rushdb-query-builder

To build an ontology-driven query, you first discover available labels and properties in your dataset, then use that schema to construct a findRecords query. This approach filters records by traversing relationships across related labels accurately.

Ontology-first discovery is the process of identifying labels and properties before executing a query. It works by loading the query specification via getSearchQuerySpec, ensuring your findRecords query targets valid schema elements and avoids field mapping errors.

Yes, you can perform semantic vector search and groupBy aggregation within the same query. The query builder supports discovering properties, filtering, and aggregating results by computed metrics alongside optional semantic search operations.

No external dependencies are required to use the ontology query builder. You simply install and connect to the MCP server, then initiate a single session to discover your dataset's ontology and construct your findRecords query.

The best way to filter records across related labels is to use ontology-driven query building. By discovering the schema first, you can safely specify relationship traversals and property filters before executing the findRecords command.