Skip to main content

​
rag

Rust AI Agent SDK RAG (Retrieval Augmented Generation) Module This module provides RAG capabilities for agents:
  • RAG - Main RAG pipeline
  • RAGConfig - Configuration for RAG
  • RAGResult - Result with answer and citations
  • Citation - Source citation
  • SmartRetriever - Intelligent document retrieval

​
Example

use praisonai::rag::{RAG, RAGConfig};

let rag = RAG::new()
.config(RAGConfig::default())
.build()?;

let result = rag.query("What is the main finding?")?;
println!("{}", result.answer);

​
Import

use praisonai::rag::*;

​
Classes

Citation

A citation referencing a source document.

ContextPack

A pack of context chunks for RAG.

ContextChunk

A single context chunk.

RAGResult

Result of a RAG query.

RAGConfig

Configuration for RAG pipeline.

RetrievalConfig

Unified retrieval configuration (Agent-first).

TokenBudget

Token budget for context management.

RetrievalResult

Result of a retrieval operation.

RAG

Main RAG pipeline.

RAGBuilder

Builder for RAG

RetrievalStrategy

Retrieval strategy for RAG.

CitationsMode

Citations mode for RAG.

​
Functions

get_model_context_window()

Get model context window size.

estimate_tokens()

Estimate token count for text.

build_context()

Build context string from chunks.

truncate_context()

Truncate context to fit token limit.

deduplicate_chunks()

Deduplicate chunks by content similarity.

Rust RAG

Rust Retrieval

Rust Vector Store

Rust Chunking