He wasnt staunchly against metaphysics just a certain kind of metaphysics and he felt that metaphysics could be reformed through epistemology, and its clear to any good commentator that he was not totally against metaphysics for example in his 1783 work Prolegomena to Any Future Metaphysics this short polemic work was written by Kant after what he had felt had been the poor reception of The Critique of Pure Reason. Prolegomena to Any Future Metaphysics stated the importance of the Critique of Pure Reason as developing a project for the science of Metaphysics.

He was against what he called, "dogmatic metaphysics". Not against metaphysics. Indeed, he claimed that he had written the First Critique to save metaphysics from the attack made on it by "the astute Mr. Hume". That is why he wrote that whether synthetic a priori judgments were possible (the main theme of his First Critique) was, "a matter of life and death for philosophy".

Kant wishes to save Metaphysics from itself, and from the kind of devastating arguments exemplified by Hume. As he makes clear in the Preface to the second edition, he wishes to do that by finding limits to legitimate ("scientific" in the sense that it is productive of knowledge) areas of discourse and investigation by reason. He seems well aware that his analysis has, by its nature, a limiting effect on metaphysics because it rejects other, traditional, areas as idle speculation and dogmatic (and illegimate)metaphysics that has brought it into disrepute.

The comments above are all fine, but Kant also gave his critical conditions for developing metaphysics as a science. There is debate whether or not the kind of a priori, objectively valid, and universal judgments found in geometry (exampled by Kant) can hold true for metaphysics. This is what Kant was looking for. The kind of dogmatic metaphysics he was against, I believe, had to do with the dogmatic insistence that certain systems of metaphysics had proved the existence of God, but Kant argued they had not. And hence, he set forth just what was in demand: 'Synthetic judgments or propositions' that were objectively valid, and universally true, just like the propositions of geometry. Metaphysics has not and still does not consist of these kind of judgments, and it is this kind of existing metaphysics that Kant was opposed to (mysticism and all the rest ... consisting of conjectures, fantasies, and illusory beliefs). The kind of proof Kant demands is not too difficult to understand, but in so much talk and study of Kant, this side of Kant is hardly if ever mentioned. That's why I offered an example of what Kant demanded in other threads. I'll repeat the example here of an objectively valid, universal judgment and also a pure philosophical one that would be true of a science of metaphysics (though one does not yet exist). A sphere holds a certain density. Infate the sphere and the density increases. Compress the sphere and the density increases. Simple. This judgment can be thought in our minds, using pure reason and nothing else ... no actual experiment would be necessary, although it could be proved by experiment, and has been. This does not imply that the judgment is empirical in nature. We can also visualize the process and the necessary consequence. I could go further with other examples should this not suffice. But it is this kind of reasoning that Kant was looking for.