Text to Image: Document integration into the AI Chat Web template

[jeffhandley]

We do not plan to integrate Text to Image directly into the AI Chat Web template, but we do need to document how it can be achieved. This issue will be resolved by capturing a how-to document with code samples showing how an existing app created with the template can be updated to integrate this modality.

Completion of this documented integration is an acceptance criterion for finishing #6638.

[jeffhandley]

@ericstj Before this integration is implemented for productization, I'm going to discuss the template strategy with @JeremyLikness, @luisquintanilla, and @stephentoub to determine which modalities we want to integrate into the template vs. which ones we want to leave for docs/samples that build on top of the template.

[jeffhandley]

@ericstj I updated the issue description to capture that we only intend to document this rather than directly integrating into the template itself.

[ericstj]

Got it, I have this sample -- https://github.com/ericstj/textToImageSample that can help with the docs. Still waiting on more feedback on that and the PR though. We shouldn't write any docs until things get finalized.

[ericstj]

Refactored the changes from that project into a single commit ericstj/imageGeneratorSample@8a97699

[JeremyLikness]

@jeffhandley @ericstj who will be responsible for driving this documentation? Is this something you've already coordinated with the docs team or are you looking to me to drive this?

[ericstj]

I don't think we've done this yet. We still have to decide how to do it. I think now that we have the ImageGeneratingChatClient that might be the way to show this.

It would be good to have a sample that already natively supports the ImageGenerationTool (so we need chat template to support responses API) then we an also show how to use ImageGeneratingChatClient to add support for image generation if your chat client doesn't natively support it.

cc @gewarren