Methodology
What this is, and what it is not
Most tools in this corner of the market sell detector evasion. We sell the opposite. Here is the whole posture, in plain terms, so you know exactly what you are paying for.
What it is
An AI voice editor. You give it a draft and a sample of how you actually write. It rewrites the draft so it sounds like you, then a deterministic checker scans the result for AI tells and forces a fix before you see it. You get the rewrite plus two honest scores and a receipt showing every fact was kept.
What it is not
It is not a humanizer that plays games against AI detectors. That arms race makes writing worse, and we will not sell it. We do not promise your text will beat any detector, ever. If you want to fool a classifier, this is the wrong tool.
The score never lies
The anti-slop score is computed by a deterministic auditor, not a model asked to grade its own work. When a rewrite cannot honestly hit your voice, the score says so and you get a caveat. We would rather lose you on a truthful number than keep you on a flattering one.
Facts are never invented (Fact-Lock)
Every rewrite is checked against your source. Names, numbers, URLs, and handles have to survive, and nothing new can be introduced. If the tool ever invents or drops a fact in a rewrite, that month is free. It is enforced by the same checker that scores the text, not a marketing promise.
Your writing trains your voice, and only yours
The samples you save sharpen your own voice match. We do not pool your writing into a shared model for other people. Your history and your voiceprint stay yours, and you can export or delete everything at any time.
Why this holds up
A voice that sounds like you is useful for as long as you write. A trick that beats this month's detector is worthless next month. We built the durable thing on purpose.