Scepter of Insight

The math is unforgiving: three mana to cast, then four more every time you want a card, all of it locked inside an artifact that produces nothing the turn it arrives. That is seven mana invested before the first draw, and every activation thereafter costs as much as some whole card-advantage spells of its era. The closest functional ancestor is the activated-draw artifact, the lineage that runs through things like Jayemdae Tome, where the permanent is a mana sink rather than a payoff. This sits near the bottom of that family: it survives a board wipe and keeps drawing, which is the entire pitch, but the activation rate is steep enough that it only made sense in a grindy, flooded game with nothing better to spend the surplus on. The design tension lives in the activation cost's split, three generic plus one blue pip, so it is reusable but never cheap, and the colored requirement means even the engine itself is not splashable into just any deck. What you are buying is durability of the engine, not efficiency of any single card. In a world that has since produced repeatable draw at far kinder rates, it reads as a relic of a slower design philosophy, from a time when a permanent that drew cards at all earned a slot regardless of what each card actually cost.