Skip to main content

llmoutputhandler

codebolt.codeboltutils.LLMOutputHandler(): void
Processes the AI model's response, manages tool executions, handles user communication, and detects when the task is completed.

Example​

const codebolt = require('@codebolt/codeboltjs').default;

const { LLMOutputHandler } = require("@codebolt/codeboltjs/utils");

const llmOutput = codebolt.llm.inference(prompt);
const outputHandler = new LLMOutputHandler(llmOutput, codebolt);

await outputHandler.sendMessageToUser();
const toolResults = await outputHandler.runTools();

if (outputHandler.isCompleted()) {
    console.log("Task completed!");
}

Methods​

The LLMOutputHandler class provides the following methods:

  • sendMessageToUser(): Sends the AI model's response message to the user
  • runTools(): Executes any tools that were called by the AI model and returns the results
  • isCompleted(): Checks if the current task has been completed based on the AI's response

Usage Notes​

  • The constructor requires llmOutput (the AI model's response) and codebolt parameters
  • Use sendMessageToUser() to communicate AI responses to the user
  • Call runTools() to execute any tool calls made by the AI
  • Check isCompleted() to determine if the workflow should continue or terminate