It might be helpful to imagine yourself very small and riding on the surface of the rotor with a magnetometer.
Torque reverses when slip changes sign. That's the RPM where your magnetometer would show flux reversing its direction of apparent motion, ie North ceases to overtake you and you commence to overtake North, passing it once per slip cycle.
The only place where slip changes sign is at zero slip, ie synchronous speed.
So torque ought to asymptotically approach zero. Well, discounting friction and windage. Make that electromagnetic torque .