C. Basso implemented a very clever limiting circuit for the output voltage, see Fig. 2.47, page 153, Second Edition. There are some other features he added for the ideal opamp throughout the book, but for the subject at hand that one is your reference. You must be careful now how you define the internal current source, to limit the peak current to reasonable values, even if you must push some coefficients to the capacitor and resistor expression (check my model). Only then the idealized diodes will be able to reasonably clamp the internal voltage levels. And the clamp mechanism must be internal, without any residue on any of the opamp pins (the pole model is internal). To the best of my knowledge, there is no other mechanism I was able to find to limit unreal current and voltages, except for what Basso came out with.
Just a side note, do not assume that a basic model “works”, only because you tested a simplified little circuit with a basic GND connection. A first or a second order model is just a distinction of no importance, if the model shows random currents and voltages: it could have some limited academic value, but for real life products what matters is realistic power rail currents and how the model works on floating applications. You risk to blow up your prototype if you do not really verify the models as close as possible to real life conditions. If you’re more of a software guy, you will always be tempted to ignore and discard real life product conditions. For the hardware side guys, that’s all that matters, we cannot afford to blow up in smoke expensive hardware and work time. Most of the opamp applications are not connected to GND, but floating. Signal conditioning, current sensing on the positive or negative rail, all circuits with ground bouncing (pretty much any application) will require floating opamps models. If you take your time to check the default IP model from QSPICE you will see an extreme error on the power rails currents, and completely random step response on the output, with gigantic and unexplained spikes.
For reference you should check this post: Generic OpAmp problems