Raycaster/ Eval

APEX-Agents · Management Consulting

Task_World130_CamilleMoingeon_4

Best published0/5Fail

APEX-Agents task Task_World130_CamilleMoingeon_4 in AI Agents for Digital Transformation. Compare dual-harness agent runs across models, scores, and public traces.

AI Agents for Digital TransformationManagement Consulting World 130Dual harnessGrader: rubric
task_2516425094194379af7c5b6d5180a608
Management Consulting World 130
message_in_console
5 models · dual config

Task prompt

What the agent was asked to do

Can you look at the Frito-Lay case study and apply their downtime reduction to HarFeast Good Group's number in the baseline file? I want to estimate what the improvement would look like for us (rounded to the nearest full percentage point). Output the information in a message here.

Published trajectories

Agent runs on this task

Curated dual-harness runs (parsed + original sandbox). Best scored run per model.

ModelHarnessScoreResultLinks
Gemini 3.1 Produal0/5Fail
GPT-5.4dual0/5Fail
GPT-5.4 minidual0/5Fail
GPT-5.4 nanodual0/5Fail
GPT-5.5dual0/5Fail

Grading rubric

Rubric criteria

Runs are graded against these criteria. Open a run for model-specific verdicts.

  1. States that the new unplanned downtime ratio for the Rockford, Illinois plant is 14%

  2. States that the new unplanned downtime ratio for the Madison, Wisconsin plant is 14%

  3. States that the new unplanned downtime ratio for the Cedar Rapids, Iowa plant is 13%

  4. States that the new unplanned downtime ratio for the Toledo, Ohio plant is 14%

  5. States that the new unplanned downtime ratio for the Kalamazoo, Michigan plant is 15%