Business Bench

Tasks · desk · Spreadsheet

Fall to spring enrollment, adds and drops by course

class-roster-changes · timeout 1200 s · source

The ask

Sent verbatim as the prompt. The agent works in a copy of the workspace below and sees nothing else.

Renata needs fall-to-spring enrollment changes for every course for the board packet. Save it as roster_changes.csv; her email explains what she wants counted.

Workspace

  • email_from_coordinator.txt
  • fall_2025_rosters.xlsx
  • spring_2026_enrollment_export.csv
  • spring_2026_waitlist.csv

Expected deliverables

  • roster_changes.csv

Checks

The task passes only when every required check passes. Reference answers, tolerances, and the planted traps are in the repository, not here.

CheckTypeRequiredFailed (frozen) / 6
requested columnscsv_columnsyes0
one row per coursecsv_set_equalyes0
fall enrolled per coursecsv_values_matchyes0
spring enrolled per coursecsv_values_matchyes0
added per coursecsv_values_matchyes0
dropped per coursecsv_values_matchyes0

Recorded results, complete desk comparison

pass pass, frozen scorer only fail timed out
SystemRepetitionsFrozen / 3Raw / 3Median minutesMean cost (USD)
Proto / DeepSeek V4.1 Flash332.00.0313
Codex / GPT-5.6 sol331.90.3866

Per-attempt check verdicts, raw and frozen verdicts, usage, receipt and artifact hashes for this task are in results/latest/attempts.jsonl under "task": "class-roster-changes".