I actually published a paper on squirrel carnivory back in 1991 in the Texas Journal of Science where a fox squirrel consumed part of a baby bluejay as the squirrel was being harassed by several other bluejays and a mockingbird that were all making alarm calls while this was going on.
A couple of years later, Joan Callahan published an extensive review of squirrel carnivory with some interesting observations (Callahan, J. R. 1993. Squirrels as predators. Great Basin Naturalist 53:137-144).
In 2002, I published another paper in the Journal of the Minn. Academy of Science on carnivorous foraging behavior of a squirrel that harvested insects off the radiators in a hotel parkinglot.
It was suggested above that the squirrel might be rabid. Carnivory in squirrels is not indicative of being rabid. With some exceptions such as the Franklin Ground Squirrel that has a particular fondness for frogs, Callahan noted that in observations of squirrel carnivory, the squirrels tended to be females that were either pregnant or nursing young (not at that particular moment, but back in a nest) as indicated by the extended nipples. The purpose for this activity pertained to dietary needs. A pregnant or lactating female squirrel can become depleted of key nutrients and minerals and one way they compensate is through carnivory.
In the 1991 paper, the bird carcass was examined after being abandoned by the squirrel The key points of attack on the carcass were areas where bone was close to the skin and not covered with much muscle. In other words and based on Callahan's study, the squirrel was after the calcium in the bones.
In the 1992 paper, the squirrel was observed with extended nipples, indicating she was pregnant or lactating...matching Callahan's pattern of most carnivory in squirrels being by females in dietary distress. Note the extended nipples in the second image.