The paper should be accepted in rough form, but publication should be held up pending approval of the data and code. Or a paper should be published in a probat

The paper should be accepted in rough form, but publication should be held up pending approval of the data and code. Or a paper should be published in a probationary form, and not certified (by the journal) until an independent lab replicates the result. A paper that isn't making adequate progress toward replication should be retracted by the publishing journal.