The question that arrives the moment a premise finally works: has someone already done this?
Not the finished manuscript. The paragraph. The point where the idea has just cohered, you are about to commit months to it, and the honest answer to "is this too close to something" is that you do not know - you only know the books you happen to have read.
Checking then costs an afternoon. Finding out at chapter thirty costs the draft.
Up to three closest existing works, each with the shared structure named in a sentence. A plain statement of which element is doing the derivative work. One concrete pivot. A verdict, including the uncomfortable ones.
Two recorded examples worth reading before you use it:
Nearly every novel shares tropes with hundreds of others, and a tool that flagged that would be useless. What makes a premise derivative is the arrangement - the same elements in the same order doing the same job. Synorigin is built to name the arrangement, not to make you afraid of tropes.
The verdict stays direct when the news is bad. That is the point - a check that tells everyone their idea is fresh is worth nothing, which is why a clean result is deliberately rare and held to the same standard of evidence as a real match.
It will not write for you, rewrite for you, or generate prose. It identifies the gap and hands it back. It also will not reproduce copyrighted text - if you ask to see the passage you should avoid, it will name the work and describe it instead.
Ideas you submit are not sold and are not used to train any model, and you can delete any chat or your whole account at any time. The privacy policy is written in plain language and states exactly where your data goes, including that it is processed across borders.