Aha! I think I found the answer to my question... the cp value of 1225 above is because Black is about to move and for Black the position is 12.25 centipawns ahead. I was operating under the falsehood that the score was positive if White was up and always negative if Black was up. (enlightenment found on stackexchange.com)
and just to confirm, from the UCI protocol: the score from the engine's point of view in centipawns.
Now when I google for this in a month I'll find this answer quickly 
I've run into a situation where I just don't understand what Stockfish 12 (compiled for Mac, installed via homebrew) is telling me with its cp number returned.
I start stockfish as normal in iTerm with "stockfish" command. Then I enter the following:
uci
ucinewgame
isready
position startpos moves d2d4 g8f6 c2c4 c7c6 d1a4 d7d5 a4c6
go depth 1
(I left off Stockfish's obvious responses)
Stockfish then responds to the "go" command with:
info string NNUE evaluation using nn-82215d0fd0df.nnue enabled
info depth 1 seldepth 1 multipv 1 score cp 1225 nodes 11 nps 11000 tbhits 0 time 1 pv b8c6
bestmove b8c6
My confusion is why is the cp value equal to 1225 when white is about to lose the queen and be very far behind? In fact, taking the queen with b8c6 yields a cp of -1182 which makes sense to me.
That 1225 is lost on me though. (yes, it's a very contrived position... I was testing around to see what the info lines looked like for a mate in # and stumbled on this)
Thanks for any light you can shed.