ggai

Generate a single biomedical entity as a reusable cutout asset (transparent alpha background, no labels) suitable for compositing into diagrams. Use when the goal is to produce one isolated subject — a cell, vessel, tissue cutaway, signaling cloud — that will later be placed into a scene, rather than rendering a complete figure.

Produce a schematic, illustration, biomedical cartoon, or pure conceptual figure without underlying data. Calls an image model directly with a structured prompt, generates one or more candidates, scores them on basic visual proxies, and saves the best. Use when the goal is "draw / illustrate / diagram / sketch" and no data frame is in scope.

Decide which R visualization engine to use for a figure: ggplot2, grid/grob, base graphics, ComplexHeatmap, circlize, ggraph/DiagrammeR, htmlwidget (plotly/leaflet), or composite (patchwork/cowplot/aplot). Load this skill when the user names a non-ggplot library, when the figure shape suggests a specialized library, or when ggplot would require contortions to reach the intended result.

Produce interactive web visualizations using `plotly`, `leaflet`, `DT`, `networkD3`, or any other `htmlwidgets`-based library. Default output is a self-contained HTML file (always works). Static PNG export is supported when the optional `webshot2` package and a headless Chrome (via `chromote`) are available.

Compose multiple plots into a single figure using `patchwork`, `cowplot`, or `aplot`. Use whenever the user wants more than one panel — labeled multi-panel figures, side-by-side comparisons, A/B/C panels with shared legend, plots stacked over a metadata strip, or any layout that exceeds what a single ggplot's `facet_*` can express.

Lift the visual quality of an existing data figure by redrawing it with an image model that uses the original ggplot as a hard constraint. Preserves data semantics (positions, groups, scales, labels) while improving typography, composition, and surface polish. Use when the user has a working figure but wants it to look like a publication, cover, or polished talk slide.

Produce circular / genomic figures using the `circlize` package — chord diagrams, circular genome browsers with multiple tracks, ideograms, Circos-style multi-track figures, and similar. Use when the figure intent is circular layout or when the user names `circlize`, `circos`, "Circos", "chord diagram", or describes a multi-track circular genomic figure.

Produce annotated heatmaps using the ComplexHeatmap package — matrix expression heatmaps with multi-track row/column annotations, dendrograms, splits, oncoPrint, and HeatmapList combinations. Use when the figure intent is a heatmap with non-trivial annotation, row or column splitting, or multiple stacked heatmaps. ggplot's `geom_tile` does not scale to these requirements.

Core ggai Agent persona for data-faithful, reviewable, iterative visual analysis.

Turn a data frame plus a natural-language instruction into a publication-quality data figure. Default engine is ggplot2; escalate to ComplexHeatmap / circlize / ggraph / base via `ggai-engine-selection` when ggplot would force contortions. Use whenever the user's goal references a tabular data source (by name, `@`-mention, file path, or "this data" anaphora) and asks for a visualization.

Meta-router for ggai goals. Load this skill first when the user's goal is ambiguous, multi-modal, or could plausibly map to more than one of: data plot, figure polish, direct illustration, glyph-augmented chart, or iterative edit. The skill classifies intent and tells you which downstream skill to load next. Skip this skill and load a task skill directly when the goal is unambiguous (e.g. a clear data-plot phrasing or a clear "draw a cartoon" phrasing).

R graphics font and multilingual text guidance for ggplot output across devices and platforms.

Guidance for adapting paper figures, screenshots, and reference images to user data in ggai.

Single-cell and spatial transcriptomics plotting guidance for ggai Agents.