Detecting whether you have light, is easy. Detecting what frequency of light you have exactly, is a whole different ballgame.
You need a mirror to focus the light, a diffraction grating (maybe a CD would work), and a photodiode array. Light hits the grating and you get a rainbow coming off. That gets focused with the mirror onto the PDA in such a way that it covers the whole strip of photodiodes. Takes a bit of calibration to figure out which wavelengths each diode is seeing. Probably get a good lesson in instrumentational amplifiers in the process. I know I did.
I built a PDA based spectrophotometer one weekend as a rainy day project. It's not terribly difficult. But I used a low end PDA that probably isn't good enough for what you want to do. Better PDAs can be had if you got the money. I posted it on another forum but unfortunately all the pictures have disappeared from that post. So it's not very helpful.
I think the code to drive the PDA is even on my github still. Same name on github except a hyphen instead of underscore. Delta-G instead of Delta_G.