Icon MCP Server

@agentic-ph

リポジトリ

基本情報

カテゴリ

その他

トランスポート

stdio

作者

agentic-ph

投稿者

John Dave Decano

Icon MCP Server について

A powerful Model Context Protocol (MCP) server that provides unified search capabilities across multiple icon libraries with fuzzy search, intelligent caching, and comprehensive filtering options.

#icon#icons

Icon MCP Server の設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "icon-search": {
      "command": "npx",
      "args": [
        "icon-mcp"
      ],
      "env": {
        "CACHE_TTL": "300000",
        "DEFAULT_SEARCH_LIMIT": "50",
        "LOG_LEVEL": "info"
      }
    }
  }
}

Icon MCP Server のツール

Search for icons by name across all or specific libraries with fuzzy matching

Get detailed information about a specific icon

Get a list of all available icon libraries

Get detailed information about a specific library

Find icons by category or tag with fuzzy matching

概要

What is Icon Mcp?

Icon Mcp is a Model Context Protocol (MCP) server that provides unified search across multiple icon libraries with fuzzy search, intelligent caching, and comprehensive filtering options. It is designed for developers integrating icon search into MCP‑compatible clients like Claude Desktop and VS Code.

How to use Icon Mcp?

Install globally via npm install -g icon-mcp, build the icon index with npm run build-icons, then start the server with npm start. Configure the server in your MCP client’s JSON (e.g., claude_desktop_config.json) by setting the command to npx icon-mcp and optionally customizing environment variables like CACHE_TTL or FUZZY_THRESHOLD.

Key features of Icon Mcp

Multi‑library search across Bootstrap Icons, Feather, Octicons, and Tabler Icons
Fuzzy search with typo tolerance using Fuse.js
Intelligent caching with configurable TTL and LRU eviction
Filtering by library, style, category, and tags
NPM package integration for automatic icon library management
Full TypeScript type safety and excellent developer experience

Use cases of Icon Mcp

Quickly find icons by name across multiple libraries during UI development
Integrate icon search into AI‑powered code assistants or chat tools
Build custom icon pickers with fuzzy matching and category filtering
Automate icon discovery for design‑to‑code workflows

よくある質問

Which icon libraries are supported?

Bootstrap Icons, Feather, Octicons, and Tabler Icons are supported out of the box. Additional libraries can be added via the provider system.

How do I add a custom icon library?

Add the library as an NPM dependency, create a provider class extending `BaseNpmProvider`, register it in the provider index, and rebuild the icon index with `npm run build-icons`.

How can I configure caching behavior?

Set the `CACHE_TTL` (default 300 000 ms) and `CACHE_MAX_SIZE` (default 1000) environment variables in your MCP client configuration.

What MCP tools are available?

The server exposes `search_icons`, `get_icon`, `list_libraries`, `get_library_info`, and `search_by_category` tools. `search_icons` supports parameters like `query`, `libraries`, `fuzzy`, `limit`, and `threshold`.

How do I verify the connection is working?

After configuration, check that the server appears in your client’s server list, test the `list_libraries` tool, and look for connection messages in the logs.

「その他」の他のコンテンツ

Screenpipe

screenpipe

YC (S26) | Open Computer History | Continuously record your company computer work, map your workflows, help you find work worth automating, and power your agents' context

22.0K

YOUB

Connect your AI assistant to your own YOUB endurance training account. Read recorded activities with full analysis, planned sessions, race goals, wellness (HRV, sleep, resting HR) and your athletic profile. Create or mov

Activepieces

activepieces

AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents

23.2K