home / mcp / scholarscope mcp server

ScholarScope MCP Server

Queries OpenAlex-backed scholarly data, fetches related works and full text where available, and exposes them via MCP for research assistants.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "eriknguyen20-scholarscope-mcp": {
      "command": "uv",
      "args": [
        "run",
        "--directory",
        "/path/to/mcp_server",
        "fastmcp",
        "run",
        "src/server.py"
      ],
      "env": {
        "OPENALEX_MAILTO": "[email protected]"
      }
    }
  }
}

ScholarScope MCP is a Python-based FastMCP server that lets you search academic literature via OpenAlex, explore related works and citations, and fetch full-text papers when available. It’s designed to power intelligent research assistants that can grow with your scholarly queries and provide actionable results directly through an MCP client.

How to use

You connect with an MCP client to ScholarScope MCP to perform fast, focused academic searches. Start the local server using your environment, then point your client to the running process. You can search for papers by keyword, author, or institution, explore related works and citations, and retrieve full text when sources permit.

Typical usage patterns include: - Initiating a search for papers by a specific keyword or author and sorting results by relevance or date. - Retrieving related works and citations for a chosen paper to broaden your literature review. - Requesting full-text access when available from configured sources.

How to install

Prerequisites: - Python (for running the server environment used in this setup) - uv and the FastMCP tooling for development and testing - An OpenAlex account email for attribution (required by the OpenAlex API usage policy)

Step-by-step setup: - Install the uv tool - Install required dependencies with uv sync - Set up the environment variable for OpenAlex attribution (OPENALEX_MAILTO) in a .env file

# Install uv if you don’t have it
pip install uv

# Install MCP-related dependencies
uv sync

# Create environment variable file
echo '[email protected]' > .env

Additional notes

Run the server via an MCP client or through an inspector tool to verify the setup locally. You’ll typically run the server with a command that launches the FastMCP-powered entry point, usually within a local project structure. Use the example command in your environment to confirm the server starts correctly.

Available tools

search_papers

Search papers by keyword, title, author, or institution and return results sorted by relevance, date, or other criteria.

get_related

Fetch related works and citation trees for a given paper to expand your literature network.

fetch_full_text

Retrieve full-text content from configured sources when available and permitted by rights.