​
overview

the Scenario class represents a test scenario that defines what the test agent should do when calling your voice agent. it includes a prompt for the agent’s behavior and a set of evaluation criteria to assess the interaction.

​
constructor

Scenario(
    name: str,
    prompt: str,
    evaluations: List[Evaluation] = []
)

​
parameters

  • name (str): the name of the scenario
  • prompt (str): the system prompt that defines what the agent should do in this scenario
  • evaluations (List[Evaluation], optional): list of evaluation criteria to assess the scenario. defaults to an empty list.

​
example usage

scenario = Scenario(
    name="order_donut",
    prompt="order a dozen donuts with sprinkles and a coffee",
    evaluations=[
        Evaluation(name="order_success", prompt="the order was successful"),
        Evaluation(name="price_confirmed", prompt="the agent confirmed the price of the order"),
    ]
)

​
notes

  • the scenario prompt should clearly define the task or interaction you want to test
  • each scenario can have multiple evaluation criteria to assess different aspects of the interaction
  • scenarios are used in combination with agents to create tests