Scopus MCP server

Scopus MCP Server

<!-- mcp-name: io.github.qwe4559999/scopus-mcp -->

中文 | English

💡 Check out the Interaction Guide & Prompt Examples to see how to chat with this tool!

This is a Model Context Protocol (MCP) server that provides access to the Elsevier Scopus API. It allows AI assistants to search for academic papers, retrieve abstracts, and look up author profiles.

Please note that requesting an Elsevier Scopus API key generally requires that your organization or institution has a subscription to Elsevier database services. Additionally, to run this tool without manual setup, your device must have the uv package manager installed.

Configuration

Setup Steps

1. Go to Elsevier Developer Portal to apply for an API key.
2. Create a config.json file in the project root (or copy from config.json.example) and fill in your key:

{
      "api_key": "YOUR_KEY_HERE"
    }

1. Edit MCPtoolconfig.json, modifying the folder path (pay attention to the slash direction).
2. Finally, import the configuration into your MCP client (e.g., Claude Desktop) by copying the content of MCPtoolconfig.json.

🚀 Quick Start (Zero Setup)

Prerequisite: You must have uv installed.

• Windows: powershell -c "irm https://astral.sh/uv/install.ps1 | iex"
• macOS/Linux: curl -LsSf https://astral.sh/uv/install.sh | sh

If you use Claude Desktop, you can skip downloading the code and just configure it directly:

1. Get Key: Get a free API Key from Elsevier Developer Portal. (⚠️ Note: Educational/Institutional email is recommended; public email domains may be rejected).
2. Configure: Edit %APPDATA%\Claude\claudedesktopconfig.json (Windows) or ~/Library/Application Support/Claude/claudedesktopconfig.json (macOS).
3. Add:

{
  "mcpServers": {
    "scopus-assistant": {
      "command": "uvx",
      "args": [
        "scopus-mcp"
      ],
      "env": {
        "SCOPUS_API_KEY": "PUT_YOUR_KEY_HERE"
      }
    }
  }
}

(If you don't have uv, see Installation for manual setup)

Using with Trae

In Trae Settings -> MCP Servers -> Click Add -> Select Manual Configuration (JSON), then paste:

{
  "mcpServers": {
    "scopus-assistant": {
      "command": "uvx",
      "args": [
        "scopus-mcp"
      ],
      "env": {
        "SCOPUS_API_KEY": "PUT_YOUR_KEY_HERE"
      }
    }
  }
}

Using with Cursor

1. Open Cursor Settings -> Features -> MCP Servers.
2. Click + Add New MCP Server.
3. Fill in the details:

• Name: scopus-mcp
• Type: command (stdio)
• Command: uvx scopus-mcp

1. Important: You need to set SCOPUSAPIKEY in your system environment variables.

Installation

1. Ensure you have Python 3.10+ installed.
2. Install dependencies:

pip install .

Usage

Running the Server

You can run the server using uvx (recommended) or directly with python.

# Using uvx
uvx --from . scopus-mcp

# Or directly
python -m scopus_mcp.server

Available Tools

1. searchscopus**

• Searches the Scopus database using the standard query syntax.
• Arguments:
• query (string): The search query (e.g., TITLE("Artificial Intelligence")).
• count (integer): Number of results to return (default: 5).
• sort (string): Sort order (e.g., coverDate).

1. getabstractdetails

• Retrieves detailed information for a specific document.
• Arguments:
• scopus_id (string): The Scopus ID of the document.

1. getauthorprofile

• Retrieves an author's profile information.
• Arguments:
• author_id (string): The Scopus Author ID.

Development

Run tests with:

pytest

License

Shortened. The full README is on GitHub.