paperclip-runner evalsRecorded evaluation

paperclip-runner eval suite

Provider-reported list cost$0.000000
Estimated model cost$0.010008

Current result for every test and model/configuration combination. Each cell uses only the most recent attempt for that combination.

Test
gpt-5.4-mini
live-codex-pinned · codex_app_server · provider default
block-taskPASS
block-task-narrative-baselineFAIL
finish-taskPASS
finish-task-narrative-baselineFAIL
request-human-confirmationPASS
request-task-reviewPASS
schedule-task-wakePASS
set-task-dependenciesPASS