You are given `n` axis-aligned rectangular buildings on a flat ground. Each building `i` is described by `(Li, Ri, Hi)` where `0 <= Li < Ri` is its left/right x-coordinate and `Hi > 0` is its height. Compute the outer silhouette (skyline) formed by the union of all buildings. Return the list of **critical points** `[x, h]` — the points where the silhouette height changes — sorted by `x` ascending. The first coordinate of each key point is the x position; the second is the height there. Consecutive key points must have **different** heights (no redundant points), and the silhouette must terminate at a key point of height `0` where the rightmost building ends. If there are no buildings, return an empty list. This is the classic 'Skyline Problem'. Aim for `O(n log n)` time. **Tie handling (important):** when multiple building edges share the same x, the algorithm must process them so the reported height is the true maximum across that x. Entering edges should be considered before leaving edges, among entering edges higher heights take effect first, and among leaving edges lower heights leave first — a sweep-line with a max-heap (and lazy deletion of ended buildings) achieves this naturally. **Example:** `buildings = [[2,9,10],[3,7,15],[5,12,12],[15,20,10],[19,24,8]]` → `[[2,10],[3,15],[7,12],[12,0],[15,10],[20,8],[24,0]]`.