The installation of an "upgraded" version of the OpLevs toward a low-frequency yaw- and length-sensing kind of OpLev has to be done for all main mirrors in KAGRA. However, the PR2 chamber is different from all the other chambers as there is only one viewport available for the (external) OpLevs.

In order to measure the reflected OpLev beam, we either have to install the beam collimeter inside the PR2 chamber so that the reflected beam goes through the only viewport outside, or to install a mirror in the inner wall (proper positioned), so that the beam, coming from outside, is subsequentally reflected directly toward the viewport.
I checked therefore the feasiblity of the second option by using the actual (as I hope) drawings of the PR2 chamber and the payload. It seems that, taken only the drawings, there is no reason why it shouldn't work. I prepared some pictures to see the paths of the OpLev beam (green line) within the chamber. By installing a mirror, it may be possible to guide the reflected beam toward the viewport.

However, there is still an open question how much space the additional structure on the inner wall (small mirror + mount) will take and whether or not it will collide with the main beam passing PR2 chamber toward the BS.