ml-generative-mattergen
Generate inorganic material structures using MatterGen, a diffusion-based generative model.
Browse reusable Agent Skills, each with a clear purpose and practical guidance.
Generate inorganic material structures using MatterGen, a diffusion-based generative model.
Project-level design skill for nxui (nxui.geoql.in). Pins the componentry.fun ground-truth direction — dark-first animated component showcase. Use BEFORE any visual work in this app.
Audit PaperLab figures for accessibility, readability, engineering plotting conventions, print safety, and style consistency across chapters.
Design safe interactive HTML labs from PaperLab notebooks, with bounded inputs, static fallbacks, and student prompts. Use when turning computational examples into guided learning activities.
Generate publication-quality figures for academic papers: LaTeX TikZ method pipelines and Blender 3D renders. Use this skill for method overview figures, pipeline diagrams, architecture diagrams, system flowcharts, 3D rendered teasers, SMPL/FBX mesh renders, skeleton visualizations, method comparisons, or any research figure need. Trigger on: "method figure", "pipeline figure", "architecture diagram", "flow diagram", "teaser", "3D render", "SMPL", "FBX", "mesh render", "Blender", "blender figure", "teaser figure", "skeleton render", "3D comparison", "方法流程图", "画图", "渲染", "三维", "draw my method", "I need a figure for my paper", "help me make a figure".
Use when the user has scientific data (or a prompt alluding to scientific data) and wants a publication-quality figure made from it. A generator drafts and renders a figure that lands a frozen communication goal; an adversarial critic critiques it hard and grades it 1-5 per axis against a fixed rubric (message, aesthetic, clarity, integrity, and a conditional domain-completeness axis), aggregates to 0-100, and decides pass; the generator revises against the critic's findings until the grade clears a threshold or the budget is hit. Both roles may consult the literature (Semantic Scholar + arXiv) to verify domain content (e.g. a pathway figure's gene set, or a benchmark's reported numbers), and conform to / grade against a named journal's figure spec fetched via web search. Not for writing a paper or analyzing a dataset, and not for editing an existing finished image — this renders a figure from data/brief and iterates on it.
Create hand-drawn style SVG flowcharts and diagrams from user descriptions. Supports aspect ratio specification (16:9, 4:3, 1:1, A4, or custom W:H ratios). Use this skill whenever the user asks to create a flowchart, diagram, architecture diagram, process chart, mind map, or any visual representation of a workflow/system using SVG. Also trigger when the user says things like "画个图", "画个流程图", "帮我可视化一下", "draw a diagram", "visualize this process", "make a chart showing how X works", or describes a process/system and asks to "turn it into a picture/chart/diagram". This skill produces SVG files with a distinctive hand-sketched aesthetic — wobbly edges, hatching fills, handwriting fonts — not sterile corporate diagrams.
Design polished text-first and terminal interfaces with stable layout and readable interaction states.
Critique user workflow and interaction sequence across product surfaces.
Updates the shadcn/ui theme by editing OKLCH CSS custom properties in apps/core/styles/tailwind.css. Use when the user wants to change colors, theme appearance, or match a brand palette.