.Net: Add tool_call_id to tool result messages in model diagnostics

[dsorokin-st]

Motivation and Context

Tool result messages in model diagnostics are missing the tool_call_id property, making it difficult to correlate tool results with their corresponding tool calls in observability tools.

Current behavior - tool result messages only have role and content:

{"role": "tool", "name": null, "content": "Partly Cloudy, 22°C", "tool_calls": []}

After this change - includes tool_call_id for correlation:

{"role": "tool", "name": "get_weather", "tool_call_id": "call_abc123", "content": "Partly Cloudy, 22°C", "tool_calls": []}

This aligns with the OpenAI API format and enables observability tools (OpenInference, Arize, Galileo) to properly correlate tool calls with their results.

Description

Updated ToGenAIConventionsFormat in ModelDiagnostics.cs to:

1. Check for FunctionResultContent in message items
2. Add tool_call_id from FunctionResultContent.CallId
3. Use FunctionResultContent.FunctionName for the name field (instead of AuthorName which is always null for tool messages)

This is a backward-compatible, additive change - existing consumers will simply ignore the new field.

Why Unit Testing Is Difficult

Unit testing this change is challenging due to ModelDiagnostics architecture:

1. Static readonly field caching: The feature flags (s_enableDiagnostics, s_enableSensitiveEvents) are static readonly fields initialized at type load time via AppContextSwitchHelper.GetConfigValue().

2. No test isolation: Once ModelDiagnostics is loaded by any test in the assembly, the flag values are permanently cached. Setting AppContext.SetSwitch or environment variables after type initialization has no effect.

3. Reflection blocked: .NET prevents modification of initonly static fields after type initialization:

   System.FieldAccessException: Cannot set initonly static field 's_enableDiagnostics' after type 'ModelDiagnostics' is initialized.
   

4. Existing precedent: The test GetInvalidResponseThrowsExceptionAndIsCapturedByDiagnosticsAsync in OpenAIChatCompletionServiceTests.cs is already skipped with [Fact(Skip = "Not working running in the console")] for the same reason.

The change has been manually verified to produce the correct output.

Contribution Checklist

• The code builds clean without any errors or warnings
• The PR title follows the format: .Net: <description>
• I have added tests - See explanation above for why unit testing is impractical
• I have updated the documentation (not applicable)

[dsorokin-st]

@microsoft-github-policy-service agree company="ServiceTitan"