What problem does it solve? Reproducing complex scientific visualization styles (SHAP beeswarm combos, raincloud plots, Taylor diagrams, chord diagrams) from scratch requires substantial matplotlib expertise and time. This Skill provides ready-to-run Python scripts that render eleven built-in figure templates with deterministic simulated data, exporting PNG, PDF, and SVG outputs. ## Core Features & Use Cases - Eleven Bundled Templates: Covers multiclass SHAP combo, paired raincloud, cross-validation ROC with CI, Taylor diagram, correlation pairgrid, prediction marginal grid, TPE 3D surface, split-violin correlation matrix, circular heatmap, urban cooling combo, and Nature-style chord diagram. - Template Resolution: A renderer script maps template ids, English aliases, or Chinese title fragments (e.g. 云雨图, 泰勒图, 和弦图) to the correct bundled script. - Workspace Output Contract: Copies the template into 绘图复刻/scripts/, runs it, and writes PNG/PDF/SVG files to 绘图复刻/outputs/ with a README entry. - Use Case: A user in the MathModel LaTeX sandbox asks to reproduce a 配对云雨图; the agent runs the renderer with the paired-raincloud id and returns the generated figure paths. ## Quick Start Ask the agent to reproduce a figure template, for example: run the mathmodel-figure-templates renderer with the paired-raincloud template to generate the paired raincloud figure in the workspace.
