For people who mostly use chatbots and are starting to track AI tools, the easiest mistake here is reading last30days like search. That sounds harmless, but it is how you lose time, budget, and attention on noise when all you wanted was one fast call on what matters next.
The contrarian part is simple: based on public repo docs, last30days is not behaving like a neutral search tool. It turns likes, upvotes, and betting odds into research ranking [C002]. If you just saw the name and assumed it was search with fresher results, you are already asking it the wrong question.
That matters because the real scene is ordinary now. A new AI tool shows up on your timeline. You are about to scroll past it, but you also do not want to be late to the one thing that might change your next move. In that moment, the wrong read is expensive. You end up following surface heat instead of signals that actually explain why something is climbing.
The proof is not one slogan. The README frames the product around people signal rather than editor signal. Then the changelog makes the weighting logic look deliberate, not decorative: version 3.4.0 on 2026-06-18 added crowd-vote weighting, and version 3.10.0 on 2026-07-04 rewrote cross-platform ranking for top community comments. Based on those public docs, this was still true at least through those versions.
So the useful takeaway is not better search. It is this: do not judge an update by the feature list. Judge it by whether it changes your next decision. For mvanhorn / last30days-skill [C001], the decision change is obvious: stop reading the order as truth-ranked, and start reading it as attention-ranked.
That boundary matters. Attention is not truth. High engagement is not the same as being correct. Use it to see what people are boosting and where money may be leaning; do not use it as your final truth machine. If you know someone who keeps treating social-signal tools like search, share this with them.