Gallery · Quick Start · Features · Spec · Verification

lanshu-animated-architecture-diagram is a Codex skill and local renderer for creating premium black-canvas technical diagrams with hand-drawn typography, editable Excalidraw output, static PNG previews, and genuinely animated GIFs.

It is designed for article explanations, system architecture diagrams, process diagrams, and DailyDoseOfDS-style black-background technical sketches.

The default visual system uses a dark canvas, moving flow highlights, pulsing modules, subtle grain, vignette, and a top-right hand-drawn signature.

Claude Loops

Memory Pack

• Generates .excalidraw, .png, and animated .gif from one JSON spec
• Produces real animation with moving glow points and pulsing module highlights
• Keeps the .excalidraw source editable and text-based
• Uses local Python rendering through Pillow
• Does not require Excalidraw, browser automation, ImageMagick, remote APIs, or external icon libraries
• Includes frame-diff verification to prove GIF motion
• Uses a fixed high-quality layout for clean technical storytelling

Each render produces:

<basename>.excalidraw
<basename>.png
<basename>.gif

The default canvas is:

1210 x 1138
20 fps
41 frames
2.05 seconds

git clone https://github.com/cclank/lanshu-animated-architecture-diagram.git
cd lanshu-animated-architecture-diagram
python3 -m pip install -r requirements.txt
python3 scripts/render_animated_diagram.py \
  --spec assets/default-spec.json \
  --outdir outputs \
  --basename sample \
  --verify

Place this folder in your Codex skills directory:

~/.codex/skills/lanshu-animated-architecture-diagram

Typical local install path:

${CODEX_HOME:-$HOME/.codex}/skills/lanshu-animated-architecture-diagram

Install the runtime dependency:

python3 -m pip install -r requirements.txt

Invoke the skill by name:

Use $lanshu-animated-architecture-diagram to turn this article into a premium hand-drawn animated architecture GIF.

Chinese prompt example:

用 $lanshu-animated-architecture-diagram 把这篇文章整理成岚叔动态架构图，输出 GIF、PNG 和 Excalidraw。

Start from the bundled template:

cp assets/default-spec.json work/my-diagram-spec.json

Render:

python3 scripts/render_animated_diagram.py \
  --spec work/my-diagram-spec.json \
  --outdir outputs \
  --basename my-diagram \
  --verify \
  --check

The --verify flag prints sampled frame differences. Nonzero changed pixels confirm that the GIF is genuinely animated.

The --check flag validates the generated PNG, GIF, and Excalidraw output contract and exits nonzero if a required property fails. It checks dimensions, GIF frame count and frame duration, sampled GIF motion, unique Excalidraw IDs, text font family, and that no external files are embedded.

The renderer uses assets/default-spec.json as a compact art-directed template.

Most edits happen in these fields:

signature
title.prefix
title.highlight
title.subtitle
inputs
core.cards
decision
output
left_panel
center_panel
right_panel

Supported icon keys:

folder
file
scan
shield
db
hash
package

For details, see references/spec-format.md.

Validate the skill structure:

python3 ${CODEX_HOME:-$HOME/.codex}/skills/.system/skill-creator/scripts/quick_validate.py \
  ${CODEX_HOME:-$HOME/.codex}/skills/lanshu-animated-architecture-diagram

Validate GIF media parameters:

ffprobe -v error -select_streams v:0 -count_frames \
  -show_entries stream=width,height,r_frame_rate,avg_frame_rate,nb_read_frames \
  -show_entries format=duration \
  -of default=noprint_wrappers=1 outputs/my-diagram.gif

Validate animation:

python3 scripts/render_animated_diagram.py \
  --spec assets/default-spec.json \
  --outdir outputs \
  --basename sample \
  --verify \
  --check

Required:

• Python 3.9+
• Pillow 10.0.0+

Install Python packages with:

python3 -m pip install -r requirements.txt

Optional:

• ffprobe for media inspection
• Excalidraw web app or editor plugin for manual editing of generated .excalidraw files

lanshu-animated-architecture-diagram/
├── SKILL.md
├── README.md
├── LICENSE
├── requirements.txt
├── agents/
│   └── openai.yaml
├── assets/
│   ├── default-spec.json
│   └── previews/
│       ├── claude-loops.gif
│       └── memory-pack.gif
├── references/
│   └── spec-format.md
└── scripts/
    └── render_animated_diagram.py

This project intentionally keeps the visual system narrow:

• Dark canvas
• Hand-drawn title treatment
• Top input strip
• Middle core pipeline
• Bottom source, layer, and pack panels
• Top-right signature
• Clean static diagram with motion added only in GIF overlays

That constraint keeps outputs consistent and polished across different architecture topics.

MIT