Leaderboard detail

Terminal Core fixed-model preview

Seed-data comparison for control-plane UI validation.

Entries
2Preview
Suite version
1.0.0-previewPreview
Top mean score
0.878Preview
Aggregation
1.0.0Preview
Preview operational data. Origin may be seed, mock-eval, or runner.
RankHarnessMeanPass rate95% CIRunsMean cost
1Structured ReAct0.878100%[0.828, 0.917]6$0.0065
2Baseline Loop0.49717%[0.292, 0.672]6$0.0055