MiniRAG »

Local hybrid search engine for your docs with MCP integration for AI agents

Why Local-First The use case that shaped this project is narrow on purpose: a personal corpus of markdown and text files — books on refactoring, prompting, and architecture — that an AI agent like Claude Desktop or Codex should be able to query as a grounded tool. No multi-tenant …

X-RAG »

Distributed Search & Indexing for Retrieval-Augmented Generation

The Problem with Single-Modality Retrieval Most RAG tutorials converge on the same architecture: embed your documents, store them in a vector database, query at runtime. That works until you have real queries, real traffic, and real tenants — at which point it starts failing in …

No posts match your filters.