {
      "scope": "run",
      "description": "current date",
      "value": "1 May 2024"
    }
  ]
}

Was this page helpful?