An indexing and search plugin for the mikser-io static site generator that uses OpenAI embeddings and either sqlite-vec or pgvector for storage. Version 1.2.0 is current, with active development. It integrates semantic search directly into the build lifecycle — indexing entities as they are processed, providing a findSimilar() runtime helper, and optionally exposing an HTTP search endpoint. This eliminates the need for external search services like Algolia, keeping the index in sync without additional infrastructure.
npm install mikser-io-vectorNo compatibility data collected yet for this library.
Verified import paths — ran on the pinned version, not inferred.
Basic configuration for mikser-io-vector with one store 'documents' indexing title and content using OpenAI embeddings.
Upgrade mikser-io to version 7 or later.
Use import syntax and ensure your project is configured for ESM.
Use client: 'better-sqlite3'.
Set dim to 1536 for text-embedding-3-small or 3072 for text-embedding-3-large.
Ensure your mikser configuration includes the 'api' plugin and a server setup.
Run npm install mikser-io-vector
Set OPENAI_API_KEY in environment or configure vector.openai.apiKey in config.
Change client to 'better-sqlite3'.
Add null safety checks or ensure entity structure matches map.