Dretske is very clear about both the aim and the structure of this paper so I'll let him guide you through it, and I'll just offer a supplementary note and a couple of questions to think about.

When introducing the idea of intensions (with a 's') as way of getting at intentionality, Dretske refers to Chisholm. Although he doesn't give a specific reference in his note 2, you can take him to be referring to the discussion you have read (see, in particular, Chalmers, p. 485, col. 2). You've also encountered the idea of intensions in connection with the example of "the morning star" and "the evening star," which differ in their intensions.

Dretske says at a couple of points that what he refers to as "original" intentionality cannot be produced at will. Is he right about this? Ask yourself why he would say that a computer could not exhibit it and whether he is right. Ask yourself also whether it could be produced by genetic engineering (e.g., the design of an organism of a sort that might have arisen from an evolutionary history).

Finally, do you he is right that his recipe is a recipe not only for representation but also for thought?