Getting StartedQuick start guide

Getting started with Semantic Kernel

In just a few steps, you can build your first AI agent with Semantic Kernel in JavaScript. This guide will show you how to…

  • Install the necessary packages
  • Create a back-and-forth conversation with an AI
  • Give an AI agent the ability to run your code
  • Watch the AI create plans on the fly

Install the package

npm install --save semantic-kernel @semantic-kernel/openai

Initialize the Kernel

import { OpenAIChatCompletionService } from '@semantic-kernel/openai';
import { FunctionChoiceBehavior, kernel, kernelFunction } from 'semantic-kernel';
 
const sk = kernel().addService(
  new OpenAIChatCompletionService({
    model: 'gpt-3.5-turbo',
    apiKey:
      'YOUR_OPENAI_API_KEY',
  })
);

Add your plugins

const temperature = kernelFunction(({ loc }) => (loc === 'Dublin' ? 10 : 24), {
  name: 'temperature',
  description: 'Returns the temperature for the given location',
  schema: {
    type: 'object',
    properties: {
      loc: { type: 'string', description: 'The location to return the temperature for' },
    },
  },
});
 
sk.addPlugin({
  name: 'weather',
  description: 'Weather plugin',
  functions: [temperature],
});

Invoke prompt

const result = await sk.invokePrompt({
  promptTemplate: 'Return the current temperature in Dublin',
  executionSettings: {
      functionChoiceBehavior: FunctionChoiceBehavior.Auto(),
  },
});

Here is the full example:

import { OpenAIChatCompletionService } from '@semantic-kernel/openai';
import { FunctionChoiceBehavior, kernel, kernelFunction } from 'semantic-kernel';
 
const sk = kernel().addService(
  new OpenAIChatCompletionService({
    model: 'gpt-3.5-turbo',
    apiKey:
      'YOUR_OPENAI_API_KEY',
  })
);
 
const temperature = kernelFunction(({ loc }) => (loc === 'Dublin' ? 10 : 24), {
  name: 'temperature',
  description: 'Returns the temperature for the given location',
  schema: {
    type: 'object',
    properties: {
      loc: { type: 'string', description: 'The location to return the temperature for' },
    },
  },
});
 
sk.addPlugin({
  name: 'weather',
  description: 'Weather plugin',
  functions: [temperature],
});
 
const result = await sk.invokePrompt({
  promptTemplate: 'Return the current temperature in Dublin',
  executionSettings: {
      functionChoiceBehavior: FunctionChoiceBehavior.Auto(),
  },
});
 
// Prints the output after executing the plugin and the given prompt
console.log(result);
OverviewKernel