Spreadsheet charts
Create and format charts to an exact specification.
20 project-authored proposals · taxonomy v1.1 · 0 new benchmark runs. Verifiers and partial-credit rules below are designs, not executed results.
- Typical applications
- Excel · Numbers · Google Sheets
- Class GUI-need label
G2· GUI-essential
Task proposals · 20
Expand to inspect the complete original wording and proposed checks. Original proposal text remains English in both interfaces.
R04.03.01.03-T01 Chart monthly revenue 2024 vs 2025 as a branded clustered column
Original instruction
In fixtures/monthly_rev.xlsx sheet 'Data' (A1:C13), insert a clustered column chart on sheet 'Chart': 2024 series #1F4E79, 2025 series #F28C28, title 'Monthly Revenue 2024 vs 2025', axis titles 'Month' and 'Revenue ($k)', legend at bottom, gap width 80%. Save.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clicktype· Text & forms- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 2 / 5
Proposed verifier
Parse chart XML: barChart barDir=col grouping=clustered, 2 series referencing Data!B2:B13/C2:C13, srgbClr fills exact, title/axis titles exact, legendPos b, gapWidth 80. Chart must carry Excel-authored c16 extension ids (absent in openpyxl output). VBA, COM and AppleScript automation disabled, no Python in the VM.
Proposed partial credit
30% correct type and series refs; 25% colors; 30% titles and legend; 15% gap width.
R04.03.01.03-T02 Plot three cities' weekly temperatures as a Numbers line chart
Original instruction
Open fixtures/climate.numbers in Numbers. From table 'Weekly' create a 2D line chart of Oslo, Lisbon and Cairo (52 weeks), y-axis from −10 to 40 °C with major steps of 10, x-axis labels every 4th week, legend at top, chart title 'Weekly mean temperature 2025'. Save.
- Platforms
mac· macOS- Applications
- Numbers
- Interfaces
desk· Desktop GUI- GUI need · task setting
G3· GUI-only- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clicktype· Text & forms- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 2 / 5
Proposed verifier
Grader exports the sheet to PDF and .xlsx via its own AppleScript: 3 line series with correct data ranges, axis min/max/steps, title; rendered chart compared to reference PNG (SSIM ≥0.90). osascript/JXA is blocked for the agent; the file must be saved by Numbers (IWA metadata).
Proposed partial credit
30% chart with 3 correct series; 30% axis scale; 20% label interval and legend; 20% title + render match.
R04.03.01.03-T03 Build a revenue-and-margin combo chart with a secondary axis
Original instruction
fixtures/quarterly_pl.xlsx 'Summary' A1:C9 holds Quarter, Revenue, Margin %. Create a combo chart: Revenue as clustered columns on the primary axis, Margin % as a line with markers on a secondary axis fixed 0%–40%, data labels on the line only (0.0%), no gridlines on the secondary axis. Save.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clicktype· Text & forms- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Chart XML: one barChart + one lineChart, line on secondary valAx with min 0 max 0.4, dLbls only on the line series with numFmt 0.0%, no majorGridlines on secondary axis; series refs exact; c16 ids present (Excel-authored). VBA, COM and AppleScript automation disabled, no Python in the VM.
Proposed partial credit
30% combo structure; 30% secondary axis with bounds; 25% labels on line only; 15% gridline setting.
R04.03.01.03-T04 Create an EBITDA bridge waterfall with total bars set correctly
Original instruction
fixtures/ebitda_bridge.xlsx 'Bridge' lists FY24 EBITDA, 6 drivers (positive and negative) and FY25 EBITDA. Insert an Excel waterfall chart where FY24 and FY25 are set as totals, increases are #2E7D32, decreases #C62828, totals #546E7A, connector lines on, data labels in $m with one decimal. Save.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clickhover· Hover & context- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
chartex part (cx:chart) layoutId='waterfall' with subtotals idx for the first and last points, data point colors per sign, connector lines enabled, label numFmt. openpyxl cannot write chartex; VBA, COM and AppleScript automation disabled, no Python in the VM.
Proposed partial credit
30% waterfall chart from right range; 30% totals set; 25% colors by sign; 15% labels/connectors.
R04.03.01.03-T05 Fit a trendline on a log-scale scatter and show equation and R²
Original instruction
fixtures/assay_curve.xlsx has Concentration (ng/mL) and Signal. Create an XY scatter (markers only) with a logarithmic x-axis base 10 from 0.1 to 1000, y-axis 0 to 2.5, a logarithmic trendline with equation and R² displayed, and axis titles with units. Save.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clicktype· Text & forms- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Chart XML: scatterChart style marker, x valAx logBase 10 min 0.1 max 1000, y min 0 max 2.5, trendline trendlineType='log' dispEq=1 dispRSqr=1; axis titles contain units; Excel-authored c16 ids required. VBA, COM and AppleScript automation disabled, no Python in the VM.
Proposed partial credit
25% scatter with correct ranges; 30% log axis and bounds; 30% trendline type + eq + R²; 15% axis titles.
R04.03.01.03-T06 Label a bubble chart's points with customer names from cells
Original instruction
In fixtures/account_matrix.xlsx create a bubble chart from 'Accounts' (x = Revenue, y = Growth %, size = Seats) and label every bubble with the customer name from column A using data labels 'Value From Cells' (no X/Y values in labels). Scale bubble size to 60%. Save.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clickhover· Hover & context- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
bubbleChart with xVal/yVal/bubbleSize refs exact; bubbleScale 60; dLbls use c15:datalabelsRange referencing Accounts!A2:A26 with showDataLabelsRange=1 and showVal/showCatName off. openpyxl cannot author datalabelsRange. VBA, COM and AppleScript automation disabled, no Python in the VM.
Proposed partial credit
35% bubble chart refs; 40% value-from-cells labels; 15% other label parts off; 10% bubble scale.
R04.03.01.03-T07 Turn a task list into a Gantt chart with stacked bars
Original instruction
fixtures/rollout_plan.xlsx 'Tasks' has Task, Start, Duration (days) for 18 tasks. Build a Gantt chart: stacked bar with Start as an invisible series, Duration filled #4472C4, tasks top-to-bottom in list order, date axis from 2026-01-05 to 2026-06-29 with major unit 14 days. Save.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clicktype· Text & forms- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
barChart barDir=bar grouping=stacked; series 1 noFill; series 2 fill #4472C4; category axis orientation maxMin; value axis min/max equal the date serials and majorUnit 14; rendered PNG compared to reference (SSIM ≥0.9). VBA, COM and AppleScript automation disabled, no Python in the VM; file must be saved by Excel (app.xml).
Proposed partial credit
30% stacked bar with invisible start; 25% order reversed; 30% date bounds and unit; 15% fill color.
R04.03.01.03-T08 Add asymmetric custom error bars from confidence-interval columns
Original instruction
fixtures/trial_means.xlsx 'Results' has Arm, Mean, CI_low, CI_high for 5 arms. Create a column chart of Mean with custom Y error bars whose plus/minus values come from helper columns you compute (CI_high−Mean, Mean−CI_low), end caps on, error bar line 1.5 pt black. Save.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clicktype· Text & forms- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
errBars errValType='cust' with plus/minus numRef pointing to helper ranges whose values equal key; noEndCap=0; line width 19050 EMU, black; Excel-authored c16 ids present. VBA, COM and AppleScript automation disabled, no Python in the VM.
Proposed partial credit
25% helper columns correct; 45% custom error bars wired to them; 30% caps and styling.
R04.03.01.03-T09 Make an interactive Numbers chart with a year slider
Original instruction
In fixtures/fleet.numbers, build an interactive column chart from table 'Fuel by Year' so each year 2019–2025 is a data set selected with a slider; categories are the 6 vehicle classes, y-axis fixed 0–120,000 L across data sets, value labels on. Save.
- Platforms
mac· macOS- Applications
- Numbers
- Interfaces
desk· Desktop GUI- GUI need · task setting
G3· GUI-only- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clickdrag· Drag & drop- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Grader-side AppleScript export to PDF for each slider position is not possible, so grader opens the file in Numbers via its own UI-automation script, steps the slider and captures 7 renders; each compared to reference (SSIM ≥0.9); chart type checked in exported .xlsx. osascript is blocked for the agent; the file must be saved by Numbers.
Proposed partial credit
30% interactive chart type; 30% data sets per year; 25% fixed axis; 15% labels.
R04.03.01.03-T10 Chart the last 12 months with a self-extending named range
Original instruction
fixtures/kpi_feed.xlsx 'Feed' grows by one row per month. Define names using INDEX/COUNTA (no OFFSET volatile functions) that cover the latest 12 rows, and build a line chart whose series and categories use those names, so new rows shift the window. Save.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
type· Text & formsclick· Point & click- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 4 / 5
Proposed verifier
Defined names use INDEX/COUNTA, no OFFSET; chart series refer to the names; grader appends 3 rows via openpyxl, re-renders via Excel COM and checks the plotted category labels are the latest 12 months each time. VBA, COM and AppleScript automation disabled, no Python in the VM; file must be saved by Excel (app.xml).
Proposed partial credit
30% names correct and non-volatile; 30% chart bound to names; 40% window shifts correctly after appends.
R04.03.01.03-T11 Recreate a published chart exactly from a reference image
Original instruction
input/target_chart.png is a chart from last year's board pack; the data is in fixtures/board_data.xlsx. Recreate it in Excel: chart type, series order, colors, axis bounds and units, number formats, label placement, legend, gridlines and title text. Place it on 'Board' at B2, size 16×9 cm.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clickvis· Visual judgmentgeom· Precise geometry- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 5 / 5
Proposed verifier
Spec diff against a hidden reference chart XML (type, series, colors ΔE2000 ≤3, axes, formats, label positions) plus Excel COM render of 'Board' compared to the target (SSIM ≥0.92); anchor and extent within 0.1 cm. VBA, COM and AppleScript automation disabled, no Python in the VM; file must be saved by Excel (app.xml).
Proposed partial credit
30% type and series; 25% axes and formats; 20% colors and labels; 25% render SSIM and size.
R04.03.01.03-T12 Compare carrier delivery times with a box-and-whisker chart
Original instruction
fixtures/deliveries.xlsx 'Times' has Carrier and DeliveryHours (6,000 rows). Insert an Excel Box & Whisker chart grouped by carrier with outlier points shown, mean markers and mean line off, inner points off, quartile calculation set to exclusive median, y-axis 0–120 h. Save.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & click- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 4 / 5
Proposed verifier
chartex layoutId='boxWhisker' with statistics quartileMethod='exclusive', visibility meanMarker=0 meanLine=0 nonoutliers=0 outliers=1; axis bounds; data refs exact. chartex cannot be authored with openpyxl; VBA, COM and AppleScript automation disabled, no Python in the VM.
Proposed partial credit
35% box chart by carrier; 35% visibility options; 15% quartile method; 15% axis bounds.
R04.03.01.03-T13 Lay out six regional line charts as aligned small multiples
Original instruction
fixtures/regional_sales.xlsx: on sheet 'Dashboard' build 6 line charts (one per region) with identical formatting, y-axis fixed 0–500, no legends, titles = region names. Arrange them in a 3×2 grid starting at B2, each exactly 8×5 cm, 0.5 cm gaps, edges aligned.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
geom· Precise geometrydrag· Drag & dropclick· Point & click- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 4 / 5
Proposed verifier
Drawing XML anchors: 6 charts, ext 8×5 cm (±0.05), grid positions and gaps within 0.05 cm; each chart's series refs the correct region; axis min/max 0/500; legend absent; titles exact. Excel-authored c16 ids required. VBA, COM and AppleScript automation disabled, no Python in the VM.
Proposed partial credit
30% 6 charts with right data; 20% identical axes/format; 35% size and grid alignment; 15% titles/no legends.
R04.03.01.03-T14 Add trend and win/loss sparklines to a KPI table
Original instruction
In fixtures/kpi_table.xlsx 'KPIs', add line sparklines in H2:H15 from B:G (monthly values) with high and low points marked red/green, and win/loss sparklines in I2:I15 from the variance columns J:O, with negative points #C00000 and an axis. Save.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clickdrag· Drag & drop- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
x14:sparklineGroups: one line group H2:H15 → B:G with high/low markers and colors, one stem (win/loss) group I2:I15 → J:O with negative color and displayXAxis=1; exact ranges per row. openpyxl does not write sparklines; VBA, COM and AppleScript automation disabled, no Python in the VM.
Proposed partial credit
40% line sparklines with correct ranges; 20% high/low markers; 40% win/loss group with colors and axis.
R04.03.01.03-T15 Repair eight broken charts after a workbook restructure
Original instruction
fixtures/monthly_pack.xlsx had rows inserted and a sheet renamed; 8 charts on 'Pack' now plot wrong ranges, two have swapped axes, three carry stale hard-typed titles. Fix each chart per input/chart_spec.md (ranges, series names linked to cells, axes, titles linked to cells) without recreating them.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clickhover· Hover & contextvis· Visual judgment- Work product
WP03· Visual layout- Conditions
hidden· Hidden/stale state- Proposed difficulty
- 5 / 5
Proposed verifier
Per-chart XML diff vs spec: series refs, tx strRef titles, axis assignment; chart object ids unchanged (fixed, not recreated); render of each chart vs reference SSIM ≥0.9. Excel-authored edits required (c16 ids preserved). VBA, COM and AppleScript automation disabled, no Python in the VM.
Proposed partial credit
48% series refs; 28% titles; 24% axis assignment; each pro-rata over the 8 charts.
R04.03.01.03-T16 Build a Pareto chart with cumulative line and 80% reference
Original instruction
From fixtures/defects.xlsx 'Log' (defect codes), build on 'Pareto' a combo chart: defect counts sorted descending as columns, cumulative % as a line on a secondary axis 0–100%, and a dashed horizontal 80% reference line as an extra series. Add a text label 'Vital few' above the bars that make up 80%. Save.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clickgeom· Precise geometrytype· Text & forms- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 5 / 5
Proposed verifier
Helper table counts and cumulative % equal key; chart: bar + 2 line series, cumulative on secondary axis 0–1, reference series constant 0.8 with prstDash dash; text box positioned over the vital-few category span (render bbox check). VBA, COM and AppleScript automation disabled, no Python in the VM; file must be saved by Excel (app.xml).
Proposed partial credit
25% helper table; 30% combo with secondary axis; 25% dashed 0.8 reference line; 20% label placement.
R04.03.01.03-T17 Show a budget hierarchy as a treemap with banner labels
Original instruction
fixtures/opex_budget.xlsx 'Budget' has Division, Department, Amount. Insert an Excel treemap with Division as level 1 and Department as level 2, banner parent labels, fill colors per division from input/division_colors.txt, data labels showing category name and value ($k). Save.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & click- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
chartex layoutId='treemap' with parentLabelLayout='banner', 2-level strDim, value refs exact; per-division data point fills match the color file; labels show name and value. VBA, COM and AppleScript automation disabled, no Python in the VM; openpyxl cannot write chartex.
Proposed partial credit
35% treemap with two levels; 20% banner labels; 30% division colors; 15% label content.
R04.03.01.03-T18 Control pair of T03: create the combo chart via the Sheets API
Original instruction
Control pair of T03. The same quarterly data is in Google Sheet 'Quarterly PL'. Using the Sheets API (credentials provided), add a combo chart with Revenue columns, Margin % line on the right axis fixed 0–0.4, and data labels on the line only.
- Platforms
web· Any OS (web)- Applications
- Google Sheets · Sheets API
- Interfaces
api· API/SDK- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
AP· API-sufficient- Interaction primitives
code· Code & commands- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 2 / 5
Proposed verifier
spreadsheets.get: one chart with basicChart COMBO, series types COLUMN and LINE, line on RIGHT_AXIS with viewWindow 0–0.4, dataLabel on the line only. Compared with T03 to measure GUI vs API success.
Proposed partial credit
35% combo chart; 35% right axis with bounds; 30% labels on line only.
R04.03.01.03-T19 Build a candlestick stock chart with a volume panel in Calc
Original instruction
In fixtures/ticker_daily.ods (synthetic prices: Date, Open, High, Low, Close, Volume), create a LibreOffice Calc stock chart of type 'Volume–Open–Low–High–Close' for the last 60 trading days, rising candles white and falling black, date axis with weekly major ticks. Save.
- Platforms
mac· macOSwin· Windowslin· Linux- Applications
- LibreOffice Calc
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clicktype· Text & forms- Work product
WP03· Visual layout- Conditions
- None declared
- Proposed difficulty
- 4 / 5
Proposed verifier
Chart object in content.xml: stock chart with volume (chart:stock-with-volume), 5 ranges covering exactly the last 60 rows, candle colors, date axis interval 7 days; render vs reference SSIM ≥0.9. Macros/UNO disabled for the agent.
Proposed partial credit
30% chart type; 30% correct 60-row ranges; 20% candle colors; 20% axis ticks.
R04.03.01.03-T20 Save a brand chart template and apply it to twelve charts
Original instruction
Format the first chart in fixtures/annual_review.xlsx to input/brand_charts.pdf (Arial 10 pt, palette of 6 hex colors in order, light grey #D9D9D9 gridlines, no chart border), save it as output/brand.crtx, then apply that template to the other 11 charts without changing their data or chart types.
- Platforms
mac· macOSwin· Windows- Applications
- Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
click· Point & clicknav· Pan/zoom/scroll- Work product
WP03· Visual layout- Conditions
long· Long-horizon- Proposed difficulty
- 5 / 5
Proposed verifier
output/brand.crtx exists and its chart XML carries the spec; all 12 charts' XML: latin typeface Arial 1000, series fills follow the palette order, gridline color D9D9D9, no chart-area border; series refs and chart types unchanged vs input. VBA, COM and AppleScript automation disabled, no Python in the VM; file must be saved by Excel (app.xml).
Proposed partial credit
25% template file correct; 50% charts conform (pro-rata over 12); 25% data/types unchanged.
Mapped research references
Existing benchmark-to-class links from the research catalog. They are not record-level annotations, executions or product endorsements.