I'm happy to see there's an option to use this with an AST as input, more tools like this should follow suit. Hopefully it can then push us to a place where the

I'm happy to see there's an option to use this with an AST as input, more tools like this should follow suit. Hopefully it can then push us to a place where there's a standard JS AST such that we don't reinvent that wheel over and over. Babel seems to be winning here, but I don't think it matters so much which one wins so long as any one does. This tool looks interesting, particularly the future direction of it, but I'm weary about efficiency claims without a single runtime metric posted. The claims may be true, initializing is costly, but so is parsing huge unrolled loops. For an optimization tool, I'd hope to see pretty strong metrics to go along with the strong claims, but maybe that is coming? Interesting work, nonetheless!