--Silverman and Tate, "Rational points on elliptic curves": the proof of Bezout's theorem (in one of the appendices) is given as a series of exercises with hints.
--Munkres, "Topology": there's a section on nets that is almost entirely exercises.
--Hartshorne, "Algebraic Geometry": A large number of results are given as either a series of exercises, or a "single" exercise with many "parts."
And the list goes on. And on.
Linguistic note: I don't think there is any two-word phrase in English that captures what you mean. "Proof via exercises" or "Exploration via exercises" or "Exposition via exercises" is probably the best you're going to get. Although you should probably look up "Moore method," if you are not familiar with this term already.
]]>