First, let us consider that the heliosphere represents, for the practical purposes of this exercise, the extent of the sun's magnetic field. In this field, there exist the other magnetic fields of celestial bodies, as well as those bodies which are composed of magnetic materials, but possess no active or permanent field of their own. As well, there are the fields associated directly with active regions on the sun, which need to be treated separately from the primary solar field. With that knowledge, we can view the heliosphere as a body of great dynamic magnetic complexity, with nested field interacting together to apply force on one another, and on the constant plasma flow through the solar system.
In practical application, only a few tools used to access publicly available resources are required for quite accurate prediction of solar flares. The first of these is a solar system simulator, such that one can track the positions of planets on particlar dates. One such tool for popular operating systems on personal computers, available for free, is Celestia. Another choice is the cross platform, web based and mobile app solar system model implementation Solar System Scope. These applications provide the geometrical configuration of the solar system at any given time, which is critical data if we are to determine the magnetic interaction of planetary bodies, the heliosphere, and active regions on the sun.
The next dataset we need to reference is the position and state of active regions on the sun. There are several different tools which can allow us to view images of the sun, but the use of STEREO and SDO gives us the most complete view, despite the loss of communication with STEREO B. Many spot groups persist through multiple rotations of the sun. With these we can simply calculate the solar longitude of the active region based on rate of rotation at it's latitude. When considering active regions which are newly formed on the far side of the sun, STEREO gives us our only view until the active region approaches the horizon, when it becomes visible by its extruding plasma loops.
With this information in hand, we can begin to model the geometrical conjunctions of active regions and planetary bodies. While a complex computer simulation of the interacting magnetic bodies would be preferable for optimal accuracy, the geometric method is simple, and provides viable results. Put plainly, the planetary bodies, by means of influence on the magnetic "shape" of the heliosphere, also influence the stability of solar active regions. The approach of an planet to conjunction with an active region, particularly an inner planet, will tend to destabilize the active region, causing flaring.
It should be noted that this method is not an exhaustive solution for predicting all solar flare events. There are other flare triggers, including sundiving comets, and the more commonly seen interactions of complex spot groups. A complete flare prediction model would include these triggers, however, the techniques described herein can be used to accurately predict flares triggered by planetary body interaction.