* Added `--repo` and `--name` flags for running only examples with a specific name, or repo (substring matching) * Fixed a race condition that caused hangs when running multiple examples at the same repo and sha * Fixed a bug where scoring was completely wrong because I had passed the arguments to `apply_diff_to_string` in the wrong order * The current evals now run quickly and without errors. Release Notes: - N/A