Namespaces

A namespace is one memory space: an isolated set of memories, entities, documents, and configuration. Give each agent, team, or domain its own namespace and memories never bleed across the boundary.

Lifecycle

Create a namespace explicitly with PUT /namespaces/{namespace_id}; commit does not auto-create them. Ids are yours to choose (support-agent, team-payments). Delete removes the namespace and everything in it.

Free workspaces hold 3 namespaces; pro raises the cap.

Identity and configuration

Namespace config shapes behavior on both sides of the memory:

• Commit side. A commit mission, extraction mode, custom extraction instructions, and chunk size steer what gets kept.
• Illuminate side. An illuminate mission and disposition (skepticism, literalism, empathy) steer how illuminate answers.
• Signals. Enable and steer consolidation.

Config updates are partial (PATCH); DELETE /config resets to defaults.

Standing instructions

Directives are per-namespace standing instructions that bias commit and search continuously: "prefer customer names over ticket numbers", "treat runbook facts as authoritative".

Templates

Export a namespace's configuration as a template and import it to stamp out consistent namespaces per customer or per environment.

Observability

Per-namespace stats, a memories timeseries, and an audit trail recording every operation the namespace served over HTTP and MCP, with its outcome, duration, and the API key or user behind it.