MCP.so
ログイン
O

Opengraph.io MCP Server

@securecoders

基本情報

カテゴリ

その他

トランスポート

stdio

作者

securecoders

投稿者

Charles Wooley

Opengraph.io MCP Server について

Opengraph.io MCP Server — Instant Link Metadata Extraction

#link-preview#unfurling#crawling#opengraph#screenshots#web-scraping#link#preview#scraping

Opengraph.io 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": {
    "opengraph": {
      "command": "npx",
      "args": [
        "-y",
        "opengraph-io-mcp"
      ],
      "env": {
        "OPENGRAPH_APP_ID": "YOUR_APP_ID_HERE"
      }
    },
    "opengraph-io-hosted-mcp": {
      "url": "https://mcp.opengraph.io/mcp",
      "headers": {
        "x-app-id": "YOUR_APP_ID_HERE"
      }
    }
  }
}

Opengraph.io MCP Server のツール

ツール一覧はまだありません

この掲載には表示できるツール一覧がまだありません。サーバーにツールがないという意味ではありません。対応機能はプロジェクトのドキュメントをご確認ください。

上のボタンでサーバーのエンドポイントに接続し、ツール一覧を取得できます。

概要

What is Opengraph.io?

Opengraph.io is a Model Context Protocol (MCP) server that exposes every OpenGraph.io API endpoint to AI agents such as Anthropic Claude, Cursor, and LangGraph. It acts as a secure bridge, allowing autonomous agents to unfurl links, scrape HTML, extract article text, capture screenshots, and generate images without exposing raw API keys.

How to use Opengraph.io?

Install globally via npm install -g opengraph-io-mcp or use the CLI installer with npx opengraph-io-mcp-install. Configure the server in your MCP client (e.g., Claude Desktop, Cursor, VS Code) using stdio transport and setting the APP_ID environment variable to your OpenGraph.io App ID. For HTTP/SSE transport, run npm start and point clients to http://localhost:3010/sse?app_id=YOUR_APP_ID.

Key features of Opengraph.io

  • Exposes all OpenGraph.io data endpoints as MCP tools
  • Includes AI-driven image generation (illustrations, diagrams, social cards, QR codes)
  • Supports multiple MCP clients: Claude Desktop, Cursor, Windsurf, VS Code, Zed, JetBrains
  • Offers both stdio and HTTP/SSE transport options
  • Provides iteration and export capabilities for generated images
  • Installs globally via npm or one‑click CLI installer

Use cases of Opengraph.io

  • AI agents automatically unfurl links and extract Open Graph metadata from URLs
  • Generate social‑media‑optimized images (OG cards, Twitter cards) directly from prompts
  • Create technical diagrams from Mermaid, D2, or Vega syntax within chat interfaces
  • Capture webpage screenshots and retrieve structured HTML extracts on demand
  • Query webpages with custom questions and receive structured answers

よくある質問

1

What is required to use this server?

An OpenGraph.io App ID is required. You can provide it via the `APP_ID` environment variable, a `.env` file, or a command‑line argument (`--app-id`).

2

Which transport options are supported?

Stdio transport (recommended) and HTTP/SSE transport. Stdio is configured for most MCP clients; SSE runs a persistent web server on port 3010.

3

Can I use it with Claude Desktop or Cursor?

Yes. Detailed configuration examples for Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, Zed, and JetBrains are provided in the README.

4

What types of images can be generated?

Illustrations, diagrams (Mermaid, D2, Vega), icons, social cards, and QR codes. Each type supports preset aspect ratios and style presets.

5

How do I iterate on or export a generated image?

Use the `Iterate Image` tool to modify an existing image by session and asset ID. Use `Export Image Asset` to retrieve the image inline as base64, optionally saving to disk when running locally.

コメント

「その他」の他のコンテンツ