Somebody defined the operation of addition—it did not arise out of pure thought alone, as is evidenced by the fact that nobody bothered to define some other operation by which two compounds could be combined to produce a lesser quantity of some other compound (at least until people began formalizing chemistry). There are an infinite number of possible operations, most of which are completely meaningless for any purpose we would put them to. Knowledge of addition isn’t knowledge at all until you have something to add.
The problem here is that you seem to be presupposing the odd idea that, in order for any proposition to be knowable a priori, its content must also have been conceived a priori. (At least for the non-Kantian conceptions of the a priori). It would be rare to find a person who held the idea that a concept be acquired without having any experience related to it. Indeed, such an idea seems entirely incapable of being vindicated. If I expressed a proposition such as “nothing can be both red and green all over at the same time” to a person who had no relevant perceptual experience with the colors I am referring to and who had failed to acquire the relevant definitions of the color concepts I am using, then that proposition would be completely nonsensical and unanalyzable for such a person. However, this has no bearing on the concept of a priori knowledge whatsoever. The only condition for a priori knowledge is for the expressed proposition be justifiable by appeal to pure reason.
The problem here is that you seem to be presupposing the odd idea that, in order for any proposition to be knowable a priori, its content must also have been conceived a priori. (At least for the non-Kantian conceptions of the a priori). It would be rare to find a person who held the idea that a concept be acquired without having any experience related to it. Indeed, such an idea seems entirely incapable of being vindicated. If I expressed a proposition such as “nothing can be both red and green all over at the same time” to a person who had no relevant perceptual experience with the colors I am referring to and who had failed to acquire the relevant definitions of the color concepts I am using, then that proposition would be completely nonsensical and unanalyzable for such a person. However, this has no bearing on the concept of a priori knowledge whatsoever. The only condition for a priori knowledge is for the expressed proposition be justifiable by appeal to pure reason.