Este repositorio contiene un servidor MCP que busca resúmenes de artículos en PubMed mediante el módulo Entrez de BioPython. Aprovecha el framework FastMCP para ofrecer funciones de búsqueda asíncrona en PubMed.
Buscar en PubMed: consulta artículos según un término de búsqueda.
Recuperar resúmenes: recupera resúmenes de artículos devueltos desde PubMed.
Operación asincrónica: utiliza la ejecución asincrónica (a través de asyncio.to_thread ) para evitar bloquear el servidor.
Related MCP server: PubMed MCP Server
Python 3.8 o superior
mcp[cli]
BioPython
Clonar el repositorio:
git clone https://github.com/AIAnytime/MCP-Servers
cd pubmed-mcp-serverDependencias de instalación:
Puede instalar los paquetes necesarios usando uv:
uv add -r requirements.txtConfigurar el correo electrónico de Entrez:
Asegúrese de haber configurado una dirección de correo electrónico válida en el código (en main.py ):
Entrez.email = "give an email address"Inicie el servidor PubMed MCP ejecutando:
uv run main.pyEste comando inicia el servidor utilizando la herramienta de línea de comandos uv (como se especifica en su configuración).
Para configurar su cliente MCP para que se conecte al servidor MCP de PubMed, cree o actualice su archivo config.json de la siguiente manera:
{
"mcpServers": {
"pubmed": {
"command": "C:/Users/aiany/.local/bin/uv",
"args": [
"--directory",
"C:/Users/aiany/OneDrive/Desktop/YT Video/pubmed-mcp-server",
"run",
"main.py"
]
}
}
}dominio:
La ruta completa a la herramienta de línea de comandos utilizada para ejecutar el servidor MCP (en este caso, uv ).
argumentos:
--directory : especifica el directorio de trabajo donde se encuentra el servidor.
"C:/Users/aiany/OneDrive/Desktop/YT Video/pubmed-mcp-server" : la ruta al directorio raíz del servidor.
"run" y "main.py" : el comando y el punto de entrada para iniciar el servidor PubMed MCP.
Una vez que el servidor esté en funcionamiento y su cliente MCP esté configurado, puede utilizar la herramienta proporcionada:
Herramienta: search_pubmed
Parámetros:
query : El término de búsqueda de PubMed (el valor predeterminado es "endocarditis" ).
max_results : Número máximo de artículos a recuperar (el valor predeterminado es 10 ).
Ejemplo de uso:
search_pubmed(query="endocarditis", max_results=10)Esto devolverá una cadena con los resúmenes de los artículos separados por nuevas líneas.
Este proyecto está licenciado bajo la licencia MIT .
Puede ajustar las rutas y los detalles según sea necesario para su configuración específica.
Search PubMed for articles matching the query.
Args:
query: The search term for PubMed.
max_results: Maximum number of articles to retrieve.
Returns:
A string containing the abstracts of found articles, separated by two newlines.
| Name | Required | Description | Default |
|---|---|---|---|
| max_results | No | ||
| query | No | endocarditis |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of behavioral disclosure. It mentions the tool 'searches' and 'retrieves' articles, implying a read-only operation, but doesn't clarify if it's safe, requires authentication, has rate limits, or details the search behavior (e.g., relevance ranking, filters). The return format is described, but key behavioral traits like error handling or performance are omitted.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is well-structured and front-loaded with the core purpose, followed by clear sections for Args and Returns. Every sentence adds value: the first states the purpose, and the others explain parameters and output. There's no redundancy or unnecessary information, making it highly efficient.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's moderate complexity (2 parameters, no output schema, no annotations), the description is minimally adequate. It covers the purpose, parameters, and return format, but lacks behavioral context like error handling, search constraints, or performance details. Without annotations or output schema, more completeness would be beneficial, but it meets basic requirements.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The description adds basic meaning for both parameters: 'query' is described as 'the search term for PubMed,' and 'max_results' as 'maximum number of articles to retrieve.' This compensates for the 0% schema description coverage by providing semantic context. However, it lacks details on query syntax, result limits, or default behaviors, keeping it at a baseline level.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's purpose: 'Search PubMed for articles matching the query.' This specifies the verb ('search'), resource ('PubMed articles'), and scope ('matching the query'). It's not a tautology and is unambiguous. However, with no sibling tools mentioned, there's no explicit differentiation from alternatives, preventing a perfect score.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides no guidance on when to use this tool versus alternatives. It lacks context about prerequisites, limitations, or scenarios where it's most appropriate. While it implies usage for searching PubMed articles, there's no explicit when/when-not advice or mention of other tools, leaving the agent with minimal usage direction.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v1.0.0search_pubmedScored across 1 tool
With only one tool, there is no possibility of confusion or overlap between tools. The tool 'search_pubmed' has a clear, singular purpose of searching PubMed for articles, making disambiguation trivial.
Since there is only one tool, naming consistency is inherently perfect. The tool name 'search_pubmed' follows a clear verb_noun pattern, and there are no other tools to compare it against for inconsistency.
A single tool for a PubMed server is too few for the apparent scope. PubMed typically involves operations like fetching article details, filtering by date or author, or retrieving citations, which are missing. This minimal set limits functionality and feels incomplete for the domain.
The tool set is severely incomplete for a PubMed server. While 'search_pubmed' covers basic searching, there are obvious gaps such as retrieving full article metadata, accessing citations, filtering results, or managing user queries. This will likely cause agent failures when more complex tasks are required.
Search 36M+ PubMed biomedical articles and ClinicalTrials.gov studies.
Search PubMed/Europe PMC, fetch articles and full text (PMC/EPMC/Unpaywall), citations, MeSH terms.
Search PubMed with precision using keyword and journal filters and smart sorting. Uncover MeSH ter…
Search biomedical literature, get article details, find related articles, and explore MeSH terms