mcpField
I
infyicon-mcp

infyicon-mcp

Remote
github.com/krupalghori44-devMedia & Design0Updated 5w ago
Repository

About infyicon-mcp

infyicon-mcp lets AI clients find free icons and retrieve embed-ready SVG, PNG URLs, or webfont classes. It serves a large icon catalog through a hosted MCP endpoint or a relay you can run locally.

Key features

  1. 01Search icons by keyword and style
  2. 02Retrieve SVG for one or multiple icons
  3. 03Get transparent PNG URLs in set sizes
  4. 04Find related icons and popular picks
  5. 05Look up UI webfont CSS classes

Use cases

  1. 01Find icons for an interface design
  2. 02Retrieve SVG markup for a project
  3. 03Get matching icon variants
  4. 04Look up webfont classes for UI elements

MCP support

Exposes read-only tools for icon search, SVG and PNG retrieval, related icons, categories, and webfont lookup. Clients can connect to the hosted endpoint or use the stdio bridge.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://infyicon.com/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

8
ParameterTypeRequiredDescription
limitintegerNoMax results per page (default 10, max 30)
querystringYesWhat to search for, e.g. "shopping cart", "doctor", "hammer" (1-4 plain words work best)
styleenumNoOptional: restrict to one style; omit to get all four styles mixedoutline | fill | color-outline | color-fill
offsetintegerNoOptional: skip this many results for pagination. Pass next_offset from the previous response to get the next page.

Details

From the project README

Infyicon MCP Server

Official MCP Registry npm Glama TDQS One-click setup

Search 161,000+ free hand-drawn vector icons from any AI assistant — Claude, ChatGPT, Cursor, VS Code / Copilot, Codex, Gemini, Windsurf, Zed, Cline, Goose, LM Studio, JetBrains and every other MCP client. Four matching styles (outline, fill, color-outline, color-fill), ready-to-embed SVG markup, transparent PNGs at 16–512 px, and a 61,000-glyph UI webfont.

Tools (v1.2.0)

ToolWhat it does
search_iconsKeyword search over 161,000+ icons, optional style filter, offset paging. Returns ids, names, page/SVG/PNG URLs, tags.
get_popular_iconsFeatured icons when there is no search term.
get_related_iconsSame-theme icons across styles — build a matching set or offer variations.
get_icon_svgFull ready-to-embed SVG markup for one icon id.
get_icon_svg_bulkSVG markup for 1–20 icon ids in one call.
get_icon_pngTransparent PNG URLs at 16 / 24 / 32 / 64 / 128 / 256 / 512 px.
list_categoriesMost popular categories with counts and category-page URLs.
search_uiconsCSS class names in the Infyicon UI webfont (<i class="ii-r-home"></i>).

Old names from v1.0/1.1 (popular_icons, related_icons, bulk_svg, uicons_lookup) are still accepted as aliases.

Quick start

Claude.ai / Claude Desktop — Settings → Connectors → Add custom connector → https://infyicon.com/mcp. Or install the desktop extension: infyicon.mcpb.

Claude Code

claude mcp add --transport http infyicon https://infyicon.com/mcp
# or as a plugin
claude plugin marketplace add krupalghori44-dev/infyicon-mcp && claude plugin install infyicon@infyicon

ChatGPT — Settings → Apps & Connectors → Advanced → Developer mode → Create → MCP URL https://infyicon.com/mcp, auth None.

Cursor — Add to Cursor or ~/.cursor/mcp.json:

{ "mcpServers": { "infyicon": { "url": "https://infyicon.com/mcp" } } }

VS Code / GitHub Copilot — .vscode/mcp.json:

{ "servers": { "infyicon": { "type": "http", "url": "https://infyicon.com/mcp" } } }

Codex CLI — codex mcp add infyicon --url https://infyicon.com/mcp

Gemini CLI (extension, recommended) — gemini extensions install https://github.com/krupalghori44-dev/infyicon-mcp — adds the server, a /icons <query> command and usage context. Or plain MCP: gemini mcp add --transport http infyicon https://infyicon.com/mcp · Gemini app — Settings → Connected apps → Add custom MCP server

Windsurf — { "mcpServers": { "infyicon": { "serverUrl": "https://infyicon.com/mcp" } } }

Cline / Roo Code — { "mcpServers": { "infyicon": { "type": "streamableHttp", "url": "https://infyicon.com/mcp" } } }

Any stdio-only client (Zed, Goose, LM Studio, JetBrains, Perplexity…) — the zero-dependency npm bridge relays to the hosted server:

{ "mcpServers": { "infyicon": { "command": "npx", "args": ["-y", "infyicon-mcp"] } } }

Run it yourself

npx -y infyicon-mcp             # stdio bridge
npx -y infyicon-mcp --http 3000 # local HTTP mirror at POST http://127.0.0.1:3000/mcp
docker build -t infyicon-mcp . && docker run -i infyicon-mcp

standalone.js relays initialize, tools/list and tools/call to https://infyicon.com/mcp (override with INFYICON_MCP_UPSTREAM), so it never goes stale. mcp.js is the reference implementation that runs on infyicon.com (Express, hand-rolled JSON-RPC, no SDK).

Repository layout

  • standalone.js — npm package infyicon-mcp (stdio / --http)
  • mcp.js — hosted server source
  • mcpb/ — Claude Desktop extension source; build with npx @anthropic-ai/mcpb pack mcpb dist/infyicon.mcpb
  • .claude-plugin/ + .mcp.json — Claude Code plugin marketplace
  • gemini-extension.json + GEMINI.md + commands/ — Gemini CLI extension (/icons command)
  • connect/connect.html — the one-click setup page served at https://infyicon.com/connect
  • server.json — Official MCP Registry manifest

Privacy

Tool arguments (search words, icon ids, style filters) are sent to infyicon.com over HTTPS; responses contain icon metadata, SVG markup and PNG URLs. No account, API key, telemetry or local file access. Privacy policy: https://infyicon.com/privacy · Contact: [email protected]

License

Server and bridge code: MIT. Icons: free with attribution — https://infyicon.com/license

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.