codeflash-agent/evals
Kevin Turcios 66187bbcc3 fix: v2 eval runner — shallow cached clones + non-interactive prompt
- Shallow clone (--no-checkout --depth 1 + fetch specific commit) instead
  of full clone — 15s vs 2+ min for large repos like codeflash-internal
- Cache clone in evals/repos/<name>/workspace/, cp -r for each run
- Use gh repo clone for private repo auth
- Fix eval prompt to skip skill's AskUserQuestion step in non-interactive mode
- Gitignore workspace/ dirs
- Update intro.md with v2 eval docs
2026-03-27 07:27:12 -05:00
..
repos/codeflash-internal-psycopg-serialization feat: eval v2 — real-repo evals cloned from git 2026-03-27 07:25:10 -05:00
templates Hello World 2026-03-24 16:14:04 -05:00
.gitignore fix: v2 eval runner — shallow cached clones + non-interactive prompt 2026-03-27 07:27:12 -05:00
run-eval.sh fix: v2 eval runner — shallow cached clones + non-interactive prompt 2026-03-27 07:27:12 -05:00
score-eval.sh Hello World 2026-03-24 16:14:04 -05:00
score.py Hello World 2026-03-24 16:14:04 -05:00