If you mostly use chatbots and recently started tracking new AI tools, this is the trap: you scroll past a new update, then wonder if you just missed the one thing that should change your next move.
The useful read on mvanhorn / last30days-skill is not better search. It turns likes, upvotes, and betting odds into research ranking. That means it is trying to sort what people are reacting to, not just what a plain search box can return.
If you only follow surface hype, you can waste time, budget, and attention in the wrong place. The quieter cost is that you keep circling whatever looks busy and miss the one change that actually shifts your decision.
A tool update is not worth tracking because it lists more features. It is worth tracking if it changes your next decision.
That is why the README and changelog matter. The README frames the ranking around upvotes, likes, and real money. Then the changelog makes the logic concrete: in v3.4.0 on June 18, 2026, the project gave more weight to crowd votes. Sixteen days later, v3.10.0 changed how top community comments from different sites were ranked. The pattern is hard to miss: people reaction is not a side signal here. It is the ranking logic.
That does not make it a truth machine. High engagement is not the same as high truth, and that is the boundary to keep in mind. The value here is faster filtering when you want to see what is pulling real attention right now.
If you know someone who uses AI chat tools to keep up with launches but keeps drowning in noise, share this with them. last30days-skill looks most useful when the question is not what exists, but what people are actually piling onto right now.