ResearchTasksAgent Task Utility and Admission IntegrityAgent Task Utility and Admission IntegrityFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedCode-agent exact-effect benchmarkCXI231STC Count1Jul 8, 2026AgentDojo (live task bank)CXI199STC1Jul 8, 2026