II.
KnowledgeFabricImpl overview
Reference · liveknowledge-fabric-impl:zep-fabric
Zep as Agent Memory Fabric overview
Zep as an agent-facing memory knowledge fabric. Persists, summarizes, and retrieves conversation history with entity extraction and temporal awareness. Enriches memories with structured entity graphs — tracking people, organizations, and concepts mentioned across conversations. REST API and SDKs for Python and TypeScript. As a knowledge fabric, Zep provides conversation-centric knowledge with temporal reasoning — understanding not just what was discussed but when and in what order.
Attributes
displayName
Zep as Agent Memory Fabric
description
Zep as an agent-facing memory knowledge fabric. Persists, summarizes,
and retrieves conversation history with entity extraction and temporal
awareness. Enriches memories with structured entity graphs — tracking
people, organizations, and concepts mentioned across conversations.
REST API and SDKs for Python and TypeScript. As a knowledge fabric,
Zep provides conversation-centric knowledge with temporal reasoning —
understanding not just what was discussed but when and in what order.
knowledgeFileFormats
- conversation-history
- entity-extracts
- fact-triples
retrievalStrategy
hybrid
knowledgePersistence
memory-service
knowledgeScopes
- user
- agent
autoExtractionSupport
true
notes
Zep's temporal awareness sets it apart from simple vector stores — it
understands that knowledge has a time dimension. Fact retrieval is
time-weighted, so recent knowledge is prioritized. Entity extraction
builds a knowledge graph of entities mentioned in conversations,
enabling relationship-aware retrieval. This makes Zep particularly
suited for long-running agent relationships where context evolution
matters.
Outgoing edges
integrates_with1
- tool:zep·ToolZep
realizes1
- layer:12-knowledge-fabric·LayerKnowledge Fabric
uses_memory_system1
- memory-system:zep-service·MemorySystemZep Memory Service
Incoming edges
None.