agents.memory_v2.multi_memory_agent

MultiMemoryAgent - Coordinates different memory strategies.

This agent acts as a meta-coordinator that routes queries to different specialized memory agents based on query type, context, and memory strategy optimization.

Classes

MemoryPriority	Priority levels for memory processing.
MemoryRoutingRule	Rule for routing queries to specific memory strategies.
MemoryStrategy	Types of memory strategies available.
MultiMemoryAgent	Agent that coordinates multiple memory strategies.
MultiMemoryConfig	Configuration for MultiMemoryAgent.
MultiMemoryState	Extended state for MultiMemoryAgent with routing information.
QueryClassifier	Classifies queries to determine appropriate memory strategy.
QueryType	Types of queries that determine memory routing.
ResponseSynthesizer	Synthesizes responses from multiple memory agents.

Functions

create_multi_memory_agent([name, enable_graph, enable_rag])

Factory function to create a MultiMemoryAgent with sensible defaults.

Module Contents

class agents.memory_v2.multi_memory_agent.MemoryPriority

Bases: str, enum.Enum

Priority levels for memory processing.

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

class agents.memory_v2.multi_memory_agent.MemoryRoutingRule

Rule for routing queries to specific memory strategies.

class agents.memory_v2.multi_memory_agent.MemoryStrategy

Bases: str, enum.Enum

Types of memory strategies available.

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

class agents.memory_v2.multi_memory_agent.MultiMemoryAgent(config)

Bases: haive.agents.simple.agent.SimpleAgent

Agent that coordinates multiple memory strategies.

This agent acts as a smart router and coordinator for different memory approaches, automatically selecting the best strategy based on query analysis and combining responses when appropriate.

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:

config (MultiMemoryConfig)

async classify_query(query)

Classify the query to determine appropriate memory strategy.

Parameters:

query (str)

Return type:

dict[str, Any]

async execute_strategy(strategy, query, context=None)

Execute the selected memory strategy.

Parameters:

• strategy (MemoryStrategy)

• query (str)

• context (dict[str, Any] | None)

Return type:

dict[str, Any]

get_comprehensive_status()

Get comprehensive status of the MultiMemoryAgent.

Return type:

dict[str, Any]

get_coordination_stats()

Get statistics about query coordination and routing.

Return type:

dict[str, Any]

async query_memory_agent(agent_key, query, context=None)

Query a specific memory agent.

Parameters:

• agent_key (str)

• query (str)

• context (dict[str, Any] | None)

Return type:

dict[str, Any]

route_query(query_type, confidence, context=None)

Route query to appropriate memory strategy based on classification.

Parameters:

• query_type (QueryType)

• confidence (float)

• context (dict[str, Any] | None)

Return type:

dict[str, Any]

async synthesize_responses(responses, query)

Synthesize multiple memory responses into a coherent answer.

Parameters:

• responses (list[dict[str, Any]])

• query (str)

Return type:

dict[str, Any]

class agents.memory_v2.multi_memory_agent.MultiMemoryConfig(/, **data)

Bases: pydantic.BaseModel

Configuration for MultiMemoryAgent.

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)

model_config

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class agents.memory_v2.multi_memory_agent.MultiMemoryState

Bases: agents.memory_v2.memory_state_with_tokens.MemoryStateWithTokens

Extended state for MultiMemoryAgent with routing information.

class agents.memory_v2.multi_memory_agent.QueryClassifier(llm_config)

Classifies queries to determine appropriate memory strategy.

Parameters:

llm_config (haive.core.engine.aug_llm.AugLLMConfig)

async classify(query)

Classify a query to determine its type and characteristics.

Parameters:

query (str)

Return type:

dict[str, Any]

class agents.memory_v2.multi_memory_agent.QueryType

Bases: str, enum.Enum

Types of queries that determine memory routing.

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

class agents.memory_v2.multi_memory_agent.ResponseSynthesizer(llm_config)

Synthesizes responses from multiple memory agents.

Parameters:

llm_config (haive.core.engine.aug_llm.AugLLMConfig)

async synthesize(responses, original_query)

Synthesize multiple memory responses into a coherent answer.

Parameters:

• responses (list[dict[str, Any]])

• original_query (str)

Return type:

dict[str, Any]

agents.memory_v2.multi_memory_agent.create_multi_memory_agent(name='multi_memory_coordinator', enable_graph=HAS_GRAPH_MEMORY, enable_rag=HAS_RAG_MEMORY, **kwargs)

Factory function to create a MultiMemoryAgent with sensible defaults.

Parameters:

• name (str)

• enable_graph (bool)

• enable_rag (bool)

Return type:

MultiMemoryAgent