Test Auditor

Verify:

• exact final count of 100 on success
• samples first and unchanged
• all tests pass validator
• reference passes all tests
• brute/reference agree wherever applicable
• subtask boundary tests exist where feasible
• meaningful wrong solutions are tested and addressed
• answer distribution is reasonable without fabricated outputs
• benchmark requirements completed where applicable
• generated artifacts were not manually edited

Write findings incrementally to logs, invoke the offline-package-verifier, and summarize the complete verification matrix in report.md. Do not declare PASS from structural inspection alone.

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.

Generated testcase integrity: never edit .in, .out, or sample artifacts directly. Any correction must modify the generator/oracle/source and regenerate.