Memory is what makes an agent an agent
Most assistants forget the moment the transcript scrolls away. NMAFC is an open-source memory layer that keeps a dated, decay-aware, source-attributed record of everything it learns — and hands back a small, precise context hundreds of sessions later.
NMAFC — memory that outlives the context window
Most assistants forget. Not because the model is weak, but because the transcript ran out. NMAFC is an open-source memory layer that compacts every turn into dated atomic facts, lets unused ones fade on a decay curve, and gives back a small, precise, source-attributed context — hundreds of sessions later.
It is a Python package with a server attached. Import it, or point an agent at it. No OpenLoft account, no vector database to provision, no data leaving your network if you use local embeddings. Read the docs for install steps, configuration and the honest limitations.
Four ways to connect
Accuracy on LoCoMo question categories
Scored by an independent LLM judge on the LoCoMo benchmark, against a vanilla RAG baseline of 46.1%, 45.6% and 44.8% respectively. The full table across all five categories — including the ones we do not lead — is in the docs, and every figure regenerates from the committed data with python summarise.py.