Skyline is telling you that the .msf file does not contain q-values (i.e. false discovery rate).

Skyline says that you can create a library from your .msf file, but you need to set the Cut-off score to 0.
The default cut-off score is 0.95.

I have attached some pictures that show where the cutoff score textbox is on either the Build Library dialog, or the Import Peptide Search dialog.

If you change the circled number to 0, you should be able to create a library from your .msf file.

I just want to make sure that every time someone suggests using a cut-off of 0, we also point out that this is not advisable unless you have already applied some form of error rate control elsewhere and then exported only a set of results with a controlled error rate to the .msf file you are trying to build your library from.

If your .msf file contains all peptide spectrum matching results without error rate control, then simply importing them all into a spectral library with a cutoff of 0 is a bad idea. You will likely end up with a spectral library where at least half the results are peptides incorrectly matched to MS/MS spectra not actually produced by those peptides.

So, take care using a cutoff of zero. It is the ultimate escape when your spectrum matching files do not contain a probability value Skyline understands. But, make sure you know what you are doing.

--Brendan