Hybrid Search Implementation

From the author

Combine vector and keyword search for improved retrieval. Use when implementing RAG systems, building search engines, or when neither approach alone provides sufficient recall.

— the description by Seth Hobson

AutoGineer summary

Written by AutoGineer from the skill’s own files, not by the author

This skill provides a structured approach to implementing hybrid search, which merges vector similarity with keyword-based retrieval. By combining these two methods, developers can achieve better recall, especially when handling domain-specific vocabulary, exact matches, or queries that pure semantic search might miss. The skill covers architectural patterns, fusion methods like Reciprocal Rank Fusion (RRF) and linear combination, and best practices for tuning and testing search performance.

What it does

Defines the architecture for combining vector and keyword search results.
Provides implementation templates for Reciprocal Rank Fusion (RRF).
Offers logic for linear combination of search scores.
Outlines database schema patterns for hybrid search using PostgreSQL.
Suggests best practices for tuning weights and evaluating search quality.
Provides guidance on when to use specific fusion methods like reranking or cascading.

When to use it

When building RAG systems that require higher retrieval accuracy.
When your current vector search fails to capture exact keyword matches.
When dealing with domain-specific terminology or product codes.
When you need to balance semantic understanding with precise term matching.
When you want to improve search recall without relying on a single retrieval method.

Advantages

Improves retrieval recall by leveraging both semantic and exact matching.
Provides flexible fusion methods to suit different search requirements.
Includes actionable templates for common database implementations.
Encourages empirical tuning and A/B testing for optimal performance.
Helps address edge cases where single-method search is insufficient.

Usage examples

“How do I combine vector and keyword search results?”: The skill provides a conceptual architecture and specific Python templates for Reciprocal Rank Fusion (RRF) and linear combination to merge ranked lists from different search methods.
“What is the best way to implement hybrid search in PostgreSQL?”: The skill includes a template for setting up a PostgreSQL schema with pgvector for vector search and GIN indexes for full-text search, enabling efficient hybrid queries.

Install

Download the skill as a ZIP:
https://autogineer.com/agent-skills/wshobson/hybrid-search-implementation/download

Unzip it into the cross-client skills folder of your project:

mkdir -p .agents/skills && unzip hybrid-search-implementation.zip -d .agents/skills

Documentation by Seth Hobson

Patterns for combining vector similarity and keyword-based search.

When to Use This Skill

Building RAG systems with improved recall
Combining semantic understanding with exact matching
Handling queries with specific terms (names, codes)
Improving search for domain-specific vocabulary
When pure vector search misses keyword matches

Core Concepts

1. Hybrid Search Architecture

Query → ┬─► Vector Search ──► Candidates ─┐
        │                                  │
        └─► Keyword Search ─► Candidates ─┴─► Fusion ─► Results

2. Fusion Methods

Method	Description	Best For
RRF	Reciprocal Rank Fusion	General purpose
Linear	Weighted sum of scores	Tunable balance
Cross-encoder	Rerank with neural model	Highest quality
Cascade	Filter then rerank	Efficiency

Templates and detailed worked examples

Full template library and detailed worked examples live in references/details.md. Read that file when you need the concrete templates.

Best Practices

Do's

Tune weights empirically - Test on your data
Use RRF for simplicity - Works well without tuning
Add reranking - Significant quality improvement
Log both scores - Helps with debugging
A/B test - Measure real user impact

Don'ts

Don't assume one size fits all - Different queries need different weights
Don't skip keyword search - Handles exact matches better
Don't over-fetch - Balance recall vs latency
Don't ignore edge cases - Empty results, single word queries

Files

ATTRIBUTION.mdAdded by AutoGineer1022 B
LICENSELicence file added by AutoGineer1.0 KiB
SKILL.md2.0 KiB
references/details.md15.8 KiB

SKILL.md