FigGenie skill

---
name: figgenie-paper-diagram
description: >
  Draw publication-quality paper figures as SVG the way top venues draw them. Systems papers (OSDI, NSDI, SOSP,
  ASPLOS): architecture / overview figures distilled from 727 human-rated figures, and **mechanism figures** — how
  one part of a system works: a walkthrough with numbered steps, before/after or step-by-step frames, a state machine
  or decision flow, alternatives side by side, the anatomy of one component or data structure, a worked example with
  real values ("架构图", "系统总览图", "画一张 overview 图", "论文插图", "机制图", "原理图", "示意图", "画一张 X 是怎么工作的图",
  "状态机 / 流程图", "before/after 对比图", "算例图") — distilled from 986 rated mechanism figures. AI papers (ICLR, ICML,
  NeurIPS, CVPR, ACL, AAAI): the Figure 1 / teaser / overview figure — a stage pipeline, a model's anatomy, a
  side-by-side contrast, a worked walkthrough on real samples, a concept schematic, a benchmark or dataset overview,
  an agent interaction, a contribution mosaic ("AI 顶会图", "Figure 1", "teaser 图", "方法总览图", "模型结构图", "框架图",
  "benchmark 总览"), distilled from 3,274 labelled Figure 1s, 1,158 user ratings, 336 semantic records, 3,210 exemplars
  and a 998-part library. Workflow: intake → a natural-language brief the user confirms → figure spec → exemplars the
  user picks (AI family) → drawing plan → hand-drawn SVG that follows corpus layouts, palettes, visual encodings and a
  parts library → render / lint / look-and-fix loop → SVG + caption draft, then iterate on feedback. Use whenever the
  user wants such a figure for a paper or talk, asks how one should be laid out, coloured or annotated, wants an
  existing figure critiqued against the venue's conventions ("把这张图重画/改风格"), or asks for a named paper style
  ("Charon 风格"). Vector TeX formulas when the figure needs the paper's symbols; an editable PowerPoint deck on
  request ("转成可编辑的 PPT"). Not for data plots or charts.
---

# figgenie-paper-diagram (FigGenie-Skill) — architecture, mechanism and AI-venue figures the way top papers draw them

You draw the SVG yourself. The scripts do not lay out for you; they measure text, render, lint and show you the
result so you can fix it. Everything is in **pt at print size**: a single-column figure is ≈ 240 pt wide, a
double-column one ≈ 504 pt (an AI-venue figure spans its venue's text width, 396–504 pt), labels 6.5–7 pt, strokes
≈ 0.5–0.6 pt. Read the corpus numbers, then draw like the exemplars, not like a slide.

**Reference, not recipe.** The corpus numbers, the templates, the exemplars and the parts library exist to strengthen
a layout you derive from the figure's own content — what it must make the reader see — never to replace that
derivation. When the material asks for something the references do not cover, draw it, and say so in the plan.

**Two families, two kinds.** `meta.family` says which corpus the figure follows: **systems** (the default: OSDI, NSDI,
SOSP, ASPLOS) or **ai** (ICLR, ICML, NeurIPS, CVPR, ACL, AAAI; `references/ai/`). `meta.kind` says what it shows:
**architecture** (the whole system; 20 layout templates in the systems family) or **mechanism** (how one part works).
Decide both in the brief. The steps below are written for a systems architecture figure; what a mechanism figure and
what an AI-venue figure do differently each get one section after the workflow — menus with frequencies, not rulebooks.

## Workflow (gates in bold)

Work in a folder per figure: `brief.md`, `spec.json`, `plan.md`, `fig.svg`, and the outputs of `check.py`.

1. **Intake → brief.** Read what the user gave (paper sections, notes, an old figure, chat). Write the brief
   (`references/brief-template.md`): what the figure is, every component with new / existing / external tags and its