def do_test(user_id: str) -> bool:
    """
    runs the test and returns whether it was successful or not
    """
    return user_id == "f6773f16854cca0f871d10e7cf6c7e84d12f3783bcfa13ccba8d2d0e48ee7cbd"

function_tool = FunctionTool.from_defaults(fn=do_test)
tools = [FunctionTool.from_defaults(fn=f) for f in (do_test,)]
llm = OpenAIMultiModal(model="gpt-4-vision-preview", api_key=OPENAI_API_KEY)
react_step_engine = MultimodalReActAgentWorker.from_tools(tools, llm=llm, verbose=True)
agent = AgentRunner(react_step_engine)

def execute_step(agent: AgentRunner, task: Task):
    step_output = agent.run_step(task.task_id)
    if step_output.is_last:
        response = agent.finalize_response(task.task_id)
        return response
    else:
        return None


def execute_steps(agent: AgentRunner, task: Task):
    response = execute_step(agent, task)
    while response is None:
        response = execute_step(agent, task)
    return response

task = agent.create_task(
    "User ID: f6773f16854cca0f871d10e7cf6c7e84d12f3783bcfa13ccba8d2d0e48ee7cbd\nrun the first test and if it fails describe the image",
    extra_state={"image_docs": [ImageDocument(image_path="/home/finn/Pictures/rhino.jpeg")]},
)

response = execute_steps(agent, task)

task2 = agent.create_task(
    "what is the first word i said to you",
    extra_state={"image_docs": []},
)

response = execute_steps(agent, task2)