| name | tool-inventory |
| description | Purpose-organised catalogue of validated data visualisation libraries, grouped by use case so a tool can be located quickly. Reference this at the start of any dataviz task to get a head start. |
Data Visualisation Tool Inventory
This is a curated, purpose-first catalogue. It exists so that when a task arrives, the right tool can be located by jumping to the matching section rather than reasoning from scratch.
This is a guide, not a cage. Every entry here has been validated as a sensible default for its category. If none of them fit, reach for whatever works — but start here, because the categories below cover the overwhelming majority of real-world dataviz work.
How to use
- Identify the task's primary purpose from the section headings below (e.g. "Graph / network", "Geospatial maps", "Story-driven animation").
- Jump to that section.
- Pick from the listed tools — each entry notes the sweet spot.
- If nothing in the matching section fits, scan adjacent sections before going off-inventory.
1. Static publication figures (Python / R / C++)
For print, PDF reports, journal figures, Typst embeds — anywhere the output is a high-DPI PNG / SVG / PDF.
2. Interactive web charts — general purpose
For embedding in blog posts, web apps, dashboards, static reports. Canvas or SVG output, runs in the browser.
3. High-performance / large-dataset web
When the dataset is big, the refresh rate is high, or latency matters.
4. Low-level / bespoke
When the visual idiom isn't a stock chart — custom layouts, unusual encodings, data art.
5. Python / R dashboards & data apps
Interactive Python and R — dashboards, notebook-embedded viz, data apps.
6. Data storytelling — animated / narrative
Persuasion and explanation. See skills/data-storytelling for the design side.
7. Graph / network visualisation
Nodes and edges — knowledge graphs, social networks, dependency trees.
8. Geospatial / maps / globes
9. Mobile
10. Exploratory / auto-EDA tools
When the user doesn't yet know what they want to show — hand them a tool that surfaces structure.
11. BI-style dashboards (self-hosted)
When the user wants a full BI tool they can host.
12. Database / schema visualisation
Not chart-of-data, but visualisation of the data's structure.
13. Diagrams-as-code
When the visual is a diagram (flow, sequence, architecture, ER) generated from text — great for docs and repos.
14. Domain-specific
15. Java / JVM
16. Upstream data preparation (not visualisation, but adjacent)
For when the data isn't ready to plot yet. Dataviz-adjacent — use before reaching for a chart tool.
For general tabular/JSON cleaning and enrichment before visualisation, defer to the separate Claude-Data-Wrangler plugin.
Discovery beyond the inventory
The catalogue above is a head start, not an endpoint. If a task genuinely doesn't fit any category here, scan these topic feeds before committing to something arbitrary:
State explicitly when going off-inventory and say why nothing on the list fit.
Selection heuristic
- Presentation medium → section (static → §1, web embed → §2/3/4, dashboard → §5/11, story → §6, graph → §7, map → §8, mobile → §9, exploration → §10).
- Data shape → narrow within the section (time-series → uPlot/Lightweight Charts, tabular → ECharts/Plotly, network → §7, gridded/geo → §8).
- Runtime → eliminate mismatches (Python-only, React-only, no-JS, etc.).
- Interactivity level → pick the smallest tool that covers it.
Hand off to skills/choose-tool for the full decision protocol.
