This section covers who a memory belongs to, who can see it, and how isolation is enforced — from scopes and multi-tenancy to consent, auditability, and the right to be forgotten.
What you will learn
- Why access control is mandatory for production memory systems
- Project, user, and property scopes — and hard vs soft isolation
- Preventing cross-user leakage at write time and query time
- Privacy-by-design, residency, consent, and deletion requirements
Chapters in this section
There are 18 chapters in this part. Open any chapter to read it on its own, or work through them in order.
- 1 Why Memory Needs Access Control
- 2 Scopes: Controlling Who a Memory Belongs To
- 3 Project-Wide Memory vs User-Scoped Memory
- 4 Property-Scoped Memory and Soft Isolation
- 5 Multi-Tenancy as a Memory Isolation Mechanism
- 6 Hard Isolation vs Soft Isolation
- 7 Preventing Cross-User Data Leakage by Design
- 8 Role-Based Access Control for Memory Systems
- 9 Enforcing Scope at Write Time and Query Time
- 10 Bounded Topics: Enforcing One Memory Per Scope
- 11 Designing Memory Systems for Regulated Industries
- 12 Data Residency and Memory Storage Location
- 13 Consent and Memory: What Should Be Rememberable
- 14 Auditability: Tracing Where a Memory Came From
- 15 Tenant-Level Backup and Restore
- 16 Memory Isolation in Multi-Agent Deployments
- 17 Privacy-by-Design in Memory Architecture
- 18 Right to Be Forgotten: Deleting Memory on Request