Healthcare Public MCP Server

Access FDA drug databases, PubMed research, and ICD-10 medical terminology for AI analysis.

Data last scanned today · Reviewed today

399 tools125JavaScriptMIT

Overview

Healthcare Public MCP Server provides AI assistants with structured access to medical and healthcare data sources. It includes tools for retrieving FDA drug information, searching PubMed and medRxiv for literature, accessing Health.gov topics, and querying clinical trial databases.

Our verdict

This server receives a high-risk security score of 39/100, largely because it lacks read-only mode and lacks verified authentication mechanisms. While it offers useful medical research features, the lack of an officially maintained vendor and the absence of a read-only configuration means users should exercise caution, as all installations currently provide full write access potential. It is best suited for experimental environments rather than sensitive medical data pipelines.

  • Supports multiple interfaces including stdio and HTTP/SSE
  • Contains four direct dependencies to reduce supply-chain surface area
  • Includes a BMI medical calculator and DICOM metadata extraction
  • Provides interactive API documentation via Swagger UI
  • Requires full write access with no read-only mode available
  • Last code commit occurred over 350 days ago
  • Maintenance handled by community contributors rather than an official healthcare organization

Setup

Claude Desktop
{
  "mcpServers": {
    "healthcare-mcp-public": {
      "args": [
        "-y",
        "healthcare-mcp"
      ],
      "command": "npx"
    }
  }
}

Paste into ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows), then restart Claude Desktop. Some servers require API keys or tokens as environment variables — check the project's README and add an "env" object if needed.

  • No read-only mode — running this grants full read/write access.
  • No built-in authentication — anyone with access to this config can invoke every tool it exposes.

Tools

ToolDescriptionRisk
fda_drug_lookupSearch and retrieve comprehensive drug information from the FDA database.low
pubmed_searchSearch medical literature from PubMed's database of scientific articles.low
health_topicsAccess evidence-based health information from Health.gov.low
clinical_trials_searchSearch for ongoing and completed clinical trials.low
lookup_icd_codeLook up ICD-10 codes and medical terminology definitions.low
medrxiv_searchSearch for pre-print articles on medRxiv.low
calculate_bmiCalculate Body Mass Index (BMI).low
ncbi_bookshelf_searchSearch the NCBI Bookshelf for biomedical books and documents.low
extract_dicom_metadataExtract metadata from a DICOM file.low

Compatibility

ClientLocalDockerRemoteRead-only
Claude Desktop
Cursor
VS Code
Windsurf
ChatGPT

Frequently asked questions

Which communication interfaces does the server support?

The server supports both stdio for command-line interface usage and HTTP/SSE interfaces for web clients.

What is the recommended installation method?

The recommended installation method is using the provided healthcare-mcp.dxt file, which allows for one-click installation in compatible MCP clients like Claude Desktop.

Are there specific configuration requirements for the manual installation?

While not mandatory, you can set up optional environment variables by creating a .env file from the provided .env.example to include your own API keys.

How can I verify that the server tools are working correctly?

You can test the functionality of all tools using the built-in npm test script, or verify individual tools like FDA or PubMed lookups using specific test commands like npm run test:fda.

Does the server include features to manage API usage performance?

Yes, the server includes an efficient caching system with connection pooling to help reduce redundant API calls and improve overall performance.

Badge

Healthcare Public MCP Server security score, rated on RepoAI

Maintain this server? Add the live badge to your README.