discovered 03 Aug 2026
RA3G-Agent
→ View on GitHubRA3G-Agent is a policy-aware RAG (Retrieval-Augmented Generation) multi-agent AI system designed for document querying without the need for external APIs. It features a robust architecture comprising a Retriever, Reasoning, and Governance agents, enabling automatic filtering of sensitive information and maintaining session memory for context. Key capabilities include a user-friendly web interface, real-time logging, automatic PDF uploads for vector storage, and full programmatic access through a REST API built with FastAPI.