True, that can be a valid use-case. Another solution would be to leverage language agnostic API discovery solutions, generating the serialisation and part of validation code (swagger/open api/...). This way you provide a nice documentation and don't have to care about language differences while keeping good productivity.