It's not a completely trivial scenario unfortunately, given that a severed limb would mostly span at least two or even more bones (e.g. upper leg, lower leg, foot) and multiple attachments. Sometimes you want to render the whole skeleton in a single render pass (when all limbs are visible on screen), sometimes you would prefer the mesh to be split (when limbs are outside the screen and you don't want to render them). So it depends on your scenario, sometimes you might even be better off preparing simple Unity physics GameObject chains for each limb (without a Spine skeleton behind it), and then when severing the limb only hiding the respective attachments in the skeleton, and showing the prepared limb GameObjects.
You can also have a look at the Spine Examples/Other Examples/SkeletonUtility Ragdoll scene to see a slightly different approach, with more direct scripting via the SkeletonRagdoll2D example component instead of using multiple SkeletonUtilityBone components.
