i think the proof is in understanding how LLMs work, not a certain prompt. this because output isnt determenistic and hence a single prompt couldnt (theoretically) reliably prove anything. thats the exactly problem :')