ArweaveRAG Agents

ArweaveRAG

ArweaveRAG enables AI agents to retrieve provenance-verified information from permanent knowledge bases stored on Arweave.

Problem Statement

Current Retrieval-Augmented Generation (RAG) systems suffer from fundamental trust and longevity issues:

• Traditional RAG systems reference URLs that can go offline

• No reliable way to verify whether cited sources are authentic

• Knowledge bases depend on centralized providers

• AI agents are trained on unreliable or manipulated data

• No proof exists for what data an agent accessed to generate responses

Solution

A decentralized RAG system that provides:

• Permanently stored documents on Arweave

• Vector embeddings with cryptographic source provenance

• Auditable retrieval receipts

• Permanent citation links (ar://...)

• Cross-session agent memory

Technical Stack

ComponentTechnology

Amadeus Integration

FeatureUsage

Key Files

• docs/CONCEPT_DECK.md — Full concept deck

• diagrams/architecture.md — Architecture diagrams

• src/rag_contract.ts — Smart contract

• src/sdk.ts — Amadeus SDK

Revenue Model

ActionFee (AMA)