MinimemoryInstall

Summary

Create a new vector database in memory with specified dimensions, distance metric, and index type.

Inputs

• databaseName (required) — Name to identify this database instance in memory
• dimensions (required) — Number of dimensions for vectors. Common: 384 (MiniLM), 768 (MPNet), 1536 (OpenAI)
• distance — Distance metric for similarity calculation
• indexType — Index type - affects speed vs accuracy

Output shape

success: boolean, message: string, name: string, dimensions: number, distance: string, indexType: string, vectorCount: number

On success, database is created in memory. On failure, success is false with an error message.

Examples

Example 1: Basic create with default name

Leave databaseName as default, use default dimensions 384 and distance cosine, flat index type. Trigger via workflow execution; operation returns success status and metadata.