run_eval.py

#!/usr/bin/env python

"""
Entry point for running evaluations on all candidate implementations for each task.

This script uses the evaluator defined in ``src/evaluator.py`` to iterate over task
candidates, run their test suites, and print a concise summary. It keeps the
functionality minimal by design, since the goal is to demonstrate how a small
evaluation harness works rather than to build a full‑featured tool.
"""

from src.evaluator import evaluate_all_candidates


def main() -> None:
    results = evaluate_all_candidates()
    print("\n=== Evaluation summary ===")
    for r in results:
        print(
            f"[{r['task_name']}] {r['candidate_name']}: "
            f"{r['passed']} / {r['total']} tests passed "
            f"({r['status']})"
        )


if __name__ == "__main__":
    main()