feat: Add support for Weaviate vector store

## Vector Store Integration Request
Weaviate is an open source document based database with first class support for vector search. Weaviate can be run locally using Docker or in the cloud using their fully managed solution.
### Resources
Weaviate instances expose multiple APIs that allow users to interact with it: A RESTful API and a gRPC API for inserting and managing your collections, and a GraphQL or gRPC search API for performing vector search.

Inserting data can be performed via the [create](https://weaviate.io/developers/weaviate/manage-data/create) or [batch import](https://weaviate.io/developers/weaviate/manage-data/import) endpoints. Note that although you can provide the embedding vector associated with the object you are creating, Weaviate also supports automatic embedding generation using third party services (these must be configured in the Weaviate config). However, the scope of this first integration should be only concerned with the use case of generating embeddings outside of Weaviate.

Vector search is performed by invoking one of the [similarity search endpoints](https://weaviate.io/developers/weaviate/search/similarity).

Look at existing vector store integrations in the codebase to understand the interface. Review Weaviate's REST and gRPC APIs for data insertion and similarity search. Start by implementing a client for Weaviate in Rust, focusing on embedding vectors generated externally. Test with a local Docker instance of Weaviate to ensure basic CRUD and search operations work.

Vom Indexierungsmodell aus dem Issue-Text verfasst.