upstash-vector-js

# Vector Documentation Skill ## Quick Start Vector is a high‑performance vector database for storing, querying, and managing vector embeddings. Basic workflow: - Install the Vector TS SDK. - Connect to a Vector instance. - Upsert vectors, query them, and manage namespaces. Example (TypeScript): ```ts import { Index } from "@upstash/vector"; const index = new Index({ url: process.env.UPSTASH_VECTOR_REST_URL!, token: process.env.UPSTASH_VECTOR_REST_TOKEN!, }); await index.upsert([{ id: "1", vector: [0.1, 0.2], metadata: { tag: "example" } }]); const results = await index.query({ vector: [0.1, 0.2], topK: 5, }); ``` For full usage, refer to the linked skill files below. ## Other Skill Files ### TS SDK Reference - `sdk-methods`: Explains SDK commands: delete, fetch, info, query, range, reset, resumable-query, upsert ### Features - `features/namespaces`: Explains namespaces and dataset organization. - `features/index-structure`: Covers hybrid and sparse index structures. - `features/filtering-and-metadata`: Details metadata storage and server-side filtering. Use these files for deeper guidance on SDK usage, advanced configurations, algorithms, and integrations.