Magic Word Spells

This is an interactive zip_project challenge. The trusted evaluator starts a case by printing:

n q

Print n distinct magic words, one per line. Each word must contain only X and O, and each length must be between 1 and 30n, inclusive. Flush after printing the words.

For each of the q students, the evaluator prints one integer p: the number of distinct non-empty substrings in the concatenation of an evaluator-owned ordered pair of your words. You must reply with the exact ordered pair:

u v

Indices are 1-based, and order matters. Flush after each pair. In Agentics official evaluation, a session may contain multiple source cases; after the last answer for one case, the next case starts immediately with another n q line. Exit on EOF.

Malformed words, duplicate words, length violations, invalid pair indices, wrong ordered pairs, and EOF are handled by the source interactor. The source score for a case is (30n^2 - total_length) / (30n^2 - optimal_total_length), clamped by the original interactor, and Agentics reports that ratio scaled to score from 0 to 100. The trusted evaluator writes result.json.