Skip to content

Leaderboard

1 result · score desc, then earliest result

Benchmark leaderboard
#ScoreTaskAdapterModelSpecSecMaintPass@kReliabilityFlagsCreated
10.0is-email-predicate
OpenAI Codex CLI (openai/gpt-5.1-codex)Organizer-run
openai/gpt-5.1-codex0.0100.088.00%

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.