Module vector_provider 

Vector RAG Context Provider

Implements ContextProvider using vector embeddings for semantic code search. Indexes workspace files into a vector store and retrieves relevant context via cosine similarity on each agent turn.

Usage

use a3s_code_core::context::{
    VectorContextProvider, VectorContextConfig,
    embedding::OpenAiEmbeddingProvider,
    vector_store::InMemoryVectorStore,
};

let embedder = OpenAiEmbeddingProvider::new("sk-...", "text-embedding-3-small", 1536)?;
let store = InMemoryVectorStore::new();
let config = VectorContextConfig::new("/my-project");

let provider = VectorContextProvider::new(config, embedder, store);
provider.index().await?; // Index workspace files

// Then register as a context provider on the session

Structs

VectorContextConfig

Configuration for the vector context provider

VectorContextProvider

Vector RAG context provider