mcpField
L
LLM RAG

LLM RAG

github.com/binarybanaOther2Updated 87w ago
Repository

About LLM RAG

LLM RAG is a local, embedded retrieval-augmented generation project for building a searchable knowledge store from code, URLs, and PDFs. It uses LlamaIndex for processing, Gemini for embeddings, and LanceDB for vector storage.

Key features

  1. 01Ingest code, URLs, and PDFs
  2. 02Process documents with LlamaIndex
  3. 03Generate embeddings with Gemini
  4. 04Store vectors in LanceDB
  5. 05Run a search server

Use cases

  1. 01Build a local search store from code
  2. 02Index URLs for retrieval
  3. 03Search PDF documents

MCP support

The project describes the knowledge store as MCP-enabled, but the provided README does not specify its MCP interface or server behavior.

Details

From the project README

LLM RAG

A RAG (Retrieval Augmented Generation) implementation using LlamaIndex for document processing, Gemini for embeddings, and LanceDB for vector storage.

Setup

This project uses uv for dependency management and direnv for environment management. To get started:

  1. Install dependencies:
# Create and activate a new virtual environment
uv venv
source .venv/bin/activate

# Install dependencies
uv pip install -e .
  1. Set up environment:
# Create .env file with your Google API key
echo "GOOGLE_API_KEY=your_key_here" > .env

# Allow direnv to load the environment
direnv allow

Usage

Data Ingestion
python -m llm_rag.ingest --source /path/to/source --type [code|url|pdf]
Search Server
python -m llm_rag.search --db /path/to/lancedb

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.