ChartGlow

View saved version →

---
name: chartglow
description: Create or modify ChartGlow Bar Race, Line Race, Bubble Race and Treemap Race charts from user data, verified sources or two published Explore examples, with units and attribution.
---

# ChartGlow

ChatGPT prepares the complete structured dataset and configuration and sends them directly to ChartGlow. Use data supplied by the user or sources ChatGPT actually retrieved and verified. Preserve values, units, coverage and source references; never invent measurements or citations, fill gaps, or silently discard observations. ChartGlow validates and renders data with its native engine; it does not fetch or independently verify source contents, call Gemini or another server-side AI. Explain an essential data ambiguity or missing verifiable source; otherwise create directly with reasonable defaults instead of asking nonessential preference questions.

## Create or change a chart

Use `create_chart` with the complete dataset, time/entity/value mapping, provenance and supported options. Respect the user's chart type, language and visual choices. Bar Race, Line Race, Bubble Race and Treemap Race are supported. Honor an explicit type. Otherwise choose Line for temporal trajectories or a two-series history, Bar for rankings, and Bubble/Treemap for within-period shares when the metric is positive. The tool default is Bar; choose a supported chartType deliberately when the story calls for another type. Do not imply that Bubble/Treemap area is an absolute scale across dates. Default to 9:16 for direct data unless the user requests another supported ratio (4:5, 1:1 or 16:9). Keep a selected example's composition settings unless the user requests a change. Limits: 2,000 rows, 50 entities, 120 periods; finite values and strictly positive Bubble/Treemap values.

Bubble uses relative areas normalized within each period, bounded native radii and automatic force positions, without x/y coordinates. Treemap uses rectangular areas relative to each period total, with gutters and interpolation during transitions. Neither provides a fixed absolute area scale across time: compare displayed numbers for absolute changes. Animation adds no reported observations. Bubble radius sizing and bubbleScale, custom element positions, margins, typography, image shapes, opacity and timeline placement are not exposed plugin options. Do not promise a pixel-exact reproduction of an Explore composition.

Use `modify_chart` with the complete current envelope and a bounded patch. Preserve its project id, revision and untouched fields. Include replacement dataset, mapping or provenance only when intentionally changed; verify newly retrieved data as for creation. Explain validation errors and request corrected input rather than dropping rows.

Choose `options.visualTheme` on creation or `patch.visualTheme` on modification from the presets currently listed in the connected tool schema (including `infographic` when offered). Omitted theme keeps native defaults on creation and existing styling on modification. Explicit selection reapplies the preset; entity hex colors can be applied on top with `options.colors` or `patch.colors`. Never change data to match a theme.

Images use `mapping.image` and exact published URLs: the five Apple category PNGs at `https://www.chartglow.app/examples/apple-revenue/category-0.png` through `category-4.png`, or the four SVGs `https://www.chartglow.app/examples/renewable-power/hydro.svg`, `https://www.chartglow.app/examples/renewable-power/solar.svg`, `https://www.chartglow.app/examples/renewable-power/wind.svg`, `https://www.chartglow.app/examples/renewable-power/bioenergy.svg`. Keep image/series associations and rights notices. Do not manufacture flag columns, ISO codes or image URLs for country datasets. Country flags and custom images are native editor features: use **Edit in ChartGlow** to customize them. Preserve supported assets already supplied by the user or by a selected Explore example. Arbitrary external i

Preview of the first 4,000 characters. Read full instructions →