Jenia,
As I recall with Agilent data, the issue is that in conversion, the scan numbers are totally exaggerated in the mzXML. I'm not sure the reason specifically for Agilent. I'd recommend try converting to mzML and maybe it resolves that issue?
Alternatively, you could convert to mzXML and then open the file in an mzXML viewer to find the appropriate scan for the library spectrum you want to add with the new scan number.
Another alternative is to export a single scan as an mgf file (or mzXML if possible) from MassHunter and simply add that single spectrum to GNPS libraries.
Ming