To get a good rule-of-thumb for this in your brain, you can think of it in terms of vector mathematics.
Consider a circular orbit. Now consider two vectors: one in the direction of that orbit, one perpendicular to it. A naive inclination change from 0 degrees to 90 degrees, with no other change to the orbit's dynamics (i.e. same eccentricity, same periapsis / apoapsis), require the velocity to change from parallel to the first vector to parallel to the second, at the same magnitude. That's a total change of SQRT(2 * starting_velocity^2) ~= 1.4 * starting velocity. If we consider a velocity that is LEO (about 7.8 km/s), our total velocity change would have to be about 10.92 km/s.
Putting a satellite into LEO from Earth's surface only costs between 9 and 10 km/s from gravity loss, steering, and wind resistance. Earth is a deceptively expensive gravity well to be doing inclination change maneuvers in.
> Can we go the other way and use the ISS as like an orbital garage to capture and fix satellites?
Moving the whole station doesn't really make sense. Even a vehicle would burn a lot of fuel, and would only really work, once again, in the same (or close to the same) inclination.
Inclination changes are so expensive that there was at least one case where an operator sent a satellite all the way to the Moon and back, because it cost less fuel that way.