Raycaster / evalsBack to SpreadsheetBench

SpreadsheetBench

13-1

1/1Pass

SpreadsheetBench task 13-1. Inspect the exact spreadsheet prompt, compare published model runs, open the agent response trace, and review grades for workbook editing.

Spreadsheet editingDual harness
ssb-13-1
SpreadsheetBench
2 models
Dual (parsed + original)
PromptWorkbook objective and answer cells
ResponseOpen each trace to inspect tool use and edits
GradeCompare score and pass/fail by model

Task prompt

What the agent was asked to do

You are solving a spreadsheet benchmark task in a real workbook. Objective: Produce the correct final workbook state for the expected answer region. What matters: - Only the values in the expected answer region will be graded. - The workbook is the answer. Instructions: 1. Read the workbook and inspect the relevant data region first. 2. Infer the required result for the provided workbook instance. 3. Write the final value(s) directly into the expected answer region. 4. Do not rely on prose, formulas in your chat response, pseudocode, or VBA as the answer unless the benchmark explicitly requires those to be written into cells. 5. If the natural-language task asks for a general method, formula, or macro, convert that into the concrete result needed for this workbook instance. 6. Keep your final text response short and only summarize the workbook cells you changed. Relevant data region(s): A1:E56 Expected answer region(s): A3:D32 Expected answer sheet(s): LISTS Task: How can I combine data from a 'RANGES' sheet to a 'LISTS' sheet, by matching duplicates based on the 'DATE' and 'REF' columns in columns B and C, and add a 'TOTAL' row to sum up the amounts? Additionally, since new ranges with headers will be added to the 'RANGES' sheet, how do I ensure to delete the old ranges in the 'LISTS' sheet before populating the new data? I have given the correct answer of the "STAGE" table in the "LISTS" table as a format reference, please help me complete the other answers. Each of the sections in "LISTS" should be sorted based on data from lowest to highest, secondary sort key is alphabetical A to Z on column C.

Published trajectories

Agent runs on this task

Curated dual-harness runs. Best scored run per model.

ModelHarnessScoreResultLinks
GPT-5.4showcasedual1/1Pass
Gemini 3.1 Produal1/1Pass