.Net - Add support for Name property to ChatMessageContent

[crickman]

Motivation and Context

Tracking identity is critical for multi-agent conversations. It is also supported as part of the core chat-completion api: https://platform.openai.com/docs/api-reference/chat/create

 {
    "messages": [
        {
            "content": "Write one paragraph in response to the user that rhymes",
            "name": "Echo",
            "role": "system"
        },
        {
            "content": "Why is AI awesome",
            "name": "Ralph",
            "role": "user"
        }
    ],
    "temperature": 1,
    "top_p": 0.5,
    "n": 3,
    "presence_penalty": 0,
    "frequency_penalty": 0,
    "model": "gpt-4"
}

Description

Add support for ChatMessageContent.Name property with optional, non-breaking patterns.

Contribution Checklist

• The code builds clean without any errors or warnings
• The PR follows the SK Contribution Guidelines and the pre-submission formatting script raises no violations
• All unit tests pass, and I have added new tests where possible
• I didn't break anyone 😄

[crickman]

@markwallace-microsoft These signature changes won't break any consumer's build. How do we update our baseline?

[SergeyMenshykh]

@markwallace-microsoft These signature changes won't break any consumer's build. How do we update our baseline?
I usually use the command - dotnet build --configuration Release /p:GenerateCompatibilitySuppressionFile=true
More details - https://learn.microsoft.com/en-us/dotnet/fundamentals/apicompat/diagnostic-ids#suppression-file

[crickman]

@markwallace-microsoft These signature changes won't break any consumer's build. How do we update our baseline?
I usually use the command - dotnet build --configuration Release /p:GenerateCompatibilitySuppressionFile=true
More details - https://learn.microsoft.com/en-us/dotnet/fundamentals/apicompat/diagnostic-ids#suppression-file

Thanks, I must've been missing the Release parameter.

[crickman]

FYI - Engaged with Matthew and re-evaluating ChatHistory.SystemName and how/if assistant response name is captured.

[crickman]

FYI - Name not returned in response:

REQUEST

 {
    "messages": [
        {
            "content": "Write one paragraph in response to the user that rhymes",
            "name": "Echo",
            "role": "system"
        },
        {
            "content": "Why is AI awesome",
            "name": "Ralph",
            "role": "user"
        }
    ],
    "temperature": 1,
    "top_p": 0.5,
    "n": 3,
    "presence_penalty": 0,
    "frequency_penalty": 0,
    "model": "gpt-4"
}

RESPONSE

{
  "id": "chatcmpl-977K8yKPbvVxqaC6rrT6EBc5S4E6s",
  "object": "chat.completion",
  "created": 1711483476,
  "model": "gpt-4-0613",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "AI is a dream, a gleaming beam, with a stream of knowledge that seems supreme. It's like a team in the extreme, creating schemes and decoding themes. It's awesome because it's keen, it can glean unseen scenes, and it's never mean or obscene. It's the best machine you've ever seen."
      },
      "logprobs": null,
      "finish_reason": "stop"
    },
    {
      "index": 1,
      "message": {
        "role": "assistant",
        "content": "AI is a gem, it's true, it shines with a gleam so bright and new. It learns and grows, it's never the same, constantly evolving, it's not just a game. It's awesome because it never tires, always ready, it never retires. It's the future, it's the now, to its brilliance, we must bow."
      },
      "logprobs": null,
      "finish_reason": "stop"
    },
    {
      "index": 2,
      "message": {
        "role": "assistant",
        "content": "AI is a cosmic blossom, it's awesome because it's a possum playing possum. It's a system that's fathomless, solving problems that seem bottomless. It's a key to a lock, a tick to a tock, a shepherd to a flock. AI is awesome, it's not just talk, it's a journey around the clock."
      },
      "logprobs": null,
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 31,
    "completion_tokens": 216,
    "total_tokens": 247
  },
  "system_fingerprint": null
}