Dear Magdaléna,
following HGVS nomenclature recommendations you should indeed specify the inserted sequence. Reporting NM_206933.4:c.14951_14952insN[81], so not specifying the 81 nucleotides is not very useful, e.g. one can not predict the consequences on protein level of the variant.
In general, the inserted sequence comes from "somewhere", so specifying it based on the reference sequence of the origin is one option. For an example, see Notes on
https://hgvs-nomenclature.org/stable/recommendations/DNA/insertion/. Another option mentioned on this page is to submit the inserted sequence with some flanking nucleotides to GenBank and use the accession.verion number obtained as source of the inserted sequence.
Best regards,
Johan den Dunnen
HUGO HGVS Variant Nomenclature Committee (HVNC)