little tool for offline wikipedia RAG

Excerpt

I was bored and handwrote a tiny 100-line bash script to let an agent search for and read articles from an offline wikipedia archive during a regular chat. It's not particularly useful, but it's definitely neat and a big step up from llama-cli. I got surprisingly good results even with LFM2.5-2B (Q8 at 0.5 temp), which is usually too weak to be of much use. (and Qwen had no trouble, obviously— see main image) Just point it at your llama-server endpoint and chat away! Or whatever. The handwritten