Bounded model checking
You will learn
How a bounded search explores every input up to k cycles, and what that proves.
Bounded model checking is exhaustive search with a horizon: every input, every state, up to k cycles -- does any of them break the assertion? Inside the bound it is a proof, not a sample; no random seed can miss what the search enumerates. The lesson's spec carries the shape: a depth, a timeout, and properties to check. The widget is a bounded search in another domain: 84 board and model logs scanned for a key that must appear in none -- 0 hits, 0 logs too big to read -- absence, checked exhaustively rather than assumed.