Represents a single memory item in the textual memory. This serves as a standardized format for memory items across different textual memory

Extended metadata for structured memory, layered retrieval, and lifecycle tracking.

ThreadPoolExecutor that automatically propagates the main thread's trace_id to worker threads.

Schema for creating a raw memory. Args: user_id: User ID this memory belongs to organization_id: Organization ID occ

Purpose: **memory provenance / traceability**. Capture the minimal, reproducible origin context of a memory item so it can be audited, t

Represents a node in a graph. Attributes: id (str): A unique identifier for the node. attributes (Dict[str, Any]): A dictiona

Manager class to handle business logic related to Users.

Base class for test cases.

Manager class to handle business logic related to Clients.

Configuration for a Language Model (LLM) model. This object specifies all the information necessary to access an LLM model to usage with Letta, e

Embedding model configuration. This object specifies all the information necessary to access an embedding model to usage with Mirix, except for

Metadata for a memory item. This includes information such as the type of memory, when it occurred, its source, and other relevant details.

Result of memorize() method.

Result of query() method.

Configuration for a Language Model (LLM) model. This object specifies all the information necessary to access an LLM model to usage with Letta, e

Request to create a message

Represents a single item in the vector database. This serves as a standardized format for vector database items across different vector datab

A record or records cannot be found given the provided search params

A record or records cannot be found given the provided search params

Embedding model configuration. This object specifies all the information necessary to access an embedding model to usage with Letta, except for

Thread class that automatically propagates the main thread's trace_id to child threads.

Manager class to handle business logic related to Organizations.

Health condition/disease with detailed health profile.

The MOSCore (Memory Operating System Core) class manages multiple MemCube objects and their operations. It provides methods for creating, sea

Query model.

Represents an edge in a graph, connecting two nodes. Attributes: node1 (Node): The starting node of the edge. node2 (Node): T

Letta's internal representation of a message. Includes methods to convert to/from LLM provider formats. Attributes: id (str): The un

Request to create a message

Data class for tracking running tasks in SchedulerDispatcher.

Defines data model for storing Message objects

Metric evaluation result.

Partitioned async in-memory queue that simulates Kafka's partitioning. Messages are routed to partitions based on user_id hash, ensuring:

Representation of an agent's internal reasoning. Args: id (str): The ID of the message date (datetime): The date the message

Client ORM class - represents a client application

Data generator providing three independent generation modes.

Pipeline master config.

Exception raised for errors in the JSON parsing process.

Service for persona operations.

Blocks are sections of the LLM context, representing a specific part of the total Memory

Blocks are sections of the LLM context, representing a specific part of the total Memory

Async in-memory queue implementation (single partition)

Background worker that processes messages from the queue as an asyncio.Task

Request-scoped context object that holds trace_id and other request data. This provides a Flask g-like object for FastAPI applications.

Represents an available tool that the LLM can invoke. NOTE: polymorphic inheritance makes more sense here as a TODO. We want a superset of tools

Representation of an agent's state. This is the state of the agent at a given time, and is persisted in the DB backend. The state has all the inf

https://platform.openai.com/docs/api-reference/chat/object

Database returns nothing

Error raised when a tool cannot be created.

Mirix's internal representation of a message. Includes methods to convert to/from LLM provider formats. Attributes: id (str): The un

Manager class to handle business logic related to Messages.

Represents a date/time range for filtering.

A message representing a request to call a tool (generated by the LLM to trigger tool execution). Args: id (str): The ID of the mess

Manager class to handle business logic related to Agents.

Manager class to handle business logic related to Blocks.

https://platform.openai.com/docs/api-reference/chat/create

Database manager for MemoryMonitorManager objects This class handles persistence, synchronization, and locking for MemoryMonitorManager insta

Service for dialogue simulation.

Service for event graph operations.

LLM response result.

Response model for memory operations.

Goal structure for both Fast & LLM.

Redis cache provider implementation. Wraps RedisMemoryClient to provide the cache provider interface used by service managers and ORM. A

Metadata for nodes returned by search, includes similarity info.

A thread-safe dictionary wrapper with optimized read-write locks. This class allows multiple concurrent readers while ensuring exclusive acc

Represents an available tool that the LLM can invoke. NOTE: polymorphic inheritance makes more sense here as a TODO. We want a superset of tools

Representation of an agent's state. This is the state of the agent at a given time, and is persisted in the DB backend. The state has all the in

A message sent by the LLM in response to user input. Used in the LLM context. Args: id (str): The ID of the message date (da

Update a block

https://platform.openai.com/docs/api-reference/chat/object

Dialogue generation config.