DroidAgent - A wrapper class that coordinates the planning and execution of tasks to achieve a user’s goal on an Android device.

DroidAgent

class DroidAgent(Workflow)

A wrapper class that coordinates between PlannerAgent (creates plans) and CodeActAgent (executes tasks) to achieve a user’s goal.

DroidAgent.__init__

def __init__(
    goal: str,
    llm: LLM,
    tools: Tools,
    personas: List[AgentPersona] = [DEFAULT],
    max_steps: int = 15,
    timeout: int = 1000,
    vision: bool = False,
    reasoning: bool = False,
    reflection: bool = False,
    enable_tracing: bool = False,
    debug: bool = False,
    save_trajectories: bool = False,
    *args,
    **kwargs
)

Initialize the DroidAgent wrapper.

Arguments:

  • goal - The user’s goal or command to execute
  • llm - The language model to use for both agents
  • max_steps - Maximum number of steps for both agents
  • timeout - Timeout for agent execution in seconds
  • reasoning - Whether to use the PlannerAgent for complex reasoning (True) or send tasks directly to CodeActAgent (False)
  • reflection - Whether to reflect on steps the CodeActAgent did to give the PlannerAgent advice
  • enable_tracing - Whether to enable Arize Phoenix tracing
  • debug - Whether to enable verbose debug logging
  • **kwargs - Additional keyword arguments to pass to the agents

DroidAgent.run

def run() -> WorkflowHandler

Run the DroidAgent workflow.

DroidAgent - A wrapper class that coordinates the planning and execution of tasks to achieve a user’s goal on an Android device.

DroidAgent

class DroidAgent(Workflow)

A wrapper class that coordinates between PlannerAgent (creates plans) and CodeActAgent (executes tasks) to achieve a user’s goal.

DroidAgent.__init__

def __init__(
    goal: str,
    llm: LLM,
    tools: Tools,
    personas: List[AgentPersona] = [DEFAULT],
    max_steps: int = 15,
    timeout: int = 1000,
    vision: bool = False,
    reasoning: bool = False,
    reflection: bool = False,
    enable_tracing: bool = False,
    debug: bool = False,
    save_trajectories: bool = False,
    *args,
    **kwargs
)

Initialize the DroidAgent wrapper.

Arguments:

  • goal - The user’s goal or command to execute
  • llm - The language model to use for both agents
  • max_steps - Maximum number of steps for both agents
  • timeout - Timeout for agent execution in seconds
  • reasoning - Whether to use the PlannerAgent for complex reasoning (True) or send tasks directly to CodeActAgent (False)
  • reflection - Whether to reflect on steps the CodeActAgent did to give the PlannerAgent advice
  • enable_tracing - Whether to enable Arize Phoenix tracing
  • debug - Whether to enable verbose debug logging
  • **kwargs - Additional keyword arguments to pass to the agents

DroidAgent.run

def run() -> WorkflowHandler

Run the DroidAgent workflow.