Skip to main content

Class: RetrievalAugmentor

Defined in: packages/agentos/src/cognition/rag/RetrievalAugmentor.ts:61

Implements​

Orchestrates the RAG pipeline including ingestion, retrieval, and document management.

Implements​

Constructors​

Constructor​

new RetrievalAugmentor(): RetrievalAugmentor

Defined in: packages/agentos/src/cognition/rag/RetrievalAugmentor.ts:89

Constructs a RetrievalAugmentor instance. It is not operational until initialize is successfully called.

Returns​

RetrievalAugmentor

Properties​

augmenterId​

readonly augmenterId: string

Defined in: packages/agentos/src/cognition/rag/RetrievalAugmentor.ts:62

Implementation of​

IRetrievalAugmentor.augmenterId

Methods​

checkHealth()​

checkHealth(): Promise<{ details?: Record<string, unknown>; isHealthy: boolean; }>

Defined in: packages/agentos/src/cognition/rag/RetrievalAugmentor.ts:1370

Returns​

Promise<{ details?: Record<string, unknown>; isHealthy: boolean; }>

Inherit Doc​

Implementation of​

IRetrievalAugmentor.checkHealth

deleteDocuments()​

deleteDocuments(documentIds, dataSourceId?, options?): Promise<{ errors?: object[]; failureCount: number; successCount: number; }>

Defined in: packages/agentos/src/cognition/rag/RetrievalAugmentor.ts:1228

Parameters​

documentIds​

string[]

dataSourceId?​

string

options?​
ignoreNotFound?​

boolean

Returns​

Promise<{ errors?: object[]; failureCount: number; successCount: number; }>

Inherit Doc​

Implementation of​

IRetrievalAugmentor.deleteDocuments

embedTexts()​

embedTexts(texts): Promise<number[][]>

Defined in: packages/agentos/src/cognition/rag/RetrievalAugmentor.ts:766

Batch-embed arbitrary texts with the augmentor's embedding pipeline. Exposed so consumers (e.g. CitationVerifier wired through the agent's verifyCitations: { retrievalAugmentor } shortcut) can share the same embedder rather than configuring a second one with a different model.

Parameters​

texts​

string[]

Returns​

Promise<number[][]>

Implementation of​

IRetrievalAugmentor.embedTexts

ingestDocuments()​

ingestDocuments(documents, options?): Promise<RagIngestionResult>

Defined in: packages/agentos/src/cognition/rag/RetrievalAugmentor.ts:312

Parameters​

documents​

RagDocumentInput | RagDocumentInput[]

options?​

RagIngestionOptions

Returns​

Promise<RagIngestionResult>

Inherit Doc​

Implementation of​

IRetrievalAugmentor.ingestDocuments

initialize()​

initialize(config, embeddingManager, vectorStoreManager): Promise<void>

Defined in: packages/agentos/src/cognition/rag/RetrievalAugmentor.ts:96

Parameters​

config​

RetrievalAugmentorServiceConfig

embeddingManager​

IEmbeddingManager

vectorStoreManager​

IVectorStoreManager

Returns​

Promise<void>

Inherit Doc​

Implementation of​

IRetrievalAugmentor.initialize

registerRerankerProvider()​

registerRerankerProvider(provider): void

Defined in: packages/agentos/src/cognition/rag/RetrievalAugmentor.ts:222

Register a reranker provider with the RerankerService.

Call this after initialization to add reranker providers (e.g., CohereReranker, LocalCrossEncoderReranker) that will be available for reranking operations.

Parameters​

provider​

IRerankerProvider

A reranker provider instance implementing IRerankerProvider

Returns​

void

Throws​

If RerankerService is not configured

Example​

import { CohereReranker, LocalCrossEncoderReranker } from '@framers/agentos/cognition/rag/reranking';

// After initialization
augmentor.registerRerankerProvider(new CohereReranker({
  providerId: 'cohere',
  apiKey: process.env.COHERE_API_KEY!
}));

augmentor.registerRerankerProvider(new LocalCrossEncoderReranker({
  providerId: 'local',
  defaultModelId: 'cross-encoder/ms-marco-MiniLM-L-6-v2'
}));

retrieveContext()​

retrieveContext(queryText, options?): Promise<RagRetrievalResult>

Defined in: packages/agentos/src/cognition/rag/RetrievalAugmentor.ts:773

Parameters​

queryText​

string

options?​

RagRetrievalOptions

Returns​

Promise<RagRetrievalResult>

Implementation of​

IRetrievalAugmentor.retrieveContext

setHydeLlmCaller()​

setHydeLlmCaller(llmCaller): void

Defined in: packages/agentos/src/cognition/rag/RetrievalAugmentor.ts:266

Register an LLM caller for HyDE hypothesis generation.

HyDE (Hypothetical Document Embedding) improves retrieval quality by generating a hypothetical answer first, then embedding that answer instead of the raw query. The hypothesis is semantically closer to the stored documents, yielding better vector similarity matches.

The caller must be set before HyDE-enabled retrieval can be used. Once set, HyDE can be activated per-request via options.hyde.enabled on retrieveContext, or it can be activated globally by passing a default HyDE config.

Parameters​

llmCaller​

HydeLlmCaller

An async function that takes (systemPrompt, userPrompt) and returns the LLM completion text. The system prompt contains instructions for hypothesis generation; the user prompt is the query.

Returns​

void

Example​

augmentor.setHydeLlmCaller(async (systemPrompt, userPrompt) => {
  const response = await openai.chat.completions.create({
    model: 'gpt-4o-mini',
    messages: [
      { role: 'system', content: systemPrompt },
      { role: 'user', content: userPrompt },
    ],
    max_tokens: 200,
  });
  return response.choices[0].message.content ?? '';
});

shutdown()​

shutdown(): Promise<void>

Defined in: packages/agentos/src/cognition/rag/RetrievalAugmentor.ts:1399

Returns​

Promise<void>

Inherit Doc​

Implementation of​

IRetrievalAugmentor.shutdown

updateDocuments()​

updateDocuments(documents, options?): Promise<RagIngestionResult>

Defined in: packages/agentos/src/cognition/rag/RetrievalAugmentor.ts:1345

Parameters​

documents​

RagDocumentInput | RagDocumentInput[]

options?​

RagIngestionOptions

Returns​

Promise<RagIngestionResult>

Inherit Doc​

Implementation of​

IRetrievalAugmentor.updateDocuments