openground query

Search documentation using hybrid semantic and keyword search

The openground query command performs hybrid search (semantic + BM25) against the local vector database.

Usage

openground query <query> [OPTIONS]

Arguments

query · string · required

Query string for hybrid search

Options

--version · string · default: "latest"

Version to filter results by.

Aliases: -v

--library · string · default: "None"

Optional library name filter. If provided, only searches within this library.

Aliases: -l

--top-k · integer · default: "5"

Number of results to return. Default value comes from configuration (query.top_k).

Aliases: -k

Constraints: Minimum value is 1

Search Algorithm

OpenGround uses hybrid search combining:

1. Semantic Search: Vector similarity using embeddings
2. BM25 (Keyword Search): Traditional full-text search

Results are ranked by combining both scores, providing relevant results even when exact keywords don’t match.

Examples

Basic Query

openground query "How to create a vector database"

Query Specific Library

openground query "authentication" --library langchain

Query Specific Version

openground query "new features" --version v2.0.0

Get More Results

openground query "embeddings" --top-k 10

Combined Options

openground query "API reference" -l mylib -v v1.5.0 -k 3

Output Format

Results are returned in markdown format with:

• Title: Page title
• URL: Source URL
• Library & Version: Which library and version
• Content: Relevant excerpt from the page

Output Example

# Result 1

**Title:** Vector Database Guide
**URL:** https://docs.example.com/vector-db
**Library:** mylib (v1.0.0)

Vector databases store high-dimensional embeddings and support 
semantic search queries. To create a vector database, you first 
need to generate embeddings for your documents...

---

# Result 2

**Title:** Embeddings Overview
**URL:** https://docs.example.com/embeddings
**Library:** mylib (v1.0.0)

Embeddings are numerical representations of text that capture 
semantic meaning. OpenGround supports multiple embedding backends...

Configuration

The default top_k value can be configured:

openground config set query.top_k 10

Performance Tips

1. Be Specific: More specific queries return better results
2. Use Library Filter: Filter by library when you know which docs to search
3. Adjust top_k: Increase for broader searches, decrease for focused results

Related Commands

• openground list - View available libraries
• openground add - Add documentation
• openground config - Configure query settings