Add chat system prompt option for text completion (#1465)

This change adds a new property to the CompleteRequestSettings class
that allows specifying a custom system prompt to use when generating
text completions using a chat model. This can be useful for setting the
tone and context of the conversation. The default value is "Assistant is
a large language model." which is the same as the OpenAI API.

The change also modifies the ClientBase class to pass the chat system
prompt to the InternalCreateNewChat method, which creates a new instance
of the OpenAIChatHistory class. This way, the chat history will include
the system prompt as the first message.

Co-authored-by: Lee Miller <[email protected]>

dotnet/src/Connectors/Connectors.AI.OpenAI/AzureSdk/ClientBase.cs

@@ -242,7 +242,7 @@ private protected async IAsyncEnumerable<ITextStreamingResult> InternalGetChatSt
     private static OpenAIChatHistory PrepareChatHistory(string text, CompleteRequestSettings? requestSettings, out ChatRequestSettings settings)
     {
         requestSettings ??= new();
-        var chat = InternalCreateNewChat();
+        var chat = InternalCreateNewChat(requestSettings.ChatSystemPrompt);
         chat.AddUserMessage(text);
         settings = new ChatRequestSettings
         {

dotnet/src/SemanticKernel.Abstractions/AI/TextCompletion/CompleteRequestSettings.cs

@@ -54,6 +54,12 @@ public class CompleteRequestSettings
     /// </summary>
     public int ResultsPerPrompt { get; set; } = 1;
 
+    /// <summary>
+    /// The system prompt to use when generating text completions using a chat model.
+    /// Defaults to "Assistant is a large language model."
+    /// </summary>
+    public string ChatSystemPrompt { get; set; } = "Assistant is a large language model.";
+
     /// <summary>
     /// Create a new settings object with the values from another settings object.
     /// </summary>