Pyticles is my messy googlecode project: http://code.google.com/p/pyticles/
OK, first problem is that the design of the force classes is ugly. Overwhelmingly forces will be one body or pair forces within the system itself so why make the two system approach, for which the only application at the moment is system boundaries, the default case for which it is designed.
The motivation for this was to be able to use generic force calculation functions – better I think to write the functions (e.g. gmm/r2) seperately and bind them in the class defintions for one-system and two-system forces.