portia.builder.plan_builder_v2

Builder for Portia plans.

PlanBuilderError Objects

class PlanBuilderError(ValueError)

Error in Plan definition.

PlanBuilderV2 Objects

class PlanBuilderV2()

Builder for Portia plans.

__init__

def __init__(label: str = "Run the plan built with the Plan Builder") -> None

Initialize the builder.

Arguments:

• label - The label of the plan. This is used to identify the plan in the Portia dashboard.

input

def input(*,
          name: str,
          description: str | None = None,
          default_value: Any | None = None) -> PlanBuilderV2

Add an input to the plan.

Arguments:

• name - The name of the input.
• description - The description of the input.
• default_value - The default value of the input.

if_

def if_(condition: Callable[..., bool] | str,
        args: dict[str, Any] | None = None) -> PlanBuilderV2

Add a step that checks a condition.

else_if_

def else_if_(condition: Callable[..., bool],
             args: dict[str, Any] | None = None) -> PlanBuilderV2

Add a step that checks a condition.

else_

def else_() -> PlanBuilderV2

Add a step that checks a condition.

endif

def endif() -> PlanBuilderV2

Exit a conditional block.

llm_step

def llm_step(*,
             task: str,
             inputs: list[Any] | None = None,
             output_schema: type[BaseModel] | None = None,
             step_name: str | None = None) -> PlanBuilderV2

Add a step that sends a query to the underlying LLM.

Arguments:

• task - The task to perform.
• inputs - The inputs to the task. The inputs can be references to previous step outputs / plan inputs (using StepOutput / Input) or just plain values. They are passed in as additional context to the LLM when it is completing the task.
• output_schema - The schema of the output.
• step_name - Optional name for the step. If not provided, will be auto-generated.

invoke_tool_step

def invoke_tool_step(*,
                     tool: str | Tool,
                     args: dict[str, Any] | None = None,
                     output_schema: type[BaseModel] | None = None,
                     step_name: str | None = None) -> PlanBuilderV2

Add a step that directly invokes a tool.

Arguments:

• tool - The tool to invoke. Should either be the id of the tool to call, the Tool instance to call, or a python function that should be called.
• args - The arguments to the tool. If any of these values are instances of StepOutput or Input, the corresponding values will be substituted in when the plan is run.
• output_schema - The schema of the output.
• step_name - Optional name for the step. If not provided, will be auto-generated.

function_step

def function_step(*,
                  function: Callable[..., Any],
                  args: dict[str, Any] | None = None,
                  output_schema: type[BaseModel] | None = None,
                  step_name: str | None = None) -> PlanBuilderV2

Add a step that directly invokes a function.

Arguments:

• function - The function to invoke.
• args - The arguments to the function. If any of these values are instances of StepOutput or Input, the corresponding values will be substituted in when the plan is run.
• output_schema - The schema of the output.
• step_name - Optional name for the step. If not provided, will be auto-generated.

single_tool_agent_step

def single_tool_agent_step(*,
                           tool: str,
                           task: str,
                           inputs: list[Any] | None = None,
                           output_schema: type[BaseModel] | None = None,
                           step_name: str | None = None) -> PlanBuilderV2

Add a step that uses the execution agent with a tool.

Arguments:

• tool - The tool to use.
• task - The task to perform.
• inputs - The inputs to the task. If any of these values are instances of StepOutput or Input, the corresponding values will be substituted in when the plan is run.
• output_schema - The schema of the output.
• step_name - Optional name for the step. If not provided, will be auto-generated.

final_output

def final_output(output_schema: type[BaseModel] | None = None,
                 summarize: bool = False) -> PlanBuilderV2

Set the final output of the plan.

Arguments:

• output_schema - The schema for the final output. If provided, an LLM will be used to coerce the output to this schema.
• summarize - Whether to summarize the final output. If True, a summary of the final output will be provided along with the value.

build

def build() -> PlanV2

Return the plan, ready to run.