The National Education Policy 2020 asks Indian universities to bring Indian Knowledge Systems (IKS) into the curriculum, and the Ministry of Education has set up an IKS Division to fund the work. A great deal of teaching material has followed. What has not followed is any shared criterion for deciding which parts of the enormous Sanskrit, Prakrit, Tamil and vernacular inheritance a university can responsibly teach as knowledge, and which parts are better described as cultural heritage, as matters of faith, or as later interpretation. Some scholars worry that IKS without such a criterion will lose academic credibility. Others answer that the demand for external validation misses the point of the exercise. This paper suggests that both concerns can be met from within the tradition itself. The Indian tradition has its own theory of what counts as valid knowledge, worked out over many centuries in the pramāṇa debates of the Nyāya, Mīmāṃsā, Buddhist and Jaina schools; it also has an explicit method of empirical testing (parīkṣā) in the Caraka-Saṃhitā and a practice of mathematical demonstration (upapatti) in the astronomical commentaries. Using these sources, I propose a four-tier classification of IKS material by epistemic status, and I show how it applies to real cases in medicine, mathematics, grammar and logic. The framework is Indian in origin and, at the same time, acceptable to any modern university. It allows IKS to be taught with both respect and rigour.