AzureAI.Community.SK.Plugin.TwilioWhatsApp 1.0.0

Azure AI Community GitHub

AzureAI Community Semantic Kernel Plugin - Twilio WhatsApp Plugin

The AzureAI.Community.SK.Plugin.TwilioWhatsApp plugin for Microsoft Semantic Kernel (SK) allows seamless integration with Twilio's WhatsApp API, enabling you to send WhatsApp messages directly from your SK-powered applications. By using this plugin, you can easily send messages to WhatsApp numbers through Twilio's API, integrating WhatsApp messaging into your workflows with minimal effort.

Prerequisites

Before using the WhatsAppServicePlugin, ensure you meet the following requirements:

1. Twilio Account
   You�ll need a Twilio account to access the Twilio API for WhatsApp messaging. If you don�t have one, you can sign up at Twilio.

2. Twilio API Credentials
   To authenticate with Twilio�s API, you must have the following credentials:

   • Account SID: Your Twilio Account SID, available in the Twilio Console.
   • Auth Token: Your Twilio Auth Token, also available in the console. You can locate these credentials in the "Account Info" section of your Twilio dashboard.

3. Twilio WhatsApp Sandbox Number
   To send WhatsApp messages, you�ll need to set up Twilio's WhatsApp sandbox. After configuring your sandbox in the Twilio console, you'll receive a sandbox number to send messages from.

   • From Number: The number from which messages will be sent (usually formatted as whatsapp:+344324234 for sandbox usage).

Before using the AzureAI.Community.SK.Plugin.TwilioWhatsApp, ensure the following are in place:

• Azure AI Semantic Kernel installed.
• Access to the Semantic Kernel Plugin system.

Installing the Plugin

To install the AzureAI.Community.SK.Plugin.TwilioWhatsApp, simply add it to your project via the kernel.

Example Code:

// Add the AzureAI.Community.SK.Plugin.TwilioWhatsApp to the Kernel
kernel.Plugins.AddFromObject(new WhatsAppServicePlugin(accountSid, authToken, fromPhoneNo));

// Send a WhatsApp message
string prompt = $"Please inform {toPhoneNo} that the office will be closed tomorrow";
var openAiPromptSettings = new OpenAIPromptExecutionSettings()
{
    ToolCallBehavior = ToolCallBehavior.AutoInvokeKernelFunctions
};

// Chat Completion Service
var chatCompletionService = kernel.GetRequiredService<IChatCompletionService>();

// Chat History
var chatHistory = new ChatHistory();
chatHistory.AddSystemMessage("You are an AI assistant.");
chatHistory.AddUserMessage(prompt);

var result = await chatCompletionService.GetChatMessageContentAsync(chatHistory, openAiPromptSettings, kernel);

Console.WriteLine(result.Content);