Benchmark Runner

Benchmark small, medium, large, maximum and problem-specific worst-case structures when practical.

Do not assume maximum size alone is worst-case. Use structural analysis to seek inputs that activate the expensive branch of the reference algorithm.

Record runtime and, where available, memory, exit status and input dimensions.

Execution constraints

Run this skill as part of a single-agent, strictly serial workflow. Do not spawn subagents and do not execute multiple OJ tasks concurrently. Any executable work must use the agent-generated, OS-adapted protected toolchain with 1 GiB RAM and 1 second limits. Prefer C++ for executable components.

I/O contract — immutable

The original I/O mechanism is part of the problem specification. Never convert file I/O to stdin/stdout or vice versa. Preserve exact input/output filenames and case. Preserve problem.xml judging input-file/output-file values and problem-properties.json inputFile/outputFile values. If standard I/O is specified, do not invent filenames. If file I/O is specified, generated solution/brute/benchmark/protected execution must use the exact filenames. Any mismatch requires a source-level fix and regeneration.

Bilingual tutorial — required

The successful default package contains both English and Vietnamese tutorials whenever tutorials are supported. Preserve supplied tutorial text in its source language and faithfully translate the missing default language. Keep problem.xml, tutorial files and problem-properties.json synchronized.