One of the promises of machine-actionable linguistic data linked to biblical texts is the enablement of new types of language learning tools. At their simplest, such tools might involve adding the necessary scaffolding to enable students to read more text than they otherwise might by providing glosses for rarer words or help on idioms, irregular morphology, and unusual syntactic constructions. Such tools, however, are hardly novel and have long been manually produced in printed form. Equivalent electronic versions don't really take advantage of what's possible. In this paper I discuss an online reading environment for Ancient Greek, and the Greek New Testament in particular, that takes advantage of the availability of open, machine-actionable resources such as treebanks and morphological analyses for more automated and consistent generation of scaffolding but which goes a step further by being adaptive to an individual student's knowledge at a given point. Such knowledge need not be explicitly provided (although it can be: to align with a particular textbook, for example). It can also be built up implicitly from what the reader is requesting more information or help on: What words are they having trouble remembering the meaning of? What forms are they having trouble parsing? The model of student knowledge is then integrated with learning tools such as spaced-repetition flash cards and parsing drills with the results of these tools then feeding back into better adapting scaffolding for reading. The online reading environment is open source and potentially applicable to a wide range of other language and texts provided the necessary linguistic data is available.