What problem does it solve? Turning a raw competitive programming problem idea into concrete, defensible numbers — difficulty, constraints, scoring format, and subtask weights — is error-prone and often skipped, leading to problems where the naive solution passes or subtasks reward nothing distinct. ## Core Features & Use Cases - Originality Check: Strips flavor from the idea and checks whether the core operation is already a known problem on Codeforces, AtCoder, or OI archives. - Constraint Separation Analysis: Computes the N where the intended solution passes the time limit but the naive solution fails, using a ~10^8 operations/second budget across every size parameter. - OI vs ICPC Scoring: Proposes a subtask ladder where each rung pays for a distinct algorithmic insight, with point weights summing exactly to the configured total. - problem.json Generation: Writes the schema-1 problem.json file (format, limits, constraints, subtasks, checker kind) that downstream skills consume for test generation and Polygon upload. - Use Case: You have an idea for a graph problem and need to decide whether it works as an OI problem with three subtasks or a single ICPC group, and what N separates the O(n log n) intended solution from the O(n^2) brute force. ## Quick Start Ask the assistant to shape your problem idea into constraints, a scoring format, and a subtask ladder written to problem.json.
