Here’s a description of how it works as currently implemented.

https://grapevine-brainstorm.vercel.app/#/about/graperank

I say “as currently implemented” bc currently I’m calculating one GrapeRank score using follows and mutes and that’s it. But the idea is that we can calculate multiple GrapeRank scores, each its own context, each using distinct data inputs which are “interpreted” using unique scripts. I probably should expand upon that point in the about section.