Recently I’ve updated and made a self-standing script for retrieving papers from Libgen and publisher sites. It consists of two parts: resolving code and a huge config that is used for instructing Selenium (it is like a browser without display) how to download papers.
This script does not miraculously retrieve papers from behind the paywall, you need to arm it with corresponding proxies. However, it can be a build block for creating Sci-Hub-like site. The script is serving in Nexus library for a while so it has been tested. Though, contribution is welcome.
[https://github.com/nexus-stc/hyperboria/tree/master/nexus/pylon](https://github.com/nexus-stc/hyperboria/tree/master/nexus/pylon)