Python: [Feature]: Support ChatKit generated image conversion #8842

Description

agent-framework-chatkit recognizes GeneratedImageItem in ThreadItemConverter, but the dispatch path currently drops it with return [] and leaves a TODO for future handling. This means generated images stored in ChatKit thread history are not provided back to the model on later turns.

The expected behavior is to convert a completed generated image into user context containing a short text preface and the image URL, while continuing to skip items that do not yet contain an image. This matches the default conversion semantics in openai-chatkit.

I am preparing a narrow implementation limited to the ChatKit converter and focused tests.