Memory

Persistent conversational memory with semantic search using PostgreSQL and pgvector, combining traditional memory storage with confidence scoring and a bidirectional knowledge graph for entity relationships

Provides persistent semantic memory and knowledge graph capabilities through PostgreSQL with pgvector for vector similarity search using Aliyun's text-embedding-v3 model. Features traditional memory storage with confidence scoring and decay algorithms, plus a bidirectional knowledge graph for entity relationships. Includes LRU caching for embeddings, hybrid search combining vector and keyword matching, memory access tracking, and comprehensive audit logging for persistent conversational context and complex knowledge representation.