Friday Memory

Conversation memory, semantic retrieval, entity graphs, context assembly, ingestion, and lifecycle maintenance. Choose individual adapters or let Friday Runtime compose the memory controller.

Distribution: friday-framework-memory · Import: friday_memory Python: 3.13.4+ · Status: alpha · License: MIT

Installation

python -m pip install "friday-framework-memory==0.1.0a4"

Friday distributions use the same release version; keep internal Friday dependencies aligned.

Quick start: store and retrieve a conversation turn

This example is ephemeral and requires no model, credentials, or database.

import asyncio

from friday_memory import InMemoryEpisodicStore, InteractionTurn

async def main() -> None:
    store = InMemoryEpisodicStore()
    await store.add_turn(
        "demo",
        InteractionTurn(
            id="turn-1", session_id="demo",
            user_message="Remember the project name.",
            assistant_message="The project is Friday.",
        ),
    )
    turns = await store.get_turns("demo")
    print(turns[0].assistant_message)

asyncio.run(main())

Expected output: The project is Friday. This does not perform semantic retrieval or durable persistence.

Choosing a memory surface

MemoryController requires explicit vector, graph, episodic and collection-management dependencies. For a full application, prefer Runtime composition over assembling those dependencies by hand. Extraction, semantic recall and maintenance can invoke configured model services; the offline example above does not.

Storage and configuration

Persistent adapters need application-owned paths and, where applicable, a compatible embedding model/dimension. Use disposable directories for experiments. Stored transcripts, memory and diagnostic content can be sensitive; do not commit them. Lifecycle operations and snapshot/restore APIs have explicit ownership and quiescence requirements; they are not interchangeable with deleting store files.

Development

Work from the monorepo root after the shared contributor setup. Run the focused suite:

uv run --no-sync pytest packages/friday-memory/tests_memory -q

Documentation and support

Runtime setup · Memory guide · Memory management · Source · Issues · Contributing

License

MIT.