displayText from second simple response ignored in simulator

Question

EDIT: this was a bug in the simulator and the bug has apparently been fixed.

I'm trying to build a Google Assistant action (skill) using Python code and the REST API. According to the docs, you can post a rich response containing one or two simple responses. When I do this, however, the simulator ignores the displayText from the second simple response and displays only the first one. Unlike text, voice is taken from both simple responses. Here is the structure I post:

{
    "conversationToken": "",
    "expectUserResponse": True,
    "expectedInputs": [
        {
            "inputPrompt": {
                "richInitialPrompt": {
                    "items": [
                        {
                            "simpleResponse": {
                                "textToSpeech": "First speech.",
                                "displayText": "First text."
                            }
                        },
                        {
                            "simpleResponse": {
                                "textToSpeech": "Second speech.",
                                "displayText": "Second text."
                            }
                        },
                    ],
                    "suggestions": []
                }
            },
            "possibleIntents": [
                {
                    "intent": "actions.intent.TEXT"
                }
            ]
        }
    ]
}