One of the challenging research questions in soft robotics is the establishment of valid mathematical models and control techniques under limited feedback conditions. A soft robotic system will typically have a continuum or monolithic body seamlessly embodying its actuation and sensing elements, accompanied by mechanical structure, and even the energy source, much like a biological system. In this study, we propose a relay sliding mode controller based on the input-output model (RSMC-IO), which is a simple but effective control approach that does not require a complete system model, to control the tip position of a soft prosthetic finger, which typifies a monolithic soft robotic positioning system. The feedback signal corresponding to the bending angle of each joint of the finger is provided by highly stretchable soft strain sensors made of a silicone (EcoFlex 0010) substrate containing channels filled with carbon paste as the conductive component. An analytical expression between the fingertip and joint angles of the finger is derived and experimentally verified. This expression is accurate enough to estimate the joint positions (i.e. joint angles) as well as the fingertip. The experimental results show the efficacy of the proposed controller, which has not been proposed for prosthetic fingers and similar soft robotic positioning systems before.