🌟 Semantic Kernel Intro Agents 🌟

Welcome to the Semantic Kernel Intro Agents project! This application demonstrates how to build interactive, intelligent agents using Semantic Kernel and Azure OpenAI. 🎉 With plugins like NewsPlugin for fetching the latest news and ArchivePlugin for data storage, this project showcases extensible and powerful AI functionality for real-world applications. 🚀

---

🚀 Features

• 🤖 Interactive AI Chat: Engage with an AI agent in real-time, powered by Azure OpenAI's GPT models.
• 📰 News Retrieval: Stay informed with the latest news headlines using the NewsPlugin.
• 💾 Data Archival: Save conversation history or other data locally with the ArchivePlugin.
• 🔧 Extensibility: Designed for flexibility—easily add custom plugins or extend existing ones.
• 📜 Real-Time Streaming Responses: Enjoy smooth, streaming responses during chat interactions.

---

📋 Requirements

Before running the project, ensure you have the following:

• ✅ .NET 8.0 SDK: Download here.
• ✅ Azure OpenAI Credentials:
  • Deployment Name: The name of your deployed Azure OpenAI model.
  • Endpoint URL: The endpoint for your Azure OpenAI service.
  • API Key: The authentication key for accessing Azure OpenAI.
• ✅ Configuration:
  • Populate the appsettings.json file with your Azure OpenAI credentials.

---

🛠️ How It Works

Program Overview

1. Configuration Loading:

   • The application reads Azure OpenAI credentials (DeploymentName, Endpoint, ApiKey) from appsettings.json for secure configuration management.

2. Semantic Kernel Setup:

   • A Kernel Builder is initialized to configure the Semantic Kernel environment.
   • The Azure OpenAI service is integrated as a chat completion backend.

3. Plugins:

   • The application integrates plugins to expand functionality:
     • NewsPlugin: Retrieves news headlines from the New York Times RSS feeds.
     • ArchivePlugin: Saves runtime data to local storage for later use.

4. Interactive Loop:

   • Users input prompts, and the AI agent responds in real-time.
   • Chat history is maintained to ensure context-aware conversations.

5. Streaming Responses:

   • The AI responses are streamed for a smoother user experience, displaying the output as it is generated.

---

📂 Project Structure

C:\Users\rsung\OneDrive\Desktop\AIResearch\semantickernel\sk-intro-agents
├── .gitignore                       # Git ignore rules
├── README.md                        # Project documentation
├── sk-intro-agents.sln              # Solution file
├── notebooks/                       # Jupyter notebooks (optional)
├── src/                             # Source code directory
│   ├── appsettings.Development.json # Environment-specific config
│   ├── appsettings.json             # Main configuration file
│   ├── ArchivePlugin.cs             # Plugin for data archival
│   ├── NewsPlugin.cs                # Plugin for fetching news
│   ├── Program.cs                   # Main entry point
│   ├── sk-intro-agents.csproj       # Project configuration
│   ├── archives/                    # Archived runtime-generated data
│   ├── bin/                         # Build output (auto-generated)
│   └── obj/                         # Build object files (auto-generated)

🔌 Plugins

📰 NewsPlugin

• Purpose: Fetches top news articles from the New York Times RSS feed for a specified category.
• Key Features:
  • Retrieves up to 10 news items.
  • Uses the SimpleFeedReader library for RSS feed parsing.
• Sample Input: tech
• Sample Output:

  1. TikTok Faces U.S. Ban After Losing Bid to Overturn New Law
  2. Bitcoin Price Surges to a Milestone: $100,000

💾 ArchivePlugin

• Purpose: Saves conversation data or any runtime information to local storage.

• Key Features:

• Writes data into text files in the archives/ folder.

• Useful for maintaining logs or preserving user interactions.

Output Example:

Data is saved as C:\archives<filename>.txt.

🔧 Configuration Details:

Edit the appsettings.json file in the src/ directory to include your Azure OpenAI credentials:

{
  "AzureOpenAI": {
    "DeploymentName": "your-deployment-name",
    "Endpoint": "https://your-endpoint.azure.com",
    "ApiKey": "your-api-key"
  }
}

🖥️ How to Run

1. Clone the Repository

git clone <repository-url> cd sk-intro-agents

2. Restore Dependencies

dotnet restore

3. Build the Project

dotnet build

4. Run the Application

dotnet run

🖱️ Interact with the AI

• Prompt: Enter hi or tech news please.

• Response: The AI will respond in real-time with interactive, context-aware replies.

Prompt: hi
Response: Hello! How can I assist you today?

Prompt: tech news please
Response: Here are some of the latest updates in the technology sector:
1. TikTok Faces U.S. Ban After Losing Bid to Overturn New Law
2. Bitcoin Price Surges to a Milestone: $100,000

🎉 Enjoy exploring the capabilities of Semantic Kernel Intro Agents! 🚀