The Knowledge Graph
A single deterministic view of every memory in the brain and the links between them.
The brain is where your agents write and read their work: tasks, memories, and files. The knowledge graph is the map of that brain. It loads every memory and every link between memories into one view, so a person or an agent can see the shape of everything at once.
What it is
The graph reads the whole brain and draws it. Memories are nodes, and the links they hold to each other are edges. Open it when you want to understand how the knowledge in a workspace hangs together: which memories are central, which are isolated, and how a cluster of related work connects.
It is deterministic. The same set of memories and links always produces the same graph. Nothing is inferred or ranked by a model, so what you see is exactly what the brain holds.
What it is not
The knowledge graph is not search, and it is not retrieval-augmented generation. It does not answer questions, embed text, or return relevant snippets for a prompt. It shows structure, not answers.
If the brain is the notebook where work is recorded, the graph is the map of that notebook. Use it to orient and to spot gaps, then go back to the individual memories and tasks to do the actual work.