handoffs.py

from __future__ import annotations

import inspect
from collections.abc import Awaitable
from dataclasses import dataclass
from typing import TYPE_CHECKING, Any, Callable, Generic, cast, overload

from pydantic import TypeAdapter
from typing_extensions import TypeAlias, TypeVar

from . import _utils
from .exceptions import ModelBehaviorError, UserError
from .items import RunItem, TResponseInputItem
from .run_context import RunContextWrapper, TContext
from .strict_schema import ensure_strict_json_schema
from .tracing.spans import SpanError

if TYPE_CHECKING:
    from .agent import Agent


# The handoff input type is the type of data passed when the agent is called via a handoff.
THandoffInput = TypeVar("THandoffInput", default=Any)

OnHandoffWithInput = Callable[[RunContextWrapper[Any], THandoffInput], Any]
OnHandoffWithoutInput = Callable[[RunContextWrapper[Any]], Any]


@dataclass(frozen=True)
class HandoffInputData:
    input_history: str | tuple[TResponseInputItem, ...]
    """
    The input history before `Runner.run()` was called.
    """

    pre_handoff_items: tuple[RunItem, ...]
    """
    The items generated before the agent turn where the handoff was invoked.
    """

    new_items: tuple[RunItem, ...]
    """
    The new items generated during the current agent turn, including the item that triggered the
    handoff and the tool output message representing the response from the handoff output.
    """


HandoffInputFilter: TypeAlias = Callable[[HandoffInputData], HandoffInputData]
"""A function that filters the input data passed to the next agent."""


@dataclass
class Handoff(Generic[TContext]):
    """A handoff is when an agent delegates a task to another agent.
    For example, in a customer support scenario you might have a "triage agent" that determines
    which agent should handle the user's request, and sub-agents that specialize in different
    areas like billing, account management, etc.
    """

    tool_name: str
    """The name of the tool that represents the handoff."""

    tool_description: str
    """The description of the tool that represents the handoff."""

    input_json_schema: dict[str, Any]
    """The JSON schema for the handoff input. Can be empty if the handoff does not take an input.
    """

    on_invoke_handoff: Callable[[RunContextWrapper[Any], str], Awaitable[Agent[TContext]]]
    """The function that invokes the handoff. The parameters passed are:
    1. The handoff run context
    2. The arguments from the LLM, as a JSON string. Empty string if input_json_schema is empty.

    Must return an agent.
    """

    agent_name: str
    """The name of the agent that is being handed off to."""

    input_filter: HandoffInputFilter | None = None
    """A function that filters the inputs that are passed to the next agent. By default, the new
    agent sees the entire conversation history. In some cases, you may want to filter inputs e.g.
    to remove older inputs, or remove tools from existing inputs.

    The function will receive the entire conversation history so far, including the input item
    that triggered the handoff and a tool call output item representing the handoff tool's output.

    You are free to modify the input history or new items as you see fit. The next agent that
    runs will receive `handoff_input_data.all_items`.

    IMPORTANT: in streaming mode, we will not stream anything as a result of this function. The
    items generated before will already have been streamed.
    """

    strict_json_schema: bool = True
    """Whether the input JSON schema is in strict mode. We **strongly** recommend setting this to
    True, as it increases the likelihood of correct JSON input.
    """

    def get_transfer_message(self, agent: Agent[Any]) -> str:
        base = f"{{'assistant': '{agent.name}'}}"
        return base

    @classmethod
    def default_tool_name(cls, agent: Agent[Any]) -> str:
        return _utils.transform_string_function_style(f"transfer_to_{agent.name}")

    @classmethod
    def default_tool_description(cls, agent: Agent[Any]) -> str:
        return (
            f"Handoff to the {agent.name} agent to handle the request. "
            f"{agent.handoff_description or ''}"
        )


@overload
def handoff(
    agent: Agent[TContext],
    *,
    tool_name_override: str | None = None,
    tool_description_override: str | None = None,
    input_filter: Callable[[HandoffInputData], HandoffInputData] | None = None,
) -> Handoff[TContext]: ...


@overload
def handoff(
    agent: Agent[TContext],
    *,
    on_handoff: OnHandoffWithInput[THandoffInput],
    input_type: type[THandoffInput],
    tool_description_override: str | None = None,
    tool_name_override: str | None = None,
    input_filter: Callable[[HandoffInputData], HandoffInputData] | None = None,
) -> Handoff[TContext]: ...


@overload
def handoff(
    agent: Agent[TContext],
    *,
    on_handoff: OnHandoffWithoutInput,
    tool_description_override: str | None = None,
    tool_name_override: str | None = None,
    input_filter: Callable[[HandoffInputData], HandoffInputData] | None = None,
) -> Handoff[TContext]: ...


def handoff(
    agent: Agent[TContext],
    tool_name_override: str | None = None,
    tool_description_override: str | None = None,
    on_handoff: OnHandoffWithInput[THandoffInput] | OnHandoffWithoutInput | None = None,
    input_type: type[THandoffInput] | None = None,
    input_filter: Callable[[HandoffInputData], HandoffInputData] | None = None,
) -> Handoff[TContext]:
    """Create a handoff from an agent.

    Args:
        agent: The agent to handoff to, or a function that returns an agent.
        tool_name_override: Optional override for the name of the tool that represents the handoff.
        tool_description_override: Optional override for the description of the tool that
            represents the handoff.
        on_handoff: A function that runs when the handoff is invoked.
        input_type: the type of the input to the handoff. If provided, the input will be validated
            against this type. Only relevant if you pass a function that takes an input.
        input_filter: a function that filters the inputs that are passed to the next agent.
    """
    assert (on_handoff and input_type) or not (on_handoff and input_type), (
        "You must provide either both on_input and input_type, or neither"
    )
    type_adapter: TypeAdapter[Any] | None
    if input_type is not None:
        assert callable(on_handoff), "on_handoff must be callable"
        sig = inspect.signature(on_handoff)
        if len(sig.parameters) != 2:
            raise UserError("on_handoff must take two arguments: context and input")

        type_adapter = TypeAdapter(input_type)
        input_json_schema = type_adapter.json_schema()
    else:
        type_adapter = None
        input_json_schema = {}
        if on_handoff is not None:
            sig = inspect.signature(on_handoff)
            if len(sig.parameters) != 1:
                raise UserError("on_handoff must take one argument: context")

    async def _invoke_handoff(
        ctx: RunContextWrapper[Any], input_json: str | None = None
    ) -> Agent[Any]:
        if input_type is not None and type_adapter is not None:
            if input_json is None:
                _utils.attach_error_to_current_span(
                    SpanError(
                        message="Handoff function expected non-null input, but got None",
                        data={"details": "input_json is None"},
                    )
                )
                raise ModelBehaviorError("Handoff function expected non-null input, but got None")

            validated_input = _utils.validate_json(
                json_str=input_json,
                type_adapter=type_adapter,
                partial=False,
            )
            input_func = cast(OnHandoffWithInput[THandoffInput], on_handoff)
            if inspect.iscoroutinefunction(input_func):
                await input_func(ctx, validated_input)
            else:
                input_func(ctx, validated_input)
        elif on_handoff is not None:
            no_input_func = cast(OnHandoffWithoutInput, on_handoff)
            if inspect.iscoroutinefunction(no_input_func):
                await no_input_func(ctx)
            else:
                no_input_func(ctx)

        return agent

    tool_name = tool_name_override or Handoff.default_tool_name(agent)
    tool_description = tool_description_override or Handoff.default_tool_description(agent)

    # Always ensure the input JSON schema is in strict mode
    # If there is a need, we can make this configurable in the future
    input_json_schema = ensure_strict_json_schema(input_json_schema)

    return Handoff(
        tool_name=tool_name,
        tool_description=tool_description,
        input_json_schema=input_json_schema,
        on_invoke_handoff=_invoke_handoff,
        input_filter=input_filter,
        agent_name=agent.name,
    )