Results Archive
July 3, 2026 ยท View on GitHub
Current results live in README.md. Latest full matrix: full-k7-2026-07-03.md.
Older reports are kept for provenance and comparison. They should not be read as the current benchmark result unless explicitly referenced by a newer report.
Full Matrices
| Report | Scope | Status |
|---|---|---|
| full-k7-2026-07-03.md | Five scenarios, Codex and Claude, git, but+skill, and jj+skill, k=7 | Current |
| full-k5-2026-07-01.md | Five scenarios, Codex and Claude, git, but+skill, and jj+skill, k=5 | Superseded |
| full-k5-2026-06-29.md | Five scenarios, Codex and Claude, git, but+skill, and jj+skill, k=5 | Superseded |
| full-k5-2026-06-28.md | Five scenarios, Codex and Claude, git and but+skill, k=5 | Superseded |
| full-k5-2026-06-27.md | Five scenarios, Codex and Claude, git and but+skill, k=5 | Superseded |
| full-k5-2026-06-26.md | Five scenarios, Codex and Claude, git and but+skill, k=5 | Superseded |
| full-k3-2026-06-22.md | Five scenarios, Codex and Claude, git and but+skill, k=3 | Superseded |
Per-Scenario Reruns
| Report | Scenario / Focus | Status |
|---|---|---|
| pilot-1-but-latest-2026-06-25.md | Pilot 1, Codex and Claude but+skill reruns | Historical |
| pilot-1-codex-but-clean-harness-2026-06-25.md | Pilot 1, clean-harness but+skill reruns | Historical |
| pilot-1-current-batch.md | Pilot 1 early batch | Historical |
| pilot-2-current-batch.md | Pilot 2 early batch | Historical |
| pilot-2-codex-but-latest-2026-06-25.md | Pilot 2, Codex but+skill reruns | Historical |
| pilot-2-claude-but-latest-2026-06-25.md | Pilot 2, Claude but+skill rerun | Historical |
| pilot-3-current-batch.md | Pilot 3 early batch | Historical |
| pilot-3-claude-but-clean-config-2026-06-25.md | Pilot 3, Claude but+skill config-isolation debug | Historical |
| pilot-4-reorder-guidance-fix-2026-06-25.md | Pilot 4 reorder guidance fix | Historical |
| pilot-5-codex-but-latest-2026-06-25.md | Pilot 5, Codex but+skill rerun | Historical |
| pilot-5-claude-but-latest-2026-06-25.md | Pilot 5, Claude but+skill rerun | Historical |