Agent Skills
› advaitpaliwal/feynman
› figure-composer
figure-composer
GitHub用于从声明、数据集或草稿结果中组装出版级多面板科学图表。涵盖面板规划、布局设计、风格应用及最终合成，确保标签清晰、数据准确并保留完整溯源信息。
Trigger Scenarios
需要规划科学图表的面板结构
生成多面板组合图
审查和修订科学图表的视觉元素与数据一致性
Install
npx skills add advaitpaliwal/feynman --skill figure-composer -g -y
SKILL.md
Frontmatter
{
"name": "figure-composer",
"description": "Compose a publication-grade multi-panel scientific figure from a claim, dataset, or draft result. Use when the task needs panel planning, consistent figure layout, figure review, or final figure assembly."
}
Figure Composer
Use this skill for multi-panel scientific figures.
Workflow:
- State the figure's single claim and list every data artifact that supports it.
- Draft a panel plan with labels, chart types, data sources, required annotations, and what each panel proves.
- Use
figure-stylefor each plot before composing the multi-panel layout. - Render, inspect, and revise for label collisions, axis readability, color meaning, caption fit, and data-source fidelity.
- Save the final figure, source data, generation code, caption, and provenance sidecar.
Do not invent data or hide excluded data in summaries.
Version History
- 54d08a3 Current 2026-07-25 07:15
