Crate adk_rag 

adk-rag

Retrieval-Augmented Generation for ADK-Rust agents.

This crate provides a modular, trait-based RAG system with pluggable embedding providers, vector stores, chunking strategies, and rerankers. A RagPipeline orchestrates the full ingest-and-query workflow, and a RagTool exposes retrieval as an adk_core::Tool for agentic use.

Features

All external backends are feature-gated. The default feature set includes only core traits, the in-memory vector store, and chunking implementations.

Feature	What it enables
gemini	GeminiEmbeddingProvider via adk-gemini
openai	OpenAIEmbeddingProvider via reqwest
qdrant	QdrantVectorStore via qdrant-client
lancedb	LanceDBVectorStore via lancedb
pgvector	PgVectorStore via sqlx
surrealdb	SurrealVectorStore via surrealdb
full	All of the above

Re-exports

pub use chunking::Chunker;

pub use chunking::FixedSizeChunker;

pub use chunking::MarkdownChunker;

pub use chunking::RecursiveChunker;

pub use config::RagConfig;

pub use config::RagConfigBuilder;

pub use document::Chunk;

pub use document::Document;

pub use document::SearchResult;

pub use embedding::EmbeddingProvider;

pub use error::RagError;

pub use error::Result;

pub use inmemory::InMemoryVectorStore;

pub use pipeline::RagPipeline;

pub use pipeline::RagPipelineBuilder;

pub use reranker::NoOpReranker;

pub use reranker::Reranker;

pub use tool::RagTool;

pub use vectorstore::VectorStore;

Modules

chunking

Document chunking strategies.

config

Configuration for the RAG pipeline.

document

Data types for documents, chunks, and search results.

embedding

Embedding provider trait for generating vector embeddings from text.

error

Error types for the adk-rag crate.

inmemory

In-memory vector store using cosine similarity.

pipeline

RAG pipeline orchestrator.

reranker

Reranker trait for re-scoring search results.

tool

Agentic retrieval tool for ADK agents.

vectorstore

Vector store trait for storing and searching vector embeddings.