Usewright
Back to results
Explore the library298 guides total · goals, work & tasks
By goal

Choose what you want to achieve.

By work or interest

Choose a setting you work or spend time in.

By task

Choose the kind of help you want.

Counts show all available guides. Filters combine to narrow your results.

Communicate clearlyGuide

Make an honest chart from a small dataset

Compare weekly counts without turning a missing value into zero, exaggerating the scale or inventing a cause.

Illustrative resourceNot tool-testedAnyone explaining a small set of comparable counts in a report or presentation
What you’re working towards

A readable eight-week bar chart, source CSV and complete text alternative that keep a missing week distinct from the recorded zero.

Download playbook .md

Your choice also saves this guide’s current worked-example checkmarks. Only this browser; no account or sync. View saved guides

Saving needs JavaScript. Reading and downloading this guide does not save it.

Included example · illustrative SVG file

Download the example chart

Original self-contained 1200 × 700 SVG with a zero baseline, visible values and a labeled missing week. Static reference output.

Not tool-tested. Review and adapt before using it for a real project.

View example file (opens in a new tab; your browser may download the file) Download example
Bar chart of synthetic weekly completed-item counts: 3 August 12, 10 August 18, 17 August 0, 24 August missing, 31 August 21, 7 September 16, 14 September 24 and 21 September 19. All dates are 2026 week starts; the count axis runs from zero to 30.
The recorded zero and the missing observation remain visibly different. Full values are also available as a text download.
A short visual explanation

See why zero and missing stay different

An 18-second silent explanation using the exact owned synthetic dataset. A recorded zero has a known value; a missing observation has no known bar height. This is authored teaching media, not footage of an AI product or a real business.

Download the silent clip · Download full transcript

Text alternative

A source table sits beside a bar chart titled Items completed each week. All eight date positions remain visible throughout. The vertical count axis runs from 0 to 30, with ticks every 5. The horizontal axis is Week starting (2026). Every week starts on a Monday.

Week starting 3 August 2026: 12 items completed; observed.
Week starting 10 August 2026: 18 items completed; observed.
Week starting 17 August 2026: 0 items completed; observed.
Week starting 24 August 2026: count not supplied; missing.
Week starting 31 August 2026: 21 items completed; observed.
Week starting 7 September 2026: 16 items completed; observed.
Week starting 14 September 2026: 24 items completed; observed.
Week starting 21 September 2026: 19 items completed; observed.

The 17 August zero is explicitly labeled at the baseline. Its zero-height baseline marker is not a positive-height bar. The 24 August count cell is empty and its status is missing; the corresponding chart position has no bar and is labeled Missing value. The source does not provide a reason for the missing count.

All counts measure completed items during that week in one fictional process, not running totals. The data supplies no cause, intervention, target, comparison group, or forecast. No full-period total or average is claimed.

The video, poster, source CSV, and text equivalent are separate static files. Rebuild and recheck them together if the data changes. The poster provides the final teaching state without motion.

The encoded clip, source data, poster and text equivalent were checked together. Video inspection is separate from consumer-AI workflow evidence.

Source files to work with

Make a start

One step at a time.

4 steps

Start with the synthetic worked example or other non-sensitive inputs. Run one prompt at a time in your chosen AI tool. Replace any bracketed placeholders with the requested material, check each reply, and supply the source or earlier outputs that step needs. Paste them again if the tool lacks that context.

Set up the worked example

Open

Open a new text-AI chat, a plain-text editor and an image or SVG viewer. A charting tool is needed to render or export a new chart; no particular app or free plan is assumed. If your AI returns SVG text, you can save and open it locally. If you cannot create or view an image, finish the data/specification draft and leave visual checks pending.

Add the sample

Download weekly-counts.csv and weekly-counts-input.md. In step 1 replace [paste weekly-counts.csv and the input brief] with the complete text of weekly-counts-input.md, which already includes the CSV, definitions and brief; it matches ‘The starting point’ (Synthetic input). In step 2 replace [paste] with that full input and the checked step 1 findings. In step 3 supply the same input plus your checked chart specification. Before step 4 save and reopen your new chart and text alternative. Paste the original CSV and reopened text under their headings; attach the saved image only if the chat can inspect images. Otherwise say ‘I cannot attach an image; review the supplied source only’ and perform the visual checklist yourself. Use ‘The intended result’ (Illustrative output) and reference downloads to compare after trying.

Start with the synthetic example, then adapt each prompt to your task.

01

Check what each row means

Read the dates, units and observation status before choosing a chart. There are eight week positions, seven known counts and one genuine zero.

Your starting prompt
Review this CSV and its data notes before charting. State what one row measures, the units, the week convention and the period. Identify duplicate or out-of-order dates, missing rows, invalid counts and missing values. Distinguish a recorded 0 from a blank with observation_status missing. Do not fill, interpolate, drop or silently convert missing values. For this fixture, retain all eight weeks and mark 24 August unknown.

DATA AND NOTES:
[paste weekly-counts.csv and the input brief]
Expect

A data check identifying eight consecutive Monday-start weeks in 2026, seven observed counts and one missing count. Each number measures items completed during that week.

Check

The ordered values are 12, 18, 0, missing, 21, 16, 24 and 19. 17 August is an observed zero; 24 August is unknown. Keep the blank CSV field and its missing status together.

If it differs

If a blank becomes zero or a date disappears, point to the original row and ask for correction. Do not estimate the missing count or add a reason for it.

02

Specify a chart the data can support

Use chronological bars for this exercise, with a count scale beginning at zero. Give the missing week a labeled gap instead of a fabricated bar.

Your starting prompt
Propose one bar-chart specification for these weekly item counts. Preserve the eight chronological positions, including the missing week. Use a count axis starting at 0 with an upper bound of 30 and ticks every 5 for this fixture; label every known value. Give the missing week an explicit Missing value label and no bar. Retain the observed zero with its 0 label. Specify a descriptive title, year, Monday-start week convention, units and synthetic source note. Add no 3D effect, trend line, forecast, causal headline or unsupported comparison.

CHECKED DATA:
[paste]
Expect

A chronological bar-chart specification with all eight positions, a 0–30 count scale in steps of 5, direct value labels and a labeled gap for 24 August.

Check

Dates/year, Monday-start convention, count units and synthetic source are stated. Zero has a 0 label; missing has no value or bar. There is no trend line, cause or forecast.

If it differs

If the proposed scale starts above zero, sorting changes chronology or missing data gets a bar, request the specified correction before making the image.

03

Build the chart and its text equivalent

Ask your chosen charting tool to implement the specification. Keep full values available as text alongside the picture. Download the reference SVG or PNG to compare after trying.

Your starting prompt
Create a static chart from the checked data and specification below. If generating SVG, return self-contained SVG with a title and description, no scripts, links, embedded images, external fonts or network resources. Keep all labels readable and inside the image. Return a short image description plus a separate full text alternative with dates, every known value, the missing week, axes and the source note. Do not claim the chart was rendered or viewed unless you actually opened the result. State whether the image updates automatically when the CSV changes.

DATA AND SPECIFICATION:
[paste]
Expect

A static chart or complete self-contained SVG source, plus a short description and full text alternative. Save new outputs as my-weekly-chart.svg or the actual exported image type, my-weekly-data.csv and my-weekly-description.txt. Preserve the reference downloads.

Check

For SVG, save only <svg through </svg> in plain UTF-8 text, excluding code-fence lines; it must have no script or external resources. The text lists all eight dates/values, axes and the missing-data note. Changing a CSV does not automatically update these static files.

If it differs

If it provides only a specification, do not call that an image: ask for complete SVG source or use an available charting tool. If rendering/export is unavailable, mark it pending. Save and reopen the actual files before the audit.

04

Open the image and audit every mark

A plausible-looking image can still plot the wrong values. Compare all eight positions with the CSV and check that the text alternative tells the same story.

Your starting prompt
Audit the final rendered chart against the original CSV and text alternative. Check chronological order, every direct value label, bar height, baseline, tick spacing, year and units. Verify that 17 August is a real 0 and 24 August has no value or bar and is labeled missing. Confirm that all eight dates are retained, no total or average treats the missing count as zero, and no causal claim was added. Check readable text and clipping at the intended display size. List actual visual observations separately from source-code checks and unperformed browser or assistive-technology tests.

ORIGINAL CSV:
[paste]

CHART:
[attach or open final image]

TEXT ALTERNATIVE:
[paste]
Expect

An audit of the reopened chart, original CSV and saved text alternative, separating actual image observations from source checks.

Check

Open the saved image yourself at its intended display size. Compare all labels and bar heights with the CSV: 0–30, evenly spaced ticks of 5, 17 August labeled 0 and 24 August explicitly missing. The seven observed values sum to 110; the full eight-week total is unknown.

If it differs

If the image cannot be seen, no visual pass is possible. For a wrong bar or clipped label, identify it, correct the chart and text together, save new revisions and reopen them to repeat the checks. Keep browser/assistive-technology tests unperformed unless actually carried out.

Optional: improve the result

Review revisions without rewriting the story

If a missing count is later supplied, record the new evidence and rebuild the image and description together.

Advanced prompt
Compare this revised source with the previous dataset, chart and description. Record exactly which dates, values or statuses changed and what supports them. If an unknown count is now supplied, recompute only the affected summaries, rebuild the plot and update the text alternative. Preserve the baseline convention and show any necessary upper-bound change explicitly. Check every plotted mark and statement again. Do not invent a reason for the changes, infer causality or describe the old missing value as if it was known earlier.

PREVIOUS SOURCE AND OUTPUT:
[paste]

REVISED SOURCE:
[paste]
Optional: use your own input

Chart your own small count dataset

Use a small non-sensitive dataset you own or may use, with comparable nonnegative counts and a clear meaning for each period. Open a fresh chat and fill the prompt below. After checking the data review, send these messages one at a time:
“Specify one chronological bar chart for my checked counts. Retain every expected time position and distinguish observed zero from missing. Start the count axis at zero and propose an upper bound/tick spacing that fits my values. Label known values, missing observations, dates/year, units and source. No invented causes, trend line, forecast or filled gaps. Wait for my review.”
“Create the checked chart as a static image if your tools can, or return complete self-contained SVG with title/description and no scripts, links, external fonts, images or network resources. Also return a short description and complete text alternative listing every date/value or missing status, axes and limitations. State what was actually rendered/viewed and whether files update automatically. CHECKED DATA AND SPECIFICATION: [paste both].”
Save the chart as my-count-chart.svg or its actual exported format, the data as my-count-data.csv and the full description as my-count-description.txt, using fresh filenames. If saving SVG text, include only the complete SVG source. Reopen these saved files and view the chart before the next message; if no viewer is available, mark that check pending.
“Audit the original data, saved chart and saved description below. Recompute every plotted value, check chronology, labels, common scale and zero baseline, missing-versus-zero treatment, readable text and clipping. Separate source inspection, actual image observations and tests not performed. ORIGINAL DATA AND NOTES: [paste]. SAVED DESCRIPTION: [paste]. SAVED CHART: [attach only if image inspection is supported, otherwise paste SVG source and state that visual checks are unperformed]. MY VIEWING OBSERVATIONS: [paste only what you actually saw or say unavailable].”
Check the picture yourself against every row. Correct, save and reopen all affected files together; do not audit an unsaved chat revision as the final image. A text-only review cannot replace visual inspection. Your data controls the date range, scale and totals; the sample answers do not apply. The original advanced prompt needs your own prior files and supported revised data before reuse. This is a private static draft, with no publishing or account changes.

Your own-input prompt
Help me make an honest static bar chart from MY DATA. First check the definition and unit of each count, time-period convention, expected dates, ordering, duplicates, missing rows and invalid values. Require comparable nonnegative counts; ask about conflicting definitions or unknown coverage rather than combining them. Keep recorded zeros separate from missing observations and never fill, interpolate or drop missing positions. Identify what the data can and cannot support; stop after the data check for my review. Later we will specify, create, save and inspect a chart and full text alternative. Do not publish or claim to view an image you cannot access.

MY DATA
[Paste the complete table/CSV with period/date, count and observed/missing status. Keep unknown counts blank and explicitly marked missing.]

DEFINITIONS AND CHART PURPOSE
[State count definition, units, expected periods/date range, period convention, reporting coverage, source, audience/question and intended display size. Mark any unknowns.]

Check your own result

  • Every expected period remains visible and every label matches your data; missing observations never become zero.
  • The count scale starts at zero and uses consistent spacing; your values determine the upper bound.
  • Any subtotal names exactly which observations it covers. Unknown periods do not become complete totals, averages or explanations.
  • The actual saved picture and full text agree with the data. Unperformed visual, browser and assistive-technology checks remain explicit.
Show, then try

See the shape of the result.

This input is synthetic. The output is a written illustration of the intended result, not a verified AI-tool run.

The starting point

Synthetic input
ORIGINAL SYNTHETIC DATA
weekly-counts.csv contains eight consecutive Monday-start weeks. Each observed count is the number of items completed during that week in one fictional process. All observed counts have the same definition and reporting coverage. No real organization or person is represented.

week_start,items_completed,observation_status
2026-08-03,12,observed
2026-08-10,18,observed
2026-08-17,0,observed
2026-08-24,,missing
2026-08-31,21,observed
2026-09-07,16,observed
2026-09-14,24,observed
2026-09-21,19,observed

MEANING
2026-08-17 is a recorded zero. The count for 2026-08-24 was not supplied; the reason is unknown. A blank count with status missing is not zero and must not be filled, interpolated or dropped from the timeline. Counts are weekly flows, not running totals. The data supplies no causes, intervention, target, comparison group or forecast.

BRIEF
Make one chronological chart for a general reader who wants to compare the eight weeks. Use all eight week positions, a count axis beginning at zero, clear dates/year and units, and an explicit missing-value note. Keep actual values and the missing week in a text alternative alongside the chart. The supplied SVG and PNG are authored reference outputs; they are not evidence of an executed AI-charting workflow.

Copies text only. For image or file tasks, use the source files named in the steps too.

Download source .txt

The intended result

Illustrative output
ILLUSTRATIVE CHART HANDOFF
The reference chart is a static bar chart with eight chronological week positions and seven observed values. The vertical count axis runs from 0 to 30 in steps of 5. Values are labeled 12, 18, 0, missing, 21, 16, 24 and 19 in order. The 24 August position is retained and explicitly marked Missing value without a bar. The observed zero on 17 August is labeled 0.

The title is Items completed each week. The horizontal label is Week starting (2026); the vertical quantity is Items completed (count). The subtitle identifies Monday-start weeks and the 3 August–21 September 2026 range. The source is original synthetic teaching data.

SHORT TEXT ALTERNATIVE
Weekly counts of completed items from 3 August to 21 September 2026, with one missing week. Full values and the missing-data note follow.

FULL VALUES
3 Aug: 12; 10 Aug: 18; 17 Aug: 0 (observed); 24 Aug: missing; 31 Aug: 21; 7 Sep: 16; 14 Sep: 24; 21 Sep: 19. All dates are 2026 week starts. The accompanying weekly-counts-description.md gives the full reading order, axes, values and missing-data note in ordinary text.

WHAT THE DATA SUPPORTS
The largest observed count is 24 on 14 September; the smallest is the recorded zero on 17 August. The seven observed weeks sum to 110 items. The complete eight-week total is unknown. No causal explanation, full-period average or forecast is supplied.

HANDOFF LIMITS
The SVG and PNG were authored and visually inspected as a reference fixture; no consumer AI-charting workflow, spreadsheet import, website embed or assistive-technology behavior was tested. The data, graphic and text alternative are separate static files. Rebuild and recheck all three if the data changes.
Download result .txt
You’re the final check

Check the worked example.

These checks use the worked example and its particular facts. With your own material, check your actual names, numbers, sources and rules; do not reuse the sample’s values as the answer.

Unsaved checks stay in this tab and reset when you reload. If you choose Save guide, its current example checkmarks are kept in this browser when storage is available.

A few things to watch for.

  • Allowing a blank CSV cell to become a zero bar or silently disappearing week.
  • Using a truncated baseline on a bar chart, which changes the apparent proportions.
  • Sorting weeks by count when the question is about time.
  • Describing an upward bar as evidence that an unmeasured action caused improvement.
  • Updating the CSV while leaving the static chart or text alternative on the old values.

What has been checked?

Documentation
Reviewed 2026-10-14. Source review is not workflow execution.
Included files
Scoped checks passed on 2026-10-14. File parsing; Fixture checks. See scope below.
AI workflow
Not run in a recorded consumer-AI workflow.
Time savings
Time savings not measured
Read the scope and evidence

weekly-counts.csv

2026-10-14 · parser · passed

Plain UTF-8 CSV with stable columns, eight rows, quoted-field round-trip and no formula-leading/control-character cells.

  • Exact three-column header and row widths
  • Eight data rows
  • CSV parse/write/parse retains every cell

Environment: Local Python standard library; CSV, XML and PNG parsers plus independent exact fixture/geometry assertions. No consumer AI or spreadsheet app.

SHA-256: 6a8d10d8d517f32ac025213fac5e1702de5eddef3da7b0971c5453beed99c2ac

weekly-counts.csv

2026-10-14 · semantic fixture · passed

All eight dated observations match independent controls. Missing and zero are distinct; Monday spacing and the seven-observation subtotal are verified.

  • Eight consecutive Monday-start weeks
  • 17 August is observed 0; 24 August is blank and missing
  • Seven known counts sum to 110, not a complete eight-week total
  • Six invalid source mutations rejected

Environment: Local Python standard library; CSV, XML and PNG parsers plus independent exact fixture/geometry assertions. No consumer AI or spreadsheet app.

SHA-256: 6a8d10d8d517f32ac025213fac5e1702de5eddef3da7b0971c5453beed99c2ac

weekly-counts-chart.svg

2026-10-14 · parser · passed

Self-contained SVG parses with expected intrinsic dimensions and no active or externally linked resources.

  • XML parse; 1200 by 700 pixels
  • No scripts, foreignObject, event handlers or href references

Environment: Local Python standard library; CSV, XML and PNG parsers plus independent exact fixture/geometry assertions. No consumer AI or spreadsheet app.

SHA-256: bbe1c73175ef86a66c156875ebc84fb0a62e6fcab2b26268f8266190b1ff063e

weekly-counts-chart.svg

2026-10-14 · semantic fixture · passed

Every plotted group, visible number and bar geometry is checked against independent data controls using a 10-pixel-per-item zero-baseline scale. This source/geometry check does not alone establish rendered readability.

  • Eight chronological groups; seven observed bars including zero height
  • Missing week has no value attribute or bar
  • All bar heights, y positions, widths and x positions match controls
  • Tick labels 0 through 30 at steps of 5
  • Four invalid plotted/label mutations rejected

Environment: Local Python standard library; CSV, XML and PNG parsers plus independent exact fixture/geometry assertions. No consumer AI or spreadsheet app.

SHA-256: bbe1c73175ef86a66c156875ebc84fb0a62e6fcab2b26268f8266190b1ff063e

weekly-counts-chart.png

2026-10-14 · parser · passed

Raster signature, intrinsic dimensions, chunk CRCs and decompression checked.

  • PNG is 1200 by 700
  • Every chunk CRC matches
  • Compressed image data decodes

Environment: Local Python standard library; CSV, XML and PNG parsers plus independent exact fixture/geometry assertions. No consumer AI or spreadsheet app.

SHA-256: 08e54b48dc524e93ef0444d01446b295cad6b6ef352e4c31db88d69c4a224324

weekly-counts-chart.png

2026-10-14 · semantic fixture · passed

Actual rendered PNG viewed on 2026-10-14: all date and value labels, the zero baseline, observed zero and missing-week annotation were legible without clipping. This is an editorial visual review, not a consumer AI/charting or browser/assistive-technology test.

  • All eight date positions visible
  • 0 on 17 August visibly differs from Missing value on 24 August
  • Bar lengths and direct labels visually agree with source
  • Title, count unit, year, source and missing-data note visible

Environment: Inkscape 1.4 raster export, followed by direct view_image inspection of the actual 1200 by 700 PNG.

SHA-256: 08e54b48dc524e93ef0444d01446b295cad6b6ef352e4c31db88d69c4a224324

weekly-counts-description.md

2026-10-14 · semantic fixture · passed

The authored text alternative was compared with all eight source observations, scale labels and the limits on totals and explanation.

  • All eight dates and values or missing status present
  • Observed zero retained
  • 110 identified only as a seven-observation subtotal
  • No complete eight-week total or causal explanation

Environment: Local Python standard library; CSV, XML and PNG parsers plus independent exact fixture/geometry assertions. No consumer AI or spreadsheet app.

SHA-256: ee49b13d7fbb60ebf4286957f4adb8c385fc5a9b5d9c3b149493ec33a4d7cb48

Government Analysis Function: data visualisation charts ↗ (opens in a new tab)

2026-10-14 · confirmed

For bar charts, the numerical axis should start at zero; a time-series discontinuity must be highlighted, and a line should not join across a missing point.

Limits: This is a scoped documentation review, not evidence of an AI run, browser/assistive-technology test or whole-page accessibility compliance.

W3C WAI: Complex Images ↗ (opens in a new tab)

2026-10-14 · confirmed

A chart can need a short identifying description and a fuller text equivalent containing the essential values and relationships.

Limits: This is a scoped documentation review, not evidence of an AI run, browser/assistive-technology test or whole-page accessibility compliance.

For context and further reading

Official documentation

Documentation reviewed: 2026-10-14. Documentation review is not tool-execution evidence. This resource remains illustrative and not tool-tested.

Keep going
Back to the library
  1. Simplify a complex taskGuide

    Find themes in customer feedback without inventing consensus

    A downloadable theme ledger that preserves source IDs and exact excerpts, counts unique codable comments transparently and keeps excluded or unclear rows visible.

  2. Create work & incomePlaybook

    Write a portfolio case study without invented metrics

    A complete third-person synthetic case study that distinguishes role, decisions, observations, revisions, delivery status and unmeasured outcomes.

  3. Communicate clearlyGuide

    Write image descriptions that fit their actual purpose

    A three-image handoff with context-specific roles, exact proposed alternatives, preserved notice text and clear implementation checks.

Copy these instructions

Your browser blocked automatic copying. Select the text below and copy it.