Semantic Kernel Starters

This starter project provides the scaffolding to create a ChatGPT plugin with the following capabilities:

• Plugin Discovery: An endpoint that serves the ai-plugin.json file required for ChatGPT to discover your plugin.
• Semantic Function Generation: A generator that automatically converts prompts into semantic function endpoints.
• Native Function Support: The ability to add additional native functions as endpoints to the plugin.

[IMPORTANT]

This repository is being consolidated. It is planned to be removed in the near future. For the most up-to-date and official samples, please use the following links instead of this repository:

• .NET Samples: https://github.com/microsoft/semantic-kernel/tree/main/dotnet/samples
• Python Samples: https://github.com/microsoft/semantic-kernel/tree/main/python/samples
• Java Samples: https://github.com/microsoft/semantic-kernel-java/tree/main/samples

The Azure Functions Python V2 programming model uses a decorator-based approach that aligns with standard Python frameworks.

Key characteristics:

• No function.json required: Triggers and bindings are defined directly in the code using decorators.
• File Organization: You can define multiple functions in a single file or organize them across multiple files using Blueprints.
• Main Entry Point: The primary functions file must be named function_app.py.
• Constraint: You cannot mix V1 and V2 programming models within the same Function App.

For detailed technical specifications, refer to the Azure Functions Python developer guide.

To begin working with the declarative process for generating product documentation, open the product-documentation.process.yaml file in Visual Studio Code.

Recommendation: It is highly recommended to use the Semantic Kernel Tools VS Code extension to interact with these process files effectively.

# Open this file in VS Code to start working with processes
product-documentation.process.yaml

You can run the application in two ways:

1. Via IDE: Press F5 in your development environment.
2. Via Terminal: Build the project and use the Azure Functions Core Tools to start it.

dotnet build
func start --csharp

To set up the environment for the Process Framework example, ensure you have .NET 8 installed. Follow these steps to build the project and configure your Azure OpenAI credentials using .NET user secrets:

1. Build the process node logic:

   dotnet build

2. Configure your Azure OpenAI deployment name:

   dotnet user-secrets set "AZUREOPENAI_DEPLOYMENT_NAME" "YOUR_DEPLOYMENT_NAME"

3. Configure your Azure OpenAI endpoint:

   dotnet user-secrets set "AZUREOPENAI_ENDPOINT" "YOUR_ENDPOINT"

4. Authenticate with Azure to access the deployment:

   az login

To manually set up and run the Azure Functions host from your terminal:

1. Create a virtual environment.
2. Install requirements.
3. Activate the environment.
4. Start the Azure Functions host.

python -m venv .venv
.venv\Scripts\python -m pip install -r requirements.txt # Note: Path to python depends on your OS
.venv\Scripts\activate
func host start

Before running this starter, ensure you have the following installed and configured:

• Python: Version >=3.8 and <3.11
• Azure Functions Core Tools: For running the function host.
• Azurite: An Azure Storage API emulator. You must run the services from the VS Code status bar.
• Semantic Kernel Tools: VS Code extension for Semantic Kernel.

To use Azure OpenAI, set the following keys via .NET Secret Manager:

• serviceType: Set to AzureOpenAI.
• serviceId: The service identifier.
• deploymentId: Your Azure OpenAI deployment ID.
• modelId: The model ID.
• endpoint: Your Azure OpenAI endpoint URL.
• apiKey: Your Azure OpenAI API key.

cd sk-csharp-console-chat
dotnet user-secrets set "serviceType" "AzureOpenAI"
dotnet user-secrets set "serviceId" "gpt-35-turbo"
dotnet user-secrets set "deploymentId" "gpt-35-turbo"
dotnet user-secrets set "modelId" "gpt-3.5-turbo"
dotnet user-secrets set "endpoint" "https:// ... your endpoint ... .openai.azure.com/"
dotnet user-secrets set "apiKey" "... your Azure OpenAI key ..."