Fundamentals

Memory Model

Neutron AI organizes memory by tenant and scope. This keeps integrations predictable across products, agents, workspaces, and users.

Core Terms

TermMeaning
tenantIdYour top-level customer, workspace, organization, or application partition.
scopeIdA narrower memory boundary such as user:<id>, workspace:<id>, project:<id>, or agent:<id>.
Memory recordA durable fact, preference, instruction, summary, or outcome that may improve future agent work.
Context packA bounded set of relevant memory returned for one task.
Privacy classA label that helps your integration decide what can be recalled, cached, shown, or omitted.
TombstoneA deletion marker that prevents removed memory from reappearing.

Choose scopes that match how your product grants access.

Product needExample scope
Personal assistant memoryuser:<user-id>
Team shared knowledgeworkspace:<workspace-id>
Project-specific agent contextproject:<project-id>
Support account historyaccount:<account-id>
Short-lived workflow memorysession:<session-id>

You can request multiple scopes for an agent task. Keep the list explicit and only include scopes the current user or agent is allowed to access.

Memory Types

Use memory type labels consistently so recall quality improves over time.

TypeUse for
preferenceStable user or workspace preferences.
factLasting product, account, or workflow facts.
instructionStanding rules that should guide future agent behavior.
summaryCondensed history from a conversation, document, or workflow.
outcomeWhat happened after a task, decision, or tool run.

Lifecycle

flowchart LR
  A["remember"] --> B["recall"]
  B --> C["agent context"]
  C --> D["reflect"]
  D --> A
  B --> E["forget"]

Use forget for deletion, correction, user privacy controls, and stale or unsafe memory. Deletions should be treated as final in your integration logic.