Having thought about this for a while but never having pulled the trigger, I had some time over Christmas to play around with cursor.ai and found it so reduced the overhead of getting a flex project up and running that now I can announce:
This is a very simple command-line program that reads 6502 assembler code from “early” dialects, especially the Microtek toolchain, and outputs a bunch of statistics. Some are basic, like the number of lines in assembler and number of bytes in the binary, but others look for possible optimizations and offer suggestions on how to update the code to take advantage of 65C02 instructions (and CE will follow).
As the Microtek tool is mostly associated with Atari’s internal work, I suspect this will not work well with your own assembler examples. That’s exactly why I’m posting here, please try this out on widely-used assemblers of the era (as opposed to modern ones like c65’s) and I’ll try to get as many working as possible.
In particular, if someone here is more familiar with the PET and C64 than I, I’d love it if you could try it out on the original code for those machines and help me get it running for them as well.