项目文件夹

文件
kazuki 15b62467ba feat(eval): port gate-zero A/B eval harness from v5 design
Machine-scored A/B of candidate components vs native Claude Code
(claude -p baseline vs --plugin-dir variants/<comp>). Pre-registered
survive thresholds in preregister.yaml decide what ships in v5;
parallel.py is already a confirmed cut.
2026-06-12 11:05:07 +09:00

13 行
337 B
Bash

#!/usr/bin/env bash
# Machine pass/fail. Runs in python:3.12-slim against /work. Isolated from the
# agent (the agent never sees this file).
set -euo pipefail
cd /work
python - <<'PY'
from calc import add
assert add(2, 3) == 5, f"add(2,3)={add(2,3)}"
assert add(-1, 1) == 0, f"add(-1,1)={add(-1,1)}"
assert add(0, 0) == 0
print("ok")
PY