Simulate Grid Infection

Implement a grid infection simulator. Base problem: - You are given a 2D grid. - `X` means infected. - `.` means healthy. - Each day, every currently infected cell simultaneously infects all 8 neighbors that are still healthy. - A cell infected on the current day cannot spread until the next day. - Return the number of days until the system becomes stable, meaning no new cells are infected. Required edge cases: - Empty grid - No initial infected cells - Already fully infected grid - Single row or single column - Multiple infection sources - `1 x 1` grid Common follow-ups: 1. Add immune cells `I` that can never be infected and block spread. 2. Add recovery: an infected cell becomes immune after `D` days, and you must return the day when no active infected cells remain. 3. Add threshold infection: a healthy cell becomes infected only if at least `T` of its 8 neighbors were infected at the start of the day. 4. Add a countdown or death state, which requires careful synchronous state transitions. The main skills tested are multi-source BFS, simulation with synchronous updates, boundary handling, and off-by-one correctness.