weaviate-query-agent

Semantic search in Weaviate uses vector embeddings to find results based on meaning rather than exact keywords. This Skill queries your local Weaviate collections with near_text, near_vector, or near_image operations, returning contextually relevant data without requiring exact term matches.

Yes, hybrid search combines BM25 keyword matching with vector similarity in a single query. This Skill supports hybrid queries that weight both semantic and keyword relevance, delivering more precise results than either approach alone.

Property-based filters narrow results by specific conditions like author, date, or category before or alongside semantic ranking. This Skill applies where conditions to refine your search space, ensuring retrieved data meets both semantic relevance and business logic requirements.

RAG retrieves relevant documents from Weaviate, then passes them to a language model to generate context-aware answers with cited sources. This Skill orchestrates the retrieval and generation pipeline, supporting vision-enabled RAG to analyze images and diagrams within your documents.

You need a running Docker-based Weaviate instance, an active connection to it, ingested data in your collections, and Python weaviate-client installed. This Skill requires these prerequisites to execute near_text, near_vector, BM25, and hybrid queries against your local deployment.

Yes, near_image queries search collections containing visual content by image or vision understanding. This Skill supports image-based retrieval and vision-enabled RAG, enabling analysis of diagrams, maps, and other visual documentation alongside text.