What problem does it solve? Creating publication-quality scientific figures that meet Nature-family journal standards requires deep knowledge of layout, typography, color semantics, and export rules, which most researchers must learn through trial and error. ## Core Features & Use Cases - Figure contract workflow: Defines the core conclusion, evidence hierarchy, panel map, and export requirements before any plotting code is written. - Dual backend support: Provides dedicated tracks for Python (matplotlib, seaborn, subplot_mosaic) and R (ggplot2, patchwork, ComplexHeatmap), with a blocking gate that enforces exclusive use of the selected backend. - Editable vector export: Enforces SVG-first output with selectable text via mandatory rcParams rules, plus PNG/PDF/TIFF secondary exports. - Use Case: A researcher needs a multi-panel results figure combining a schematic, heatmap, and forest plot for a Nature Machine Intelligence submission; the skill guides backend selection, applies restrained semantic palettes, and delivers journal-ready SVG output. ## Quick Start Ask the assistant to create a Nature-style multi-panel figure from your data and specify whether you want to use Python or R.
