ElBruno.Connectors.SqliteVec 0.5.3

ElBruno.Connectors.SqliteVec

A MEAI-native VectorStoreCollection implementation for sqlite-vec, based on the original Semantic Kernel implementation. Store, query, and search vector embeddings using SQLite with full support for the Microsoft.Extensions.VectorData abstractions.

Features

• ✅ VectorStoreCollection implementation — full CRUD + vector similarity search
• ✅ Dependency injection — register collections with AddSqliteVecCollection<TKey, TRecord>
• ✅ Attribute-based mapping — use [VectorStoreKey], [VectorStoreData], [VectorStoreVector]
• ✅ Filtered retrieval — query records with LINQ-style expressions
• ✅ Text splitting — built-in TextSplitter.SplitParagraphs for chunking documents
• ✅ Optional embedding generation — works with IEmbeddingGenerator<string, Embedding<float>> or raw float vectors
• ✅ Multi-target — supports .NET 9 and .NET 10

Installation

dotnet add package ElBruno.Connectors.SqliteVec

Quick Start

Define your record type

using Microsoft.Extensions.VectorData;

public class GlossaryEntry
{
    [VectorStoreKey]
    public string Key { get; set; } = string.Empty;

    [VectorStoreData]
    public string Term { get; set; } = string.Empty;

    [VectorStoreData]
    public string Definition { get; set; } = string.Empty;

    [VectorStoreVector(Dimensions: 4)]
    public ReadOnlyMemory<float> Embedding { get; set; }
}

Register and use the collection

using ElBruno.Connectors.SqliteVec;
using Microsoft.Extensions.DependencyInjection;
using Microsoft.Extensions.VectorData;

var services = new ServiceCollection();
services.AddSqliteVecCollection<string, GlossaryEntry>(
    collectionName: "glossary",
    connectionString: "Data Source=mydata.db");

var provider = services.BuildServiceProvider();
var collection = provider.GetRequiredService<VectorStoreCollection<string, GlossaryEntry>>();

// Create the collection
await collection.EnsureCollectionExistsAsync();

// Upsert a record
await collection.UpsertAsync(new GlossaryEntry
{
    Key = "1",
    Term = "Embedding",
    Definition = "A numerical representation of data in a vector space.",
    Embedding = new ReadOnlyMemory<float>([0.1f, 0.2f, 0.3f, 0.4f])
});

// Get by key
var entry = await collection.GetAsync("1");

// Vector similarity search
var query = new ReadOnlyMemory<float>([0.1f, 0.2f, 0.3f, 0.4f]);
await foreach (var result in collection.SearchAsync(query, top: 5))
{
    Console.WriteLine($"{result.Record.Term}: {result.Score:F4}");
}

Text splitting

using ElBruno.Connectors.SqliteVec;

var chunks = TextSplitter.SplitParagraphs(
    new[] { "Your long document text goes here..." },
    maxWordsPerChunk: 100);

API Reference

Key Methods

Samples

• BasicSample — CRUD operations and vector search with raw float vectors
• IntermediateSample — TextSplitter, multiple collections, filtered retrieval, and embedding generation

License

This project is licensed under the MIT License — see the LICENSE file for details.

👨‍💻 About the Author

Bruno Capuano (aka El Bruno) is a passionate developer, AI enthusiast, and content creator who loves building practical solutions and sharing knowledge with the community.

🌐 Connect & Explore

• 🐙 GitHub: github.com/elbruno - More cool projects and open-source contributions
• 📝 Blog: elbruno.com - Technical articles, tutorials, and insights
• 🎙️ Podcast: No Tiene Nombre - Spanish tech podcast
• 🎥 YouTube: youtube.com/elbruno - Video tutorials and demos
• 💼 LinkedIn: linkedin.com/in/elbruno - Professional updates and articles
• 🐦 Twitter/X: @elbruno - Quick tips and tech discussions

💡 Support This Project

If you find this library useful:

• ⭐ Star this repo on GitHub
• 📢 Share it with your network
• 🐛 Report issues or suggest features
• 🤝 Contribute via pull requests

Happy coding! 🚀