You can define energy in an accelerating frame, and you do it every day. The surface of the earth is an accelerating frame.
Sometimes you say a frame is close enough to inertial and just treat like it is inertial even though it isn't inertial and hope for the best.
Other times you just have to sit down and learn how to do physics in a noninertial frame. In GR the inertial frames are infinitesimally large, so you can do calculus fine, but anything other than looking at a rate at a point requires a noninertial frame for a finite sized coordinate patch.
But you can handle acceleration and handle noninertial frames without doing GR.
But there are things that break about energy. For instance, energy is not conserved in general Relativity, that's life. Potential energy doesn't belong to an object it belongs to a system that is exerting instantaneous forces on each other. So in special relativity it doesn't make sense unless that energy exchange is happening at the same event, contact forces or interactions that happen at the same place and time together.
So things you are familiar with are going to change if you study enough physics. Potential energy will go away and you will have energy densities in space and Lagrangians instead. Energy won't be conserved, but that will describe the actual world you live in. It is essential to life, things acquire energy as gravity contracts regions, that energy is then available thermodynamically and it gets expelled from a region leading to things not reversing.
The lack of energy conservation in GR is about how energy density relates to energy, and it comes from the geometry of space being strange. It isn't because of a violation of local (infinitesimal) energy conservation.