anti-ai-writing

Edits any piece of writing to remove AI tells. Strips banned words, banned phrases, and banned structures (em dashes, the default rule of three, bullet lists with bold titles, formulaic transitions) at the sentence level, plus discourse-level tells that survive a surface edit (stated takeaways, single-track structure, manufactured closure, atmospheric openers). Includes a strict review mode that audits a draft and returns a verdict instead of a rewrite. Returns a clean rewrite plus a diagnosis