Introduction
This Skill creates editable scientific SVG artwork from a research story and target publication requirements, supporting multi-panel result figures, mechanism diagrams, and graphical abstracts. It reads journal column width, typography, margins, and color constraints, then produces two or three variants with different layouts and visual treatments while keeping labels as live SVG text.
Use Cases
It fits manuscript figures, method pipelines, and TOC graphics for common Nature, Science, Cell, ACS, IEEE, and Elsevier specifications. Real data plots can begin as correctly sized placeholders for later replacement. Color palettes are checked for colorblind readability, and named layers remain easy to edit in Inkscape or Ipe.
Template
Provide the research question, main result, panel count, content or data for each panel, target journal, column width, lab colors, and required labels. Example: create a four-panel RLHF figure with a training pipeline, loss curves, benchmark comparison, and human evaluation.
