net10.0
OpenSearch vector adapter for Koan: index provisioning, kNN search, metadata filtering and health checks via the OpenSearch REST API.
Published
Jun 12, 2026
4mo ago
Total Downloads
3.9K
3,913
Install
dotnet add package Sylin.Koan.Data.Connector.OpenSearch --version 0.17.0
| Product | Versions |
|---|---|
| .NET |
README
Sylin.Koan.Data.Connector.OpenSearch
OpenSearch adapter for Koan vector data.
- Target framework: net10.0
- License: Apache-2.0
Capabilities
- Automatic index provisioning with dense vector mappings
- kNN vector search with bulk operations
- Operator-aware metadata filtering pushed into the kNN query — the
Filter→ query-DSL translation is shared with Elasticsearch viaKoan.Data.SearchEngine - Health checks and orchestration-aware configuration
Install
dotnet add package Sylin.Koan.Data.Connector.OpenSearch
Example
using Koan.Data.Abstractions.Filtering;
await Vector<MyDocument>.Save("doc-1", embedding, metadata: new { category = "docs" });
// Metadata filter pushed into the kNN query (the unified Filter AST).
var results = await Vector<MyDocument>.Search(
embedding,
topK: 5,
filter: Filter.Eq("category", "docs"));
The filter argument also accepts the JSON metadata-filter DSL ({ "category": "docs" }) or a
dictionary; all three are parsed by the one schemaless reader. An operator the adapter cannot push
down is a hard error, never a silent match-all.
Links
- Vector pathway + filter model:
~/decisions/DATA-0097-vector-pathway-parity.md - Embedding↔vector seam:
~/decisions/AI-0036-embedding-vector-seam.md
Changes from Previous Version
v0.8.0v0.17.0
Added Dependencies
Sylin.Koan.Data.SearchEngine(net10.0)
Adoption Assessment
Assess this package against its supported frameworks, license, maintenance signals and dependency graph before making it a project standard.
Compatibility metadata lists 1 target framework; confirm the exact target used by your application.
Related Guides & Workflows
Learn how to evaluate, integrate, or audit this package in your .NET project.