skills.rest
zester4
zester4COMMUNITY
Explore all skills by this author →
GitHub

upstash-vector-js

Implements Upstash Vector operations using the TypeScript SDK for embeddings and similarity search.

2|1|Updated Jun 9, 2026
Search Tags:#typescript#semantic search#embeddings#rag#vector database#similarity search#upstash
One-click install
npx skills add https://github.com/zester4/zilmate --skill upstash-vector-js-zester4
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: upstash-vector-js
Source: https://github.com/zester4/zilmate/tree/main/.agents/skills/upstash-vector-js
Command: npx skills add https://github.com/zester4/zilmate --skill upstash-vector-js-zester4

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires @upstash/vector, and includes references (resource) components.

What problem does it solve? Developers building semantic search or RAG applications need accurate guidance on using the Upstash Vector TypeScript SDK, including upserting vectors, querying with filters, and managing namespaces without hitting common pitfalls. ## Core Features & Use Cases - Full SDK Method Coverage: Documents upsert, query, fetch, delete, range, resumable-query, info, and reset operations with TypeScript examples and pitfalls. - Index Structure Guidance: Explains dense, sparse, and hybrid indexes, fusion algorithms (RRF, DBSF), and custom reranking workflows. - Metadata Filtering & Namespaces: Covers SQL-like filter syntax, nested object access, and namespace isolation for multi-tenant datasets. - Use Case: When building a semantic search feature, use this Skill to correctly upsert embedded documents, query with metadata filters like "genre = 'fantasy'", and paginate large result sets with resumable queries. ## Quick Start Ask how to upsert and query vectors with the Upstash Vector TypeScript SDK, including metadata filtering and namespace usage.

Frequently Asked Questions about upstash-vector-js

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I upsert and query vectors with the Upstash Vector TypeScript SDK?▼

Create an Index instance with your REST URL and token, then call index.upsert with id, vector, and optional metadata. Query with index.query passing a vector or data string, topK, and optional metadata filters.

How to filter Upstash Vector query results by metadata?▼

Pass a SQL-like filter string to the query filter parameter, supporting operators like =, !=, IN, CONTAINS, GLOB, and HAS FIELD. You can access nested fields with dot notation and combine conditions with AND/OR.

Does Upstash Vector support hybrid search with sparse vectors?▼

Yes, hybrid indexes combine dense and sparse vectors, requiring both components on upsert. Queries fuse results using RRF or DBSF algorithms, and sparse vectors are limited to 1,000 non-zero entries.

Can I upsert raw text instead of vectors in Upstash Vector?▼

Yes, use the data field instead of vector, but only if the index has a configured embedding model. Check index.info() for denseIndex.embeddingModel or sparseIndex.embeddingModel before relying on auto-embedding.

Why does my Upstash Vector query return fewer results than topK?▼

Highly selective metadata filters consume a filtering budget, so post-filtering may reduce results below topK. Sparse queries also return fewer results when no overlapping indices exist between vectors.

How do namespaces work in Upstash Vector?▼

Namespaces partition an index into isolated subsets, created automatically on first upsert. Use index.namespace("name") to scope operations; omitting it writes to the default empty-string namespace.

Related Skills

Explore More Software Engineering Skills→
zepdbzepdb
43

zeppelin

Index and search text with Zeppelin vector store and BM25 full-text indexing.

Official
Advanced
Aktoh-CyberAktoh-Cyber

AgentDB Vector Search

Perform semantic vector search and retrieval with AgentDB for RAG workflows.

Official
Advanced
robertwaltosrobertwaltos

qmd

Index and search local files with BM25, vector embeddings, and reranking.

Community
Intermediate
ancolemanancoleman
503

using-vector-databases

Implement vector databases for AI/ML semantic search and RAG systems.

Community
Advanced