Streaming Tool Calls problem in Observation

Question

In flux streaming request, ChatClientObservationContext and ToolCallingObservationContext do not have a parent-child inheritance relationship, perhaps due to multithreading considerations. ChatClientObservationContext has an important indicator, spring.ai.chat.client.conversation.id, which could be passed to ToolCallingObservationContext to print when a tool call occurs. However, in the current Spring AI version 1.1.0, this doesn't seem to be possible yet. Is there any solution for this?

However, for call request, ChatClientObservationContext and ToolCallingObservationContext do have a parent-child inheritance relationship. This allows me to obtain the parent class ChatClientObservationContext, and thus the spring.ai.chat.client.conversation.id can be obtained.

public static ChatClientObservationContext findParentClientContext(Observation.ContextView ctx) {
        if (ctx == null) {
            return null;
        }
        try {
            ObservationView current = ctx.getParentObservation();
            while (current != null) {
                Observation.ContextView parentCtx = current.getContextView();
                if (parentCtx instanceof ChatClientObservationContext clientCtx) {
                    return clientCtx;
                }
                current = parentCtx.getParentObservation();
            }
        }catch (Throwable ignore) {
        }
        return null;
    }


public class ToolObservationHandler implements ObservationHandler<ToolCallingObservationContext> {
    @Override
    public void onStop(ToolCallingObservationContext context) {
           ChatClientObservationContext clientCtx = findParentClientContext(context);
                      // note: for flux streaming request, clientCtx will always be null
                      if (clientCtx != null) {
                          String conversationId = Optional.ofNullable(
                                          clientCtx.getHighCardinalityKeyValue(
                                                  ChatClientObservationDocumentation.HighCardinalityKeyNames.CHAT_CLIENT_CONVERSATION_ID.asString()))
                                  .map(KeyValue::getValue)
                                  .orElse("");
                      }
      }
}


// flux streaming request
 public Flux<String> flux(String message, Long chatId) {
        AtomicBoolean hasSentSeparator = new AtomicBoolean(false);

        return chatClient
                .prompt(promptService.getDefault())
                .user(message)
                .toolCallbacks(toolCallbackProvider.getToolCallbacks())
                .tools(localToolsService)
                .advisors(a -> a.param(ChatMemory.CONVERSATION_ID, "abc-1"))
                .stream()
                .content().contextWrite(ctx -> ctx.put(ChatMemory.CONVERSATION_ID, "abc-1"));
    }

// `call` request
  public String chat(String message, Long chatId) {
        return chatClient
                .prompt(promptService.getDefault())
                .user(message)
                .toolCallbacks(toolCallbackProvider.getToolCallbacks())
                .tools(localToolsService)
                .advisors(a -> a.param(ChatMemory.CONVERSATION_ID, "abc-1"))
                .call()
                .content();
    }

Start with ChatClientObservationContext, ToolCallingObservationContext, and the ObservationHandler.onStop entry point, comparing the provided Flux streaming and call examples. Trace how parent observations and the conversation ID are handled in each path; done should be a defined, verified outcome for conversation-ID visibility during streaming tool calls.