In theory (note the word “theory”), it should be possible to make a rig in Poser that creates the anaglyph optically at the time of rendering. You’d need some very specific objects - effectively an anaglyph converted fro the camera you were using. You’d need to create a beamsplitter (which is a half-silvered glass mirror) to allow only 50% of the light through one side and combine it in such a way that the camerra sees the ready-made anaglyph.

Position this to the donny camera and parent it properly, and you’ll have a rig that can be moved, animated etc. without any need for Python intervention. It would only need one render cycle, and no need for specific software to combine the two images in post.

That’s the theory.