agents.conversation.directed.agent¶

Directed conversation agent where participants respond to mentions and direct questions. from typing import Any Uses structured output models for robust speaker selection and interaction tracking.

Classes¶

`DirectedConversation`	Directed conversation where agents respond to mentions and questions.
`DirectedConversationConfig`	Configuration for directed conversation behavior.
`InteractionPattern`	Track interaction patterns between speakers.
`MentionType`	Types of mentions detected in messages.
`SpeakerMention`	Structured representation of a speaker mention.
`SpeakerSelectionResult`	Structured output for speaker selection logic.

Module Contents¶

class agents.conversation.directed.agent.DirectedConversation¶

Bases: haive.agents.conversation.base.agent.BaseConversationAgent

Directed conversation where agents respond to mentions and questions.

Uses structured output models for robust speaker selection and tracking. Participants speak when: - They are directly mentioned (@name) - A question is directed at them - They haven’t spoken in a while (configurable)

classmethod create_classroom(teacher_name='Teacher', student_names=None, topic="Today's lesson", config=None, **kwargs)¶

Create a classroom-style directed conversation.

Parameters:

teacher_name (str) – Name of the teacher
student_names (list[str] | None) – List of student names
topic (str) – Lesson topic
config (DirectedConversationConfig | None) – Optional configuration for directed conversation
**kwargs – Additional conversation arguments

get_conversation_state_schema()¶

Use extended state schema.

Return type:: type

process_response(state)¶

Track interaction patterns using structured models.

Parameters:: state (haive.agents.conversation.directed.state.DirectedState)
Return type:: dict[str, Any]

select_speaker(state)¶

Select speaker based on mentions and context using structured models.

Parameters:: state (haive.agents.conversation.directed.state.DirectedState)
Return type:: dict[str, Any]

class agents.conversation.directed.agent.DirectedConversationConfig(/, **data)¶

Bases: pydantic.BaseModel

Configuration for directed conversation behavior.

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Parameters:: data (Any)

class agents.conversation.directed.agent.InteractionPattern(/, **data)¶

Bases: pydantic.BaseModel

Track interaction patterns between speakers.

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Parameters:: data (Any)

class agents.conversation.directed.agent.MentionType¶

Bases: str, enum.Enum

Types of mentions detected in messages.

Initialize self. See help(type(self)) for accurate signature.

class agents.conversation.directed.agent.SpeakerMention(/, **data)¶

Bases: pydantic.BaseModel

Structured representation of a speaker mention.

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Parameters:: data (Any)

class agents.conversation.directed.agent.SpeakerSelectionResult(/, **data)¶

Bases: pydantic.BaseModel

Structured output for speaker selection logic.

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Parameters:: data (Any)