PubMedSearch

Provides tools to search PubMed and retrieve detailed paper information via MCP protocol.

python

6

GitHub Stars

python

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "gradusnikov-pubmed-search-mcp-server": {
      "command": "mcp",
      "args": [
        "dev",
        "pubmed_search_mcp_server.py"
      ]
    }
  }
}

PubMedSearch MCP Server is a Model Content Protocol server that enables you to search PubMed and retrieve detailed information about academic papers. It provides practical tools to query the PubMed database and obtain metadata such as title, authors, journal, abstract, DOI, and more, making it easy to integrate scholarly data into your MCP workflows.

How to use

You interact with PubMedSearch MCP Server through an MCP client. Start a local or development server, then issue search requests by providing keywords or author names. You can retrieve detailed results for individual papers, including the title, authors, journal, abstract, and DOI. Use the server to build workflows that discover relevant literature and fetch complete paper details for downstream tasks like summarization, annotation, or bibliography creation.

Practical usage patterns you can follow:

Start a development server and test search queries from your MCP client.
Submit keyword searches to find papers by title or abstract content.
Search by author names to locate works from specific researchers.
Retrieve full metadata for targeted papers to populate databases or reference lists.

How to install

Prerequisites you need before installation:
- Node.js and npm or npx for client-side tooling (for installation via Smithery).
- Python 3 and pip for manual setup and dependencies.
- A working MCP client to connect to the server.

npx -y @smithery/cli install @gradusnikov/pubmed-search-mcp-server --client claude

Manual installation steps you can follow:

Clone the project repository
Install dependencies
Start the MCP server in development mode or configure the client to connect to it

git clone <repository-url>
cd pubmed-search-mcp-server

pip install fastmcp requests python-dotenv

Create a configuration file if you need to customize settings. You can place environment variables in a .env file at the project root to configure the server as needed.

Start the server in development mode from the project root:

mcp dev pubmed_search_mcp_server.py

Alternatively, you can add the server to your Claude Desktop configuration. For example, if you are using Windows Subsystem for Linux (WSL), you may set it up like this in your CLAUDE_DIRECTORY/claude_desktop_config.json file:

"pubmed-search": {
            "command": "wsl.exe",
            "args": [
                "bash",
                "-c",
                "source /home/[user]/anaconda3/etc/profile.d/conda.sh && conda activate mcp && mcp run /home/[user]/pubmed-search-mpc-server/pubmed_search_mcp_server.py"
            ]
        },

Additional notes and troubleshooting

A minimal environment file (.env) can be created in the project root if you need to configure runtime options for the server. When running in development mode, you can modify environment variables to enable debugging or to point the server at test data sources.

Available tools

search

Queries PubMed by keywords in title/abstract or author names to find relevant papers.

retrieve

Fetches detailed metadata for papers, including title, authors, journal, abstract, and DOI.