no-ai-slop

Why

It names the patterns instead of asking for "more human", and that's the whole difference. Binary contrasts ("It's not X. It's Y"), throat-clearing openers, faux-insight setups, colon reveals, dramatic one-line fragments, importance puffery, synonym cycling, fake-profound endings — twenty or so, each one a thing you can point at in a draft. A named pattern is checkable. "Sounds like AI" is a feeling, and you can't edit against a feeling.

Three modes matter in practice: it edits and tells you what it removed, it detects without editing, and it will deliberately generate slop if you want to see the patterns exaggerated.

Newest thing in the hub — added on the strength of the method, not yet on a piece I've published.

Watch out for

A pattern-stripper is only as good as its false-positive rate. Several of these patterns are fine prose when you meant them; the risk is flattened voice, not slop.