TL;DR lots of energy, and super-strength materials.
To harness the power of a volcano you need some advanced materials and drilling techniques. In several novels there are materials that either natively or through electromagnetic or handwavy modifications are "indestructible" for this purpose (i.e. they can withstand pressure, torque and heat well enough); Laurence Dahners' stade, David Adams' indestructium and so on. Drive as many double pipes as you need inside a volcano, then pump water down the inner pipe, get water vapour off the outer pipe, and hey pronto!, you can run a turbine setup.
This is much more difficult because you need to interfere with the "hot spot" under the volcano. And the hot spot has an incredible quantity of heat you need to get rid of. You don't want that heat in the environment, so you need tuned radiative coolers (basically, black bodies heated at exactly one thousand Celsius degrees, surrounded by materials that reflect infrared below 8 micrometers but are trasparent above that. This allows to radiate heat through a clear atmosphere and into space).
The difficulty here is that you would need an enormous radiative surface.
The mechanism for both is the opposite of the above. You need to heat up a hot spot; this would be done by increasing the radioactive heat plume underneath. Some sort of "neutrino laser" would need to be focused in the exact volume, to increase thermal emission through reverse beta decay. This assumes that suitable isotopes are present in that volume and in sufficient quantities. Also, since neutrinos are absorbed only with great difficulty, this method is horribly inefficient, requiring a monstrous amount of energy, and a measurable negative effect would manifest for a significant distance around the focus.
Other means of transferring energy deep underground might involve focused seismic waves, or very powerful nuclear fusion devices delivered through shafts (indestructium drills again required).
In some places you might just need to open a shaft and let internal pressure do the rest (for example, Dahners' stade could be used to drive a pipe, stazed in cylindrical sections, at practically any depth. Once enough material was removed from the inside of the pipe, the pressure would do the rest.
Drilling again is required, plus some way of exactly mapping stresses in rock. Once you know how a fault line is holding, you can frack the key points to release compressive stresses a little at a time, converting a five-minute 7.0 Richter scale quake into a five-year long sequence of piloted low-threshold temblors. Or you can cut around the fault line, again providing release (the underground compression will close the cut, relieving itself. Then you reopen the cut. An earthquake can move the fault line by up to two meters: if you provide those two meters by way of a cut, the compressive force will go into sealing the cut).
With enough energy and waste heat management, you can maybe do this with a laser (the water table would be a significant problem though: and you need to drive the cut all the way to the depth of the epicenter. For the San Andreas fault line, that's at least fifteen kilometers).