Skip to main content
Glama
emi-dm

PubMed MCP Server

by emi-dm
OverviewSchemaRelated ServersScoreDiscussions
Python
Remote

PubMed-MCP

A Model Context Protocol (MCP) server that provides tools for searching PubMed articles using the NCBI Entrez API.

Author: Emilio Delgado Muñoz

Features

  • Search PubMed for articles based on queries

  • Retrieve detailed information including title, authors, abstract, journal, and publication date

  • Returns results in JSON format

  • Configurable maximum number of results

Related MCP server: PubMed MCP Server

Architecture

graph TB
  A[User] --> B[MCP Server<br/>pubmed_server.py]
  B --> C[search_pubmed function]
  C --> D[Entrez.esearch<br/>Search in PubMed]
  D --> E[PubMed database<br/>NCBI]
  E --> F[List of PMIDs]
  F --> G[Entrez.efetch<br/>Fetch details]
  G --> E
  G --> H[Article XML records]
  H --> I[Data processing]
  I --> J[Extraction of:<br/>- Title<br/>- Authors<br/>- Abstract<br/>- Journal<br/>- Date]
  J --> K[List of articles<br/>in JSON format]
  K --> L[Response to user]

  subgraph "Dependencies"
    M[BioPython<br/>requirements.txt]
    N[FastMCP<br/>requirements.txt]
  end

  B -.-> M
  B -.-> N

  subgraph "Configuration"
    O[Entrez.email<br/>Configured in code]
  end

  C -.-> O

  style A fill:#e1f5fe
  style L fill:#c8e6c9
  style E fill:#fff3e0

Installation

  1. Clone this repository:

    git clone <repository-url>
cd PubMed-MCP

  2. Install dependencies:

    uv sync

  3. Configure your email in pubmed_server.py:

    Entrez.email = 'your-email@example.com'  # Replace with your actual email

VS Code Configuration

To use this MCP server locally in VS Code, the project includes a pre-configured .vscode/mcp.json file. This file tells VS Code how to run the MCP server.

The configuration is already set up to use uv for running the server:

{
  "servers": {
    "pubmed-mcp": {
      "command": "uv",
      "args": ["run", "${workspaceFolder}/pubmed_server.py"]
    }
  }
}

Requirements for VS Code Integration

  • VS Code with MCP extension support

  • uv package manager installed

  • Python virtual environment set up

Alternative Configuration

If you prefer to use pip instead of uv, you can modify the .vscode/mcp.json file:

{
  "servers": {
    "pubmed-mcp": {
      "command": "python",
      "args": ["${workspaceFolder}/pubmed_server.py"]
    }
  }
}

Make sure your virtual environment is activated when using this configuration.

Requirements

  • Python 3.11+

  • BioPython

  • FastMCP

Usage

Run the MCP server:

python pubmed_server.py

The server will start and listen for MCP protocol messages on stdin/stdout.

Available Tools

search_pubmed

Searches PubMed for articles matching the given query.

Parameters:

  • query (string): The search query

  • max_results (integer, optional): Maximum number of results to return (default: 10)

  • title (bool, optional): If true (default) search in Title field

  • abstract (bool, optional): If true (default) search in Abstract field

  • keywords (bool, optional): If true (default) expand search with Author Keywords ([ot]) and MeSH Headings ([mh])

Field logic:

  • title=True and abstract=True -> query applied as (your terms)[tiab]

  • Only title=True -> (your terms)[ti]

  • Only abstract=True -> (your terms)[ab]

  • Both false -> no field tag (all fields)

  • keywords=True -> OR-expanded with (your terms)[ot] OR (your terms)[mh]

Example refined queries:

query = "breast cancer metastasis"
title=True, abstract=True, keywords=True -> (breast cancer metastasis)[tiab] OR ((breast cancer metastasis)[ot] OR (breast cancer metastasis)[mh])
title=True, abstract=False, keywords=False -> (breast cancer metastasis)[ti]
title=False, abstract=False, keywords=True -> (breast cancer metastasis) OR ((breast cancer metastasis)[ot] OR (breast cancer metastasis)[mh])

Returns: A list of article objects containing:

  • pmid: PubMed ID

  • title: Article title

  • authors: List of author names

  • abstract: Article abstract

  • journal: Journal name

  • publication_year: Year of publication

  • publication_month: Month of publication

  • url: PubMed URL

Configuration

Before using the tool, you must set your email address in the Entrez.email variable. This is required by NCBI's Entrez API.

License

This project is open source. Please check the license file for details.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Available Tools

1 tool
search_pubmedA

Search PubMed and return a list of article JSON objects.

Parameters: query: Free-text user query; boolean operators (AND/OR/NOT) supported by PubMed. max_results: Maximum number of records to retrieve (retmax). title: If True, include Title field in search restriction (ti / tiab). abstract: If True, include Abstract field in search restriction (ab / tiab). keywords: If True, expand search to Author Keywords (ot) and MeSH Headings (mh).

Field logic: - title and abstract both True => core search uses [tiab] - only title True => uses [ti] - only abstract True => uses [ab] - neither title nor abstract True => no restriction (all fields) - keywords True => additionally OR with [ot] and [mh] versions of the query

Returns: List[dict]: Each dict contains pmid, title, authors, abstract, journal, publication_year, publication_month, url.

ParametersJSON Schema
NameRequiredDescriptionDefault
queryYes
max_resultsNo
titleNo
abstractNo
keywordsNo

TDQS

A4.2/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description carries full burden. It effectively describes the search behavior, field logic, and return format, though it lacks details on rate limits, authentication needs, or error handling.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is well-structured with clear sections (Parameters, Field logic, Returns) and avoids redundancy. It could be slightly more concise by integrating some details, but overall it's efficient and front-loaded.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 5-parameter search tool with no annotations or output schema, the description is quite complete—covering parameters, logic, and return values. Minor gaps include lack of error handling or performance details, but it's largely sufficient.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 0%, so the description must compensate. It thoroughly explains all 5 parameters, including their purposes, defaults, and interactions (e.g., field logic for title/abstract/keywords), adding significant value beyond the bare schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the specific action ('Search PubMed') and resource ('article JSON objects'), with no siblings to distinguish from. It provides a complete picture of what the tool does.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description implies usage for searching PubMed articles but provides no explicit guidance on when to use this tool versus alternatives, prerequisites, or error conditions. With no sibling tools, this is less critical but still a gap.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.

  1. 1 tool update
    • First observedsearch_pubmed
Install Server
RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

TDQS

A4/5.0
Disambiguation5/5

With only one tool, there is no possibility of ambiguity or overlap between tools. The single tool 'search_pubmed' has a clearly defined and distinct purpose for searching PubMed articles.

Naming Consistency5/5

Since there is only one tool, naming consistency is inherently perfect. The tool name 'search_pubmed' follows a clear verb_noun pattern that would be appropriate if more tools were added.

Tool Count2/5

A single tool for a PubMed server is too minimal for the apparent scope. While search is a core function, a complete PubMed interface would typically include tools for fetching article details, citations, related articles, or filtering by metadata. One tool feels thin and incomplete.

Completeness2/5

The tool surface is severely incomplete for a PubMed domain. It only provides search functionality, missing essential operations like retrieving specific articles by PMID, fetching citations, accessing related articles, or filtering by date/journal. This will cause significant agent failures when trying to perform comprehensive PubMed tasks.

Maintenance

ActivityInactive
ResponsivenessSyncing

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

Related MCP Servers

  • PubMed MCP Server

    openpharma-org
    A
    license
    B
    quality
    D
    maintenance
    Enables AI assistants to search and retrieve biomedical research articles from PubMed's database of over 35 million citations, including metadata, abstracts, MeSH terms, and full-text PDFs when available.
    1
    1
    MIT

  • PubMed MCP Server

    aeghnnsw
    F
    license
    Not graded
    quality
    C
    maintenance
    Enables searching PubMed's biomedical literature database and retrieving article metadata, abstracts, and full content through the E-utilities API. Supports advanced queries, batch operations, and multiple output formats with automatic rate limiting.
    1
    -

  • NCBI Literature Search MCP

    vitorpavinato
    F
    license
    Not graded
    quality
    D
    maintenance
    Provides seamless access to over 35 million PubMed scientific articles through natural language queries for research discovery and analysis. It enables tools for advanced searches, retrieving article details, and exploring related research within the life sciences and biomedical fields.
    12
    -

Appeared in Searches

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/emi-dm/PubMed-MCP'

If you have feedback or need assistance with the MCP directory API, please join our Discord server