Product #10 — Now Available
CAIN Memory
Governed persistent memory for autonomous agents. Write. Manage. Retrieve. Verify. Trust.
Step 1
Write
Agent stores memory
Step 2
Admit
Admission control evaluates
Step 3
Trust
Trust level assigned
Step 4
Use
Retrievable with ranking
Real-Time System State
-
Memories
-
Trusted
-
Unverified
-
Revoked
-
Expired
Loading system state...
Interactive Memory Demo
Store a memory and see it pass through CAIN governance.
Memory Governance Decision
IDENTITY:
-
TENANT:
-
POLICY:
-
SECURITY:
-
MEMORY:
-
PROVENANCE:
-
RESULT:
-
Trust Levels
UNVERIFIED
Newly written, not yet assessed. Default for agent-generated content.
OBSERVED
Seen but not validated. Human and system content starts here.
VALIDATED
Cross-referenced and found consistent. Requires explicit validation.
TRUSTED
Authoritative. Can be used for consequential decisions. Not auto-granted.
REVOKED
Explicitly invalidated. Cannot be restored without re-admission.
EXPIRED
Time-based decay. Configurable expiration per memory.
Memory Dashboard
-
Total Active
-
Trusted
-
Unverified
-
Revoked
-
Quarantined
Click "List Recent" to load memories
API Reference
POST /fabric/memory/store
{"agent_id": "agent-1", "memory_type": "semantic", "content": "...", "source": "agent"}
GET /fabric/memory/{memory_id}
curl -H "X-API-Key: key" https://.../fabric/memory/mem_abc123
POST /fabric/memory/query
{"query_text": "server config", "min_trust_level": "validated"}
Full OpenAPI spec available at /openapi.json