skills/43-wentorai-research-plugins/skills/writing/latex/tikz-diagrams-guide/SKILL.md
Create publication-quality scientific diagrams with TikZ in LaTeX
npx skillsauth add brycewang-stanford/Awesome-Agent-Skills-for-Empirical-Research tikz-diagrams-guideInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
A skill for creating publication-quality scientific diagrams directly in LaTeX using the TikZ package. Covers basic drawing commands, flowcharts, neural network architectures, data flow diagrams, and integration with PGFplots for camera-ready figures.
\documentclass{article}
\usepackage{tikz}
\usetikzlibrary{arrows.meta, positioning, shapes.geometric, calc, fit}
\begin{document}
\begin{tikzpicture}
% Your drawing commands here
\end{tikzpicture}
\end{document}
% Lines and shapes
\draw (0,0) -- (3,0) -- (3,2) -- cycle; % Triangle
\draw[thick, ->] (0,0) -- (4,0); % Arrow
\draw[dashed, blue] (0,0) circle (1.5); % Dashed circle
\filldraw[fill=gray!20, draw=black] (2,1) ellipse (1 and 0.5);
% Nodes (text labels with optional shapes)
\node[draw, rectangle, minimum width=2cm] (A) at (0,0) {Input};
\node[draw, circle] (B) at (3,0) {Process};
\draw[->] (A) -- (B);
% Relative positioning (requires positioning library)
\node[draw, rectangle] (C) [right=2cm of B] {Output};
\draw[->] (B) -- (C);
\begin{tikzpicture}[
block/.style={rectangle, draw, fill=blue!10, text width=5em,
text centered, rounded corners, minimum height=3em},
decision/.style={diamond, draw, fill=green!10, text width=4em,
text centered, inner sep=0pt, aspect=2},
line/.style={draw, -Stealth}
]
\node[block] (data) {Collect Data};
\node[block, below=1cm of data] (clean) {Clean \& Preprocess};
\node[decision, below=1cm of clean] (valid) {Valid?};
\node[block, below=1cm of valid] (analyze) {Analyze};
\node[block, right=2cm of valid] (fix) {Fix Issues};
\path[line] (data) -- (clean);
\path[line] (clean) -- (valid);
\path[line] (valid) -- node[right] {Yes} (analyze);
\path[line] (valid) -- node[above] {No} (fix);
\path[line] (fix) |- (clean);
\end{tikzpicture}
\begin{tikzpicture}[
neuron/.style={circle, draw, minimum size=0.8cm, fill=orange!20},
layer/.style={rectangle, draw, dashed, inner sep=0.3cm}
]
% Input layer
\foreach \i in {1,2,3,4} {
\node[neuron] (I\i) at (0, -\i*1.2) {};
}
% Hidden layer
\foreach \j in {1,2,3} {
\node[neuron, fill=blue!20] (H\j) at (3, -\j*1.2 - 0.6) {};
}
% Output layer
\foreach \k in {1,2} {
\node[neuron, fill=green!20] (O\k) at (6, -\k*1.2 - 1.2) {};
}
% Connections
\foreach \i in {1,2,3,4} {
\foreach \j in {1,2,3} {
\draw[->] (I\i) -- (H\j);
}
}
\foreach \j in {1,2,3} {
\foreach \k in {1,2} {
\draw[->] (H\j) -- (O\k);
}
}
% Labels
\node[above=0.5cm of I1] {Input};
\node[above=0.5cm of H1] {Hidden};
\node[above=0.5cm of O1] {Output};
\end{tikzpicture}
\usepackage{pgfplots}
\pgfplotsset{compat=1.18}
\begin{tikzpicture}
\begin{axis}[
xlabel={Epoch},
ylabel={Loss},
legend pos=north east,
grid=major,
width=8cm, height=6cm
]
\addplot[blue, thick, mark=none] table {
1 0.95
5 0.72
10 0.45
20 0.22
30 0.15
50 0.08
};
\addlegendentry{Training}
\addplot[red, thick, dashed, mark=none] table {
1 0.98
5 0.75
10 0.52
20 0.35
30 0.30
50 0.28
};
\addlegendentry{Validation}
\end{axis}
\end{tikzpicture}
1. Font consistency:
- Use the same font family as your document body
- Minimum 8pt for axis labels and annotations
- Match font size to caption text
2. Color considerations:
- Use colorblind-friendly palettes (avoid red-green only)
- Ensure figures are readable in grayscale
- Use patterns or line styles as secondary differentiators
3. Size and resolution:
- TikZ produces vector output (PDF) -- always sharp
- Set figure width to match column width (single or double)
- Use consistent sizing across all figures in the paper
4. Labeling:
- Label all axes with units
- Use (a), (b), (c) for sub-figures
- Place legends inside the plot area when possible
% standalone.tex -- compile separately, include as PDF
\documentclass[tikz, border=2mm]{standalone}
\usetikzlibrary{arrows.meta, positioning}
\begin{document}
\begin{tikzpicture}
% ... your diagram ...
\end{tikzpicture}
\end{document}
% In your main document:
% \includegraphics{standalone.pdf}
| Library | Purpose |
|---------|---------|
| positioning | Relative node placement (right=of, below=of) |
| arrows.meta | Modern arrow tip styles |
| shapes.geometric | Diamond, trapezium, ellipse nodes |
| calc | Coordinate calculations |
| fit | Fit a node around a set of other nodes |
| decorations.pathreplacing | Braces, snakes, zigzag lines |
| backgrounds | Draw behind other elements |
| matrix | Grid-based node layouts |
tools
Autonomous multi-round research review loop using MiniMax API. Use when you want to use MiniMax instead of Codex MCP for external review. Trigger with "auto review loop minimax" or "minimax review".
tools
Autonomous research review loop using any OpenAI-compatible LLM API. Configure via llm-chat MCP server or environment variables. Trigger with "auto review loop llm" or "llm review".
tools
Autonomous multi-round research review loop using MiniMax API. Use when you want to use MiniMax instead of Codex MCP for external review. Trigger with "auto review loop minimax" or "minimax review".
tools
Autonomous research review loop using any OpenAI-compatible LLM API. Configure via llm-chat MCP server or environment variables. Trigger with "auto review loop llm" or "llm review".