vecsearch

Vector semantic search identifies similar notes in a memory knowledge base by comparing vector embeddings. It returns a structured table of matching results ranked by distance metrics to help you quickly locate related information.

Yes, vector search supports optional deduplication to identify and surface similar notes within memory collections. This allows you to avoid content duplication and triangulate overlapping information across projects.

Semantic search filters results by applying distance-based ranking while restricting the query scope to specific note types or projects. This ensures context-aware retrieval across your memory knowledge base.

You need a pre-built vector index within your basic-memory knowledge base before running the search. The command parses your query arguments to retrieve the top matching notes from this index.

Vector search requires an existing vector index and is limited to retrieving notes within the basic-memory knowledge base. It returns distance metrics for ranking but does not generate new content or modify the stored notes.