| name | nature-figure |
| description | Nature-style publication figure and scientific schematic workflow for Python or R, including figure intent, panel composition, annotations, reproducible export, and source-aware review. Use for submission figures, graphical abstracts, or method schematics. |
Nature Figure
Use scientific-visualization as the integrity baseline, then apply the target venue contract.
- Define the message of the figure in one sentence and list the data or source evidence supporting it.
- Plan panels as a coherent visual argument: overview, comparison, mechanism, validation, or limitation. Give each panel one job.
- Preserve raw values and analysis provenance. Keep a record of transformations, statistical annotations, units, color mapping, and export settings.
- Prefer vector output for diagrams and text; preserve raster resolution for images without implying that upsampling adds detail.
- Review visual hierarchy, color accessibility, font size, caption completeness, and claim strength at final size.
If a schematic is generated from prose, mark it as a conceptual illustration and do not present it as measured data. Do not copy third-party figure assets without a compatible license and attribution.
This is a ScanSci adaptation of the nature-figure workflow from Yuan1z0825/nature-skills.