Icons logo

Icons

Community
laguagu
icons

Find, fetch, and install the right icon or logo from the right source — brand marks, country flags, file-type icons (PDF, DOCX, ZIP), and UI glyphs — and keep them visually consistent with the app. Use when the project's icon library has no match, when svgl comes up empty, or when the user asks for a flag, a file-type badge, a brand logo, or just "an icon for X". Covers the Iconify search API (200k+ icons across flags, file types, logos and UI sets), the svgl shadcn registry for full-colour brand logos, family and stroke-weight matching so a borrowed icon does not look pasted in, and how to reach svgrepo through browser automation when curl and WebFetch are blocked. Triggers on "lisää ikoni", "maan lippu", "flag icon", "PDF icon", "file type icon", "brand logo", "sign in with Google/GitHub", "language switcher", "svgl", "iconify", "find an icon". For overall visual direction rather than sourcing one specific mark, use ui-signature; for installing shadcn components generally, use shadcn.

Overview

Publisherlaguagu
Repositoryclaude-code-nextjs-skills
Skill nameicons
Stars
64
Forks
18
Bundled files
Instructions only
LicenseMIT
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.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

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

Installation

Install the 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.
  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/laguagu/claude-code-nextjs-skills.git /tmp/claude-code-nextjs-skills
mkdir -p .claude/skills
cp -r /tmp/claude-code-nextjs-skills/skills/icons .claude/skills/icons
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable 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 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 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.

Sourcing icons

An icon is wrong when it is the wrong family, not when it is the wrong shape. Pick the source by what kind of mark it is, then match it to what the app already uses.

Route by mark type

NeedSourceWhy
App UI glyph (menu, upload, search)project's existing library — Lucide in shadcn projectsnever add a second family for a glyph the current one has
Country flagIconify circle-flags / flag / flagpackreal flag geometry, MIT, no emoji-font dependency
File type (PDF, DOCX, ZIP)Iconify vscode-icons / catppuccin / material-icon-theme (colour), or lucide:file-text (mono)pick colour or mono to match the surrounding UI, not both
Brand logo, full colour, light/dark variantssvglcurated, has wordmarks and theme variants
Tech / infra logo (OpenShift, Docker, Postgres, Redis)Iconify logos (1800+, CC0), devicon, skill-iconssvgl has almost none of these; logos covers the whole stack in colour
Brand logo, monochrome long tailIconify simple-icons (3400+, CC0)covers brands svgl does not
Anything elseIconify search across all sets200k+ icons, one API

Check the project's own library first: lucide:file-text, lucide:flag, lucide:file and friends exist, and one family stays one family.

Iconify — the general engine

Public API, no auth, no key. Search returns prefix:name ids.

bash
# search everything
curl -s "https://api.iconify.design/search?query=pdf&limit=32"

# constrain to sets you actually want (this is the important flag)
curl -s "https://api.iconify.design/search?query=pdf&prefixes=vscode-icons,catppuccin,lucide"

# what is in a set / who owns the licence
curl -s "https://api.iconify.design/collections?prefixes=circle-flags,simple-icons"

# fetch the SVG itself
curl -s "https://api.iconify.design/circle-flags:fi.svg" -o public/flags/fi.svg

Browse visually at https://icon-sets.iconify.design when the user should choose. The collections response carries each set's licence — nearly all of them are permissive; Twemoji is CC BY 4.0 and needs visible attribution, which is the one that costs you something.

Iconify SVGs come out at width="1em" height="1em", and monochrome sets use currentColor — so className="size-4 text-muted-foreground" just works in Tailwind. Colour sets (vscode-icons, circle-flags, twemoji) ignore currentColor by design.

Getting them into the project

Default: write the SVG to a file and inline it. No runtime dependency, no network call, no bundle growth beyond the markup.

bash
curl -s "https://api.iconify.design/vscode-icons:file-type-pdf2.svg" -o components/icons/pdf.svg

Only reach for a package when the project needs dozens of icons from one set:

  • @iconify/react renders <Icon icon="circle-flags:fi" /> but fetches from the Iconify API at runtime — an external request per icon set, on every visitor. Do not ship that to production without saying so; pair it with @iconify-json/<prefix> offline data, or skip it.
  • unplugin-icons + @iconify-json/<prefix> compiles icons in at build time. This is the right choice for a bundler-based project pulling in a whole set.

svgl — brand logos

Still the best source for product and company logos in colour, with light/dark variants.

Register the svgl registry once in components.json, then add logos by name — svgl's docs are explicit that the full https://svgl.app/r/<name>.json URL is not passed to the CLI:

jsonc
// components.json
{ "registries": { "@svgl": "https://svgl.app/r/{name}.json" } }
bash
bunx --bun shadcn@latest add @svgl/<name>     # several names in one command, space-separated

components.json must exist — if it does not, stop and tell the user to run bunx --bun shadcn@latest init first.

Resolve the exact name before installing:

  • https://api.svgl.app?search=<query> — search
  • https://api.svgl.app/categories then https://api.svgl.app/category/<category> — browse
  • https://api.svgl.app/svg/<name>.svg — raw markup

svgl is software and brand logos only. Its categories are Software, Framework, AI, Library, Payment, Design and similar — there are no flags, no file types, no UI glyphs. ?search=finland returns SVG not found, and that is expected, not a bug: switch to Iconify rather than retrying svgl with a different phrasing.

Docs: https://svgl.app/docs/shadcn-ui

Making a borrowed icon fit

A single icon from a foreign family is the most common tell of pasted-in work. Match, in this order:

  1. Stroke vs fill — do not drop a filled icon into an outline set. Most Iconify sets ship both (teenyicons:pdf-outline / teenyicons:pdf-solid, mingcute:pdf-line / mingcute:pdf-fill); take the variant that matches.
  2. Stroke width and grid — Lucide is 24px grid, 2px stroke. Tabler and Hugeicons sit close enough to pass; Material and Carbon do not, they read thinner and boxier.
  3. Colour — a mono icon inherits currentColor; a colour icon is a logo-like object and needs the same treatment as a logo (own size, own alignment, not tinted).
  4. Optical size — colour file-type icons carry their own padding and look smaller at the same box; bump one step (size-5 against a row of size-4) and check by eye.

When mixing is unavoidable, mix by category, not per icon: all flags from one flag set, all file types from one file-type set, all UI glyphs from the project library.

Give every icon an accessible name or hide it. An icon-only button is unlabelled to a screen reader unless you add aria-label on the button — the SVG's own title does not substitute. An icon sitting next to text is decorative and repeats what the text already says: aria-hidden="true". Icon component libraries do not do this for you; inlined raw SVG certainly does not.

Flags specifically

  • Never use emoji flags (🇫🇮) when Windows is in the audience — Segoe UI Emoji ships no flag glyphs, so they render as two letters (FI). They look fine on macOS and iOS, which is exactly why this ships broken.
  • circle-flags for avatar-style circles, flagpack for rounded rectangles with a border, flag (flag-icons) for plain rectangles. Pick one and stay in it.
  • Do not search for the country name?query=finland returns only emoji sets (twemoji:flag-finland, openmoji:…), never the real flag sets, because those are indexed by ISO code. Construct the id directly from ISO 3166-1 alpha-2, lowercase: circle-flags:fi, flagpack:se. flag needs an explicit ratio suffix — flag:fi-4x3 or flag:fi-1x1; bare flag:fi is a 404.
  • Language flags are politically loaded — for a language switcher prefer the language name over a flag.

JSX conversion

Inlining raw SVG into JSX: stroke-widthstrokeWidth, classclassName, style strings → objects, <style> block rules inlined onto elements.

Rename every id per mark. Iconify serves the same mask id to every icon in a set — circle-flags:fi and circle-flags:se both arrive as <mask id="SVGuywqVbel">. Inline both into one language switcher and the second mask wins for both flags. Suffix them (SVGmask-fi, SVGmask-se) and update the matching mask="url(#…)" reference. Same trap with gradients in colour logos.

When nothing is found — svgrepo, flaticon

Neither can be fetched with curl or WebFetch. Verified 2026-08-19: svgrepo.com returns HTTP 429 to every programmatic request, browser user-agent or not; flaticon.com returns 403. Do not retry with header tricks — the block is not user-agent based.

svgrepo does work through a real browser. If browser automation is available (the claude-in-chrome tools), that is a legitimate route, not a scraping workaround — the pages render normally and same-origin fetch is allowed:

  1. Navigate to https://www.svgrepo.com/vectors/<term>/ and read the result list. Icon links have the form https://www.svgrepo.com/svg/<id>/<slug>.
  2. Open the icon page and read its LICENSE: field (shown next to COLLECTION: and UPLOADER:, e.g. CC0 License). Check it per icon — svgrepo mixes CC0, MIT, CC BY and non-commercial packs on the same site.
  3. From a page on that origin, fetch('https://www.svgrepo.com/download/<id>/<slug>.svg') returns the raw SVG (200, image/svg+xml). Write it into the project yourself.

Without browser automation, ask the user to download the file in the browser and drop the .svg in the project. Do not suggest svgapi.com — it is svgrepo's official API but starts at $9.90/mo for 5,000 calls, and without a key returns {"error": "Your domainHash is not valid."}. The free tier is a case-by-case grant for non-commercial accessibility projects only.

flaticon stays off limits: downloads need an account, and its free tier requires visible attribution in the shipped product. Ask the user for the file if they specifically want one from there.

Mixed licensing plus mixed drawing style is why both remain a last resort rather than a first stop, even where the browser route works.

Better fallbacks before giving up on Iconify: search a synonym (document for file, invoice for receipt), search unprefixed to see every set, or browse the nearest category at icon-sets.iconify.design. 200k icons is usually enough.

Local organisations, universities and public bodies are in no icon set at all — a regional university, a national research centre or a municipality returns nothing from svgl or Iconify, and that is the expected answer, not a failed search. Ask the user for the official asset, or use a lettermark in a circle. Never invent a logo for a real organisation.

Frequently asked questions

What does the Icons AI skill do?

Find, fetch, and install the right icon or logo from the right source — brand marks, country flags, file-type icons (PDF, DOCX, ZIP), and UI glyphs — and keep them visually consistent with the app. Use when the project's icon library has no match, when svgl comes up empty, or when the user asks for a flag, a file-type badge, a brand logo, or just "an icon for X". Covers the Iconify search API (200k+ icons across flags, file types, logos and UI sets), the svgl shadcn registry for full-colour brand logos, family and stroke-weight matching so a borrowed icon does not look pasted in, and how to...

Why use Icons on TypingMind?

Because you install it once and use it with any model. 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 Icons in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/laguagu/claude-code-nextjs-skills/tree/main/skills/icons. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use 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 Icons?

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

Is the Icons AI skill free?

Yes. It is published on GitHub by laguagu under the MIT 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 👇