Now we want to add some filtering cap from VDD(5V) to ground on chip. There is only space left under the PAD. Some support adding some PIP caps. But others concern about the CP test security: the probing needle may add stress on the PIP cap and make it short. Does anyone have such experience? Thanks.
This is between you, your foundry, your probe floor and
assembly house. The whole production chain needs to be
OK with circuit-under-pad and have qualified the reliability
under the kinds of post-fabrication abuse that the pad
will receive. Not just time=0 shorting, but latent damage
needs to be proven absent at acceptable quality levels.
Possibly? Yes. Provably, to the satisfaction of your end
customer? That's the question for you.
Not easily provable, as dick_freebird indicated. Better recommend to add an external 0.1¢ ceramic cap (with a lot larger cap value than attainable on silicon) close to the supply pins - exactly where you want it.