skills.rest
shinnytech
shinnytechOFFICIAL
Explore all skills by this author →
WebsiteGitHub

hybrid-search-implementation

Combine vector similarity and keyword retrieval with fusion techniques like RRF.

Updated Jan 13, 2026
Search Tags:#information retrieval#rag#vector search#hybrid search#search engine#reranking#keyword search
One-click install
npx skills add https://github.com/shinnytech/caiwenqiang-member-rank --skill hybrid-search-implementation-shinnytech
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: hybrid-search-implementation
Source: https://github.com/shinnytech/caiwenqiang-member-rank/tree/main/.cursor/skills/hybrid-search-implementation
Command: npx skills add https://github.com/shinnytech/caiwenqiang-member-rank --skill hybrid-search-implementation-shinnytech

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill addresses the limitations of relying on a single search method (either vector similarity or keyword matching) by combining their strengths to achieve more comprehensive and accurate retrieval of information.

Core Features & Use Cases

  • Improved Recall: Combines semantic understanding with exact keyword matching to ensure fewer relevant results are missed.
  • Flexible Implementation: Provides templates and examples for integrating hybrid search into RAG systems, search engines, and applications requiring nuanced search capabilities.
  • Use Case: When searching a large knowledge base, a query like "What are the safety protocols for operating the X-1000 machine?" might miss documents that only mention "X-1000" but not "safety protocols" (vector search limitation), or documents that discuss safety but not the specific "X-1000" model (keyword search limitation). Hybrid search ensures both aspects are considered.

Quick Start

Implement hybrid search using Reciprocal Rank Fusion by providing lists of results from vector and keyword searches to the reciprocal_rank_fusion function.

Frequently Asked Questions about hybrid-search-implementation

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

FAQPage Schema
How does hybrid search improve RAG system retrieval accuracy?▼

Hybrid search combines vector similarity for semantic understanding with keyword matching for exact terms, overcoming single-method limitations to improve recall and precision in RAG systems and search engines.

How do I implement reciprocal rank fusion for combined search results?▼

Implement reciprocal rank fusion by providing lists of results from vector and keyword searches to the fusion function, which merges the rankings to produce a unified list of relevant documents.

Can I use hybrid search with PostgreSQL and Elasticsearch?▼

Yes, you can use hybrid search with PostgreSQL and Elasticsearch, as the Skill provides specific implementation templates and examples for integrating combined vector and keyword retrieval methods into these platforms.

What is the best way to combine vector search and keyword search?▼

The best way to combine vector and keyword search is using fusion techniques like Reciprocal Rank Fusion (RRF), linear combination, or cross-encoder reranking to merge results based on your specific precision and recall needs.

When should I not use a single vector search method?▼

You should not use a single vector search method when queries contain specific identifiers or exact keywords, because semantic matching might miss documents that lack contextual synonyms but contain the exact search terms.

Related Skills

Explore More Software Engineering Skills→
ArogyaReddyArogyaReddy

hybrid-search-implementation

Combine vector similarity and keyword matching with RRF, Linear, Cross-encoder, or Cascade fusion.

Community
Advanced
artgaurav16420-ossartgaurav16420-oss

advanced-rag-optimizer

Optimizes Retrieval-Augmented Generation pipelines for source-backed answers with chunking, hybrid search, reranking, and evaluation workflows.

Community
Advanced
as4584as4584
2

hybrid-search-implementation

Combine vector and keyword search results using RRF, linear combination, or cross-encoder reranking.

Community
Advanced
Division6066Division6066

hybrid-search-implementation

Combine vector similarity with keyword matching for hybrid search retrieval.

Community
Advanced