What problem does it solve? Writing correct, limit-fitting C++ for a competitive programming problem is error-prone: choosing the wrong complexity class, missing edge cases, or shipping an unverified greedy can sink an entire problem package. This Skill turns an intended algorithm (or a pasted statement) into a working model solution, a brute-force oracle, and per-subtask solutions with stress testing. ## Core Features & Use Cases - Model and brute-force implementation: Produces sol-main.cpp for the intended algorithm and sol-brute.cpp matching the statement, plus one real solution per OI subtask rung. - Complexity budgeting: Maps maximum N to feasible complexity classes (from O(N!) to O(log N)) so the chosen algorithm fits a 1–2 second limit before any code is written. - Stress testing guidance: Runs the model against the brute oracle on random tests mixing tiny N and the largest N the oracle can finish, catching wrong greedy or unproved invariants. - Constant-factor optimization reference: Ships a black-magic toolbox (pragmas, fast I/O, cache layout, SIMD intrinsics) for solutions that are asymptotically optimal but still TLE. - Use Case: While preparing a Polygon package, you have an intended O(N log N) solution and need the model, an O(N²) brute oracle, and subtask ladders coded and stress-tested before test generation begins. ## Quick Start Ask the assistant to implement the intended solution and a brute-force oracle in C++ for your problem, then stress test them against each other.
