Hi! No we don’t use FROST. Essentially we implemented the basic protocol described in the introduction of [GS23], where pre-signature attacks are mitigated using re-randomization of presignatures using a random beacon, as described in [S23]. The idea is that later on we could incorporate some of the more advanced optimizations described in [GS23].
thx alex. this points to a github search on dfinity/ic, meaning there is probably multiple components that comprise the threshold schnorr implementation.
for others coming across this, i think good entrypoints are