I would like to say first that I have no medical training and so, I'm far from an expert. However, I did look into the possibility of retractable claws in the past, and came to several conclusions.
The answer, in my opinion, can be divided into two cases:
1. Retractable claws in a human hand structured as it is normally.
In this case, I don't believe such claws are possible. The human hand is a complex thing and there are tendons and nerves that run through the back of the hand that would have to be 'rewired' or moved for such an arrangement to work. The structure as it is cannot support anything more than there is.
2. Retractable claws in a human hand with alternative structure.
This case, I believe, holds more potential. Such 'claws' would have to be made out of or bone or better, keratin, with a likewise ability to be regrown if broken and be sharpened manually. They would not be as long as Wolverine's claws, as they can only be as long as the cavities holding them and the size of such cavities is highly restricted by the size of the part of the hand they sprout from.
There are three options for the placing of such claws:
1) Claws located in the fingers.
This is the location I initially researched, as my need was for feline-like claws.
Such claws would have to be short and probably curved, as there is not enough room within a finger, even a restructured one. I immediately ruled out the tip of the finger because it wasn’t what I was looking for. I wanted normal fingernails and hidden claws. This left the middle and proximal phalanges of the fingers. While the proximal phalanx of the fingers is longer and offers room of larger claws, I believe the middle one would offer more mobility, but essentially, the same would work for both.
I believe the structural changes necessary would be as such:
(pictures modified by me from one found here)
The phalanx bone would have to be split into two bones and the 'sheath' and mechanism would have to be fitted between them. (As in, muscles, tendons and the claw itself). The tendon running through the back of the finger would have to be rerouted, maybe split along the bones and mended at the next knuckle, so that when the hand is curled and the claws extend, the tendon would be held out of the way on both sides. Same for nerves.
the claw mechanism itself would have to resemble somewhat of a feline's claw mechanism so that they are unsheathed when the fingers are curled.
Note that this would undoubtedly make the fingers thicker, and the corresponsive phalanges more fragile and more susceptive to compression pain.
2) Claws located along the metacarpal bones (the back of the hand).
In this approach, we can either employ the same method as with the finger claws, namely splitting the metacarpal bones and fitting the mechanism between the split bones, or fit the mechanism between the existing bones.
there is also the option of making the bones hollow and the spines be sheathed inside them. However, I am not sure how the sheath/unsheath mechanism would work in this case. Perhaps by combination of a ring of muscles at the exit of the sheath and a piston-like effect of a liquid or gas filled gland. This, however, would make the mechanism more complex than my level of anatomic knowledge can scope. It would also require lubrication within the ducts. It can be achieved by oil glands or some liquid like saliva.
In any case, these claws would need to be straight and would not be much longer than the finger claws.
3) Claws located along the forearm bones.
This, I feel, is the most plausible of the choices, since it requires less modifications. In this case, since there are already two bones in the forearm, the structural difference would have to be made between the bones to accommodate a long, narrow sheath in which one claw or spear like protrusion will be housed. It too will be made of keratin or bone, but I cannot envision more than a single such claw in each forearm. It can, of course, be as long as the forearm itself which would make it more viable as a weapon.
The unsheathing mechanism would have to consist of rings of muscles and tendons, a lubricating system and a fleshy sheath, since the normal claw sheath found in felines would not support such a claw. It is simply too long and would have to absorb or filter debris to some extent, as a means of hygiene. The forearms would look thicker, but not by much.
Such a claw would probably have a detachment mechanism since it can be torn or broken. The mechanism would guarantee there would not be lasting damage to the muscles and tendons of the retraction mechanism. The claw would then regrow, much like a horn or a nail.