rlama

Why we included this project

rlama is a command-line tool for anyone who keeps Ollama running on their own hardware and wants to ask questions over their own documents without shipping them to a cloud service. You build a retrieval-augmented generation index from a folder of files or a crawled website, then query it through your local models. A few extra commands let you watch a directory or a web page and refresh the index when new content shows up. Since everything goes through a local Ollama endpoint, it fits teams that need answers from internal material and care about keeping that material off remote servers. The project is on a maintenance pause right now, so treat it as a capable starting point for a self-hosted document QA setup rather than a product you can count on for frequent updates.