Endpoint URL

POST /api/v1/invoke_agent

Description

This endpoint allows you to invoke an agent using specific parameters, including expert ID, ability ID, model ID, session ID, input data, and additional configuration options.

Request Headers

• bbot-api-key: Your API key, which can be generated from the Profile section of the B-Bot hub.

Request Body

The request body should be a JSON object with the following structure:

{
    "expert_id": 37,
    "ability_id": 80,
    "model_id": 6,
    "temperature": 0,
    "session_id": "12333333",
    "input_data": "use the Assistant Consultation Tool to call an Assistant for help regarding a question",
    "user_id": "123", 
    "document_urls": [],
    "conversation_history": [],
    "tool_config": {
        "tavily_max_results": 1
    },
    "tool_activation": {
        "tavily_search": true,
        "wolfram_alpha": true,
        "document_retriever": false,
        "notion_connector": false}
}

Parameters

• expert_id (integer): The ID of the expert you want to invoke.
• ability_id (integer): The ID of the ability to be used by the agent.
• model_id (integer): The ID of the language model to be used.
• session_id (string): The session ID for tracking the interaction.
• input_data (string): The input query or command for the agent.
• temperature (integer): the temperature the Agent should work on.
• user_id (string): The ID of the user making the request. (Your own User ID, will be mixed with the user ID that you provide so you can store multiple sessions of multiple users)
• document_urls (array): A list of document URLs to be used by the agent.
• conversation_history (array): A list of previous interactions with the agent.
• tool_config (object): Configuration options for various tools available to the agent.
  • tavily_max_results (integer): The maximum number of results from the Tavily tool.