rag-from-scratch and ragmate-lagacy
These are complementary tools: the first provides educational foundations for understanding RAG architecture (embeddings, retrieval, generation), while the second applies those concepts as a practical code-indexing server that performs retrieval-augmented completions for editors.
About rag-from-scratch
pguso/rag-from-scratch
Demystify RAG by building it from scratch. Local LLMs, no black boxes - real understanding of embeddings, vector search, retrieval, and context-augmented generation.
Implements a modular, JavaScript-based RAG pipeline with progressive learning examples covering embeddings, in-memory vector indexing, and retrieval strategies including hybrid search, multi-query decomposition, and query rewriting with LLM fallbacks. Built entirely with local models (via node-llama-cpp) and includes reusable library components for caching, normalization, and result fusion techniques like reciprocal rank fusion.
About ragmate-lagacy
ragmate/ragmate-lagacy
Local RAG server for code editors. Scans your codebase, builds a local context index, and connects to any external LLM for context-aware completions and assistance.
Implements local semantic search over your codebase using embeddings and file change tracking, injecting relevant code snippets into JetBrains AI Assistant prompts via an HTTP bridge. Supports any LLM provider (OpenAI, Mistral, local models) with pluggable embedding models, and automatically reindexes your project while respecting Git branch context—all running in Docker without external cloud dependencies.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work