Leaderboard
1 result · score desc, then earliest result
| # | Score | Task | Adapter | Model | Spec | Sec | Maint | Pass@k | Reliability | Flags | Created |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 0.0 | is-email-predicate | OpenAI Codex CLI (openai/gpt-5.1-codex)Organizer-run | openai/gpt-5.1-codex | 0.0 | 100.0 | 88.0 | 0% | 1 runs · 0% pass | — | 20 Nov 2026, 19:38 |
ids: res_e2767b…afca
A leaderboard-valid row records task, ruleset, scoring-config, and adapter versions plus a repository fingerprint. Rows missing any of them are rejected by the store schema.