Since every modern browser, Javascript runtime and filesytem supports unicode in their URIs/filenames I think it's fair to accept this is as a correct path

We could add an option to try to substitute certain characters using a function like (https://github.com/sindresorhus/transliterate) but it still would not support languages with another script such as Chinese or Japanese

In this case we should find a solution for finding the matching route when retreiving it through the cms.maybeGet function. Because it's resulting a notfound:

const data = await cms.maybeGet(MySchema({path: slug}))

This snippet is resulting in a notfound when the slug contains č.

To fix this we can decode the received slug:

const data = await cms.maybeGet(MySchema({path: decodeURI(slug)}))