How to add plugged.in MCP Proxy to Claude Desktop

Unified MCP hub that aggregates all your MCP servers into one connection, with knowledge, memory, and tools. Paste the config into ~/Library/Application Support/Claude/claude_desktop_config.json and restart Claude Desktop.

Last updated June 14, 2026 · 131★ · stdio · apikey · official

Claude Desktop config for plugged.in MCP Proxy

npx -y @pluggedin/pluggedin-mcp-proxy@latest --pluggedin-api-key YOUR_API_KEY

{
  "mcpServers": {
    "plugged-in-mcp-proxy": {
      "command": "npx",
      "args": [
        "-y",
        "@pluggedin/pluggedin-mcp-proxy@latest"
      ],
      "env": {
        "PLUGGEDIN_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

Setup steps

1Open Claude Desktop → Settings → Developer → Edit Config (this opens ~/Library/Application Support/Claude/claude_desktop_config.json).
2Paste the plugged.in MCP Proxy config below under the top-level "mcpServers" key.
3Fill in any placeholder secrets (API keys, paths) in the snippet.
4Save the file, then fully quit and reopen Claude Desktop.
5Open a chat and confirm plugged.in MCP Proxy's tools appear under the 🔌 tools menu.

Before you start

Node.js 18+ (v20+ recommended)
An API key from the plugged.in App (https://plugged.in/api-keys)

What plugged.in MCP Proxy can do in Claude Desktop

pluggedin_discover_tools

Smart discovery of available tools with caching for instant results; supports force_refresh and per-server discovery via server_uuid.

pluggedin_ask_knowledge_base

RAG search across your documents with AI filtering capabilities.

pluggedin_send_notification

Send notifications with optional email delivery.

pluggedin_create_document

Create AI-generated documents in your library.

pluggedin_list_documents

List documents with filtering options.

pluggedin_search_documents

Search for specific documents by query.

pluggedin_get_document

Retrieve a specific document's full content by ID.

pluggedin_update_document

Update or append to an existing document.

Security

Requires a PLUGGEDIN_API_KEY from the plugged.in App. The proxy validates commands against an allowlist (node, npx, python, python3, uv, uvx, uvenv), uses execFile() to avoid shell injection, sanitizes arguments and environment variables, and applies SSRF protection on URLs (blocking localhost, private IP ranges, and internal ports). In Streamable HTTP mode, enable --require-api-auth to require a Bearer token; rate limiting (60 tool calls/min, 100 API calls/min) and HTML/XSS sanitization are built in.

plugged.in MCP Proxy + Claude Desktop FAQ

Where is the Claude Desktop config file?

Claude Desktop reads MCP servers from ~/Library/Application Support/Claude/claude_desktop_config.json. Paste the plugged.in MCP Proxy config there under the "mcpServers" key and restart the client.

Is plugged.in MCP Proxy safe to use with Claude Desktop?

Is this an MCP server or just a proxy?

It is an MCP server that acts as a hub/proxy: it exposes its own built-in tools (discovery, RAG/knowledge, notifications, document management, clipboard memory) and also aggregates and routes to the tools of any downstream MCP servers you've configured in the plugged.in App.

What transports are supported?

It runs as STDIO by default and can also run as a Streamable HTTP server (default port 12006) with optional Bearer-token authentication (--require-api-auth) and a stateless mode. Downstream servers may use STDIO, SSE, or Streamable HTTP.

Do I need an API key?

Yes. A PLUGGEDIN_API_KEY from the plugged.in App is required; the proxy fetches server configurations and discovered capabilities from the plugged.in App's APIs.

View repo Full plugged.in MCP Proxy page