GPTsApp SkillsCodex Reset
Menu
Codex Reset
← Back to results

plugin directory · Use or study

testing-data-visualizations

Choose the smallest reliable test mix for a chart, dashboard or interactive visualization without turning every pixel into a brittle contract.

openai/pluginsplugins/build-web-data-visualization/skills/testing-data-visualizations/SKILL.md

Read SKILL.md ↗Open source directory ↗

Before you use it

Input
The analytical and interaction risks, chart technology, data transformations, user workflows, viewports, degraded states and export behavior that matter to the product.
Expected output
A layered plan assigning data logic, component contracts, selected visual states and integrated workflows to unit, component, visual or end-to-end tests with deterministic fixtures.
Prerequisites
Stabilize viewport, data, fonts, theme, locale, timezone and animation before visual assertions. Mock at network or loader boundaries while keeping transforms and rendering real. The source's broad WebGL, mobile and story scenarios are planning examples, not capabilities supplied or tested by this directory.

Client setup and runtime behavior are not verified here; client installation is not individually established. The use notes below describe source preparation, inputs and limits.

Publisher documentation, static file observations and environment tests are different kinds of evidence. Runtime behavior has not been observed.

Use this method

Source-based preparation, not an installation or a completed run. Nothing is executed here.

Good fit for
Designing or repairing tests where the main object is a chart, dashboard, diagram or visualization workflow rather than a generic page.
Not for / limits
Not a mandate for full-page screenshot suites or testing third-party rendering internals. Live services, every browser, WebGL hardware and the many adjacent reference files were not exercised; visual baselines remain useful only when deterministic and intentionally reviewed.
Prepare
List the failures that would make the visualization analytically wrong or unusable, including transforms, interactions, mobile layout, empty or stale data and export behavior.
Apply
Place each risk at the cheapest layer that proves it clearly, mock only at a stable boundary, and prepare canonical, edge and stress fixtures with deterministic rendering conditions.
Check
Run one golden integrated flow plus the highest-risk failure states, inspect any visual diff rather than auto-accepting it, and remove duplicate or flaky assertions that protect no public contract.
Sources read for this note

These exact files were read as text at the displayed commit. Other files and external tools were not reviewed. A repository LICENSE observation does not confirm Skill-specific applicability or dependency rights.

Study the method

Compare the four test layers, fixture classes and explicit non-goals. The local references explain why one golden flow and a few high-risk failures are usually more trustworthy than duplicated snapshots at every layer.

8 files in the observed scope; 0 in scripts/, 6 in references/, 0 in assets/.

This is a sample of observed paths, not the full directory. Zero files in scripts/ does not mean no execution, network access or file writes. Markdown and other folders may describe or contain executable behavior.

Source version and requirements

Observed
2026-09-07T14:02:40.375Z
Source commit
1e285826e604f66f7208f7ac4dba0fe8341d1f57
Content SHA-256
fba1423c24dac1f401c76d85cbc39d7b354bd2fb18cc9055d62798e6cca6d28e
License
Unknown
Author attribution
Repository owner: openai. Individual authorship is not independently established.

This page describes a pinned source snapshot, not the latest repository state. Changes after this commit have not been reconciled here. A repository license does not necessarily cover external models, datasets or services.

Open current repository ↗
Reference note and manual copy