Transparent Icons logo

Transparent Icons

Community
neuromechanist
transparent-icons

Use this skill when the user asks to "make an icon", "generate an icon", "create a scientific icon", "make a transparent icon", "make a minimal icon", "icon for a figure", "icon for a paper", "generate an icon set", "batch icons", or wants flat scientific icons (brain, neuron, DNA, EEG cap, microscope) with transparent backgrounds in a journal style. Generates PNG icons with gpt-image-2 through the Codex CLI (or the OpenAI Images API), or Atlas Cloud with explicit opt-in, keeps the model's native alpha, and reads palette and model settings from the project theme made by figures:figure-bible.

Overview

Publisherneuromechanist
Repositoryresearch-skills
Skill nametransparent-icons
Stars
51
Forks
9
Bundled files
5
LicenseBSD-3-Clause
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 5 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by neuromechanist on GitHub. Read the source before you install it.

Installation

Install the Transparent Icons AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/neuromechanist/research-skills/tree/main/plugins/figures/skills/transparent-icons
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/neuromechanist/research-skills.git /tmp/research-skills
mkdir -p .claude/skills
cp -r /tmp/research-skills/plugins/figures/skills/transparent-icons .claude/skills/transparent-icons
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Transparent Icons in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Transparent Icons on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Transparent Icons is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Transparent Icons

Generate flat scientific icons with transparent backgrounds, styled by the project's figures/theme.json, for use as panel elements in figures:scientific-figure, in figures:svg-primitives schematics, or on their own in posters, slides, and grants. Sister skills are invoked by their fully qualified figures:<skill> names through the host's skill-invocation mechanism. Keep host-specific command or tool syntax in the host adapter, not in shared skill instructions.

Backends

Generation goes through the shared backend in plugins/figures/lib/image_backend.py, the same one figures:ai-full-figure uses.

  1. Codex CLI image_gen (default when codex is logged in). No API key needed. The model and effort come from the theme's model_preferences (default gpt-5.6-luna at xhigh), which lands good icons at the cheapest tier in one to two minutes each.
  2. OpenAI Images API (--backend api) with OPENAI_API_KEY.
  3. Atlas Cloud (--backend atlas) with ATLASCLOUD_API_KEY. This is explicit-only; it never changes the Codex-first auto route. The default model is black-forest-labs/flux-schnell, and the asynchronous result is polled with a bounded timeout.

The backend preflights Codex with a 10 second timeout and explains the fix when the binary hangs (approve it once from Terminal.app, or copy codex and codex-code-mode-host to a folder you own, xattr -c both, and set CODEX_BIN). --backend fake renders placeholder icons for tests.

Transparency

Icons are requested with a transparent background and the model's alpha channel is kept as is. --transparency-method auto (default) runs no local processing when the PNG already has a clean cutout, so light strokes are never eaten by a second threshold pass. If the returned image is opaque, auto falls back to threshold (Pillow near-white removal); birefnet (opt-in, --with rembg --with onnxruntime) gives cleaner edges on complex foregrounds.

Usage

Free-form subject with the project theme:

bash
uv run --with pillow python scripts/generate_icon.py \
    "a human brain with EEG electrodes" --theme figures/theme.json -o icons/brain_eeg.png

From the icon bible (curated templates with prompt hints):

bash
uv run --with pillow python scripts/generate_icon.py --template brain-eeg --theme figures/theme.json -o icons/brain_eeg.png
uv run --with pillow python scripts/generate_icon.py --list-templates
uv run --with pillow python scripts/generate_icon.py --category neuroscience --theme figures/theme.json -o icons/

Batch with a placeholder:

bash
uv run --with pillow python scripts/generate_icon.py "a flat icon of a {item}" \
    --batch "neuron,synapse,electrode" --theme figures/theme.json -o icons/

Other flags: --colors "teal,coral" overrides the theme palette for one run; --size 1024 (square); --backend auto|codex|api|fake|atlas; --atlas-model; --codex-bin; --timeout; --verbose; --print-prompt. A log is written next to each output as <out>.codex.log.

Atlas example (explicit opt-in):

bash
ATLASCLOUD_API_KEY=... uv run --with pillow python scripts/generate_icon.py \
    --template neuron --backend atlas --atlas-model black-forest-labs/flux-schnell \
    -o icons/neuron.png

Theme bible

Icons should never be generated without a theme when they need to match other figures. Invoke figures:figure-bible to scaffold and validate figures/theme.json (schema at plugins/figures/schemas/theme.schema.json). The icon prompt uses palette, stroke.weight_px, style_tokens, negative_tokens, composition, and model_preferences. examples/icon_set.py generates a small consistent set from one theme.

Quality assurance

Invoke figures:figure-qa after generation unless the user passed no-qa. Its raster branch checks the alpha channel and corners (--expect-transparent), palette compliance against the theme (--palette figures/theme.json), resolution, and dominant colours, and the vision pass judges whether the icon reads as the intended concept, which is the most common miss.

Additional resources

  • references/icon-bible.md: template schema and category catalog
  • references/prompt-patterns.md: prompt patterns and failure modes for icons
  • scripts/generate_icon.py: CLI entry point
  • scripts/icon-templates.json: bible templates loaded by --template and --category
  • examples/icon_set.py: themed icon set (--smoke for a one-icon check)
  • Shared code: ../../lib/image_backend.py, ../../lib/prompting.py, ../../lib/theme.py

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Transparent Icons AI skill do?

Use this skill when the user asks to "make an icon", "generate an icon", "create a scientific icon", "make a transparent icon", "make a minimal icon", "icon for a figure", "icon for a paper", "generate an icon set", "batch icons", or wants flat scientific icons (brain, neuron, DNA, EEG cap, microscope) with transparent backgrounds in a journal style. Generates PNG icons with gpt-image-2 through the Codex CLI (or the OpenAI Images API), or Atlas Cloud with explicit opt-in, keeps the model's native alpha, and reads palette and model settings from the project theme made by figures:figure-bible.

Why use Transparent Icons on TypingMind?

Because you install it once and use it with any model. Transparent Icons is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Transparent Icons in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/neuromechanist/research-skills/tree/main/plugins/figures/skills/transparent-icons. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Transparent Icons?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Transparent Icons?

As many as you like. As long as a model supports skills, you can use Transparent Icons with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Transparent Icons AI skill free?

Yes. It is published on GitHub by neuromechanist under the BSD-3-Clause license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇