Medical MCP Server (India-Optimized)

A Model Context Protocol (MCP) server that provides comprehensive medical information by querying multiple authoritative medical APIs including FDA, WHO, PubMed, and RxNorm. Optimized for India with default country preferences and India-specific medical research context.

🌐 Globally Available: This package is published on npm and can be used by anyone worldwide with a single command - no cloning, no building, no complex setup!

🎯 Use Anywhere with npx

npx -y india_mcp

That's it! No installation needed. The package will be automatically downloaded from npm and run.

📦 Installation

Install Globally from npm

npm install -g india_mcp

Or Use with npx (No Installation Required)

npx india_mcp

🚀 Quick Start

Using with MCP Clients (Claude Desktop, Cline, etc.)

Add this configuration to your MCP settings file:

Claude Desktop Configuration

Windows: %APPDATA%\Claude\claude_desktop_config.json
macOS/Linux: ~/Library/Application Support/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "india_mcp": {
      "command": "npx",
      "args": ["-y", "india_mcp"]
    }
  }
}

Cline (VS Code Extension) Configuration

Windows: %APPDATA%\Code\User\globalStorage\saoudrizwan.claude-dev\settings\cline_mcp_settings.json
macOS/Linux: ~/Library/Application Support/Code/User/globalStorage/saoudrizwan.claude-dev/settings/cline_mcp_settings.json

{
  "mcpServers": {
    "india_mcp": {
      "command": "npx",
      "args": ["-y", "india_mcp"]
    }
  }
}

Alternative: Using Globally Installed Package

If you've installed globally with npm install -g india_mcp:

{
  "mcpServers": {
    "india_mcp": {
      "command": "node",
      "args": ["<path-to-global-npm>/node_modules/india_mcp/build/index.js"]
    }
  }
}

Find your global npm path:

• Windows: %APPDATA%\npm\node_modules\india_mcp\build\index.js
• macOS/Linux: Run npm root -g to find the path

Direct Command Line Usage

# Default stdio mode (for MCP clients)
india_mcp

# HTTP mode (for testing)
india_mcp --http

# HTTP mode with custom port
india_mcp --http --port=3000

🇮🇳 India-First Configuration

This server is configured with India as the default country preference:

• Health Statistics: Automatically defaults to India (IND) for WHO health indicators
• Medical Literature: Option to include India-specific research context in PubMed searches (enabled by default)
• Clinical Guidelines: Includes Indian medical organizations (ICMR, IMA, etc.) in guideline searches
• Regional Focus: Prioritizes India-relevant medical data while maintaining global access

You can always override these defaults by specifying different country codes or disabling India context.

🔒 Security Features

Localhost-Only Binding: This server is configured for maximum security with localhost-only access:

• Stdio Mode (Default): Inherently localhost-only process communication
• HTTP Mode: Binds to 127.0.0.1 only, blocks all external connections
• IP Filtering: Validates all incoming connections against localhost addresses
• CORS Restrictions: Only allows localhost origins
• Security Logging: All blocked connection attempts are logged

See SECURITY.md for detailed security configuration.

Features

This MCP server offers specialized tools for querying medical information from reliable sources:

💊 Drug Information Tools

search-drugs

Search for drug information using the FDA database.

Input:

• query (string): Drug name to search for (brand name or generic name)
• limit (optional, number): Number of results to return (1-10, default: 5)

Output:

• Drug information including brand name, generic name, manufacturer, route, dosage form, and purpose

Example:

Drug Search Results for "Advil"

Found 1 drug(s)

1. **ADVIL**
   Generic Name: IBUPROFEN
   Manufacturer: PFIZER CONSUMER HEALTHCARE
   Route: ORAL
   Dosage Form: TABLET
   Purpose: For temporary relief of minor aches and pains...
   Last Updated: 20210902

get-drug-details

Get detailed information about a specific drug by NDC (National Drug Code).

Input:

• ndc (string): National Drug Code (NDC) of the drug

Output:

• Comprehensive drug information including warnings, drug interactions, and clinical pharmacology

📊 Health Statistics Tools

get-health-statistics

Get health statistics and indicators from WHO Global Health Observatory (defaults to India).

Input:

• indicator (string): Health indicator to search for (e.g., 'Life expectancy', 'Mortality rate')
• country (optional, string): Country code (e.g., 'IND', 'USA', 'GBR') - defaults to 'IND' (India)
• limit (optional, number): Number of results to return (1-10, default: 5)

Output:

• Health statistics with values, ranges, and temporal data

Example:

Health Statistics: Life expectancy at birth (years)

Country: IND
Found 5 data points

1. **IND** (2019)
   Value: 69.7 years
   Numeric Value: 69.7
   Date: 2019-12-31

🔬 Medical Literature Tools

search-medical-literature

Search for medical research articles in PubMed (India-focused by default).

Input:

• query (string): Medical topic or condition to search for
• max_results (optional, number): Maximum number of articles to return (1-10, default: 5)
• include_india_context (optional, boolean): Include India-specific research context in search (default: true)

Output:

• Medical research articles with titles, PMIDs, journals, and publication dates

Note: By default, searches include India-specific context to prioritize research relevant to Indian healthcare. Set include_india_context to false for global-only results.

Example:

Medical Literature Search: "diabetes treatment"

Found 5 article(s)

1. **Novel Approaches to Diabetes Management**
   PMID: 12345678
   Journal: New England Journal of Medicine
   Publication Date: 2024-01-15

search-google-scholar

Search for academic research articles using Google Scholar.

Input:

• query (string): Academic topic or research query to search for

Output:

• Academic research articles with titles, authors, abstracts, journals, years, citations, and URLs

Example:

Google Scholar Search: "machine learning healthcare"

Found 5 article(s)

1. **Machine Learning in Healthcare: A Systematic Review**
   Authors: Smith J, Johnson A - Journal of Medical AI
   Year: 2023
   Citations: Cited by 45
   URL: https://scholar.google.com/...
   Abstract: This systematic review examines the application of machine learning...

Note: This tool uses web scraping to access Google Scholar since it doesn't provide a public API. It includes rate limiting protection and stealth measures to avoid detection.

🏥 Drug Nomenclature Tools

search-drug-nomenclature

Search for drug information using RxNorm (standardized drug nomenclature).

Input:

• query (string): Drug name to search for in RxNorm database

Output:

• Standardized drug information with RxCUI codes, synonyms, and term types

🩺 Clinical Guidelines Tools

search-clinical-guidelines

Search for clinical guidelines and practice recommendations from medical organizations (includes Indian medical organizations by default).

Input:

• query (string): Medical condition or topic to search for guidelines
• organization (optional, string): Specific medical organization to filter by (e.g., 'Indian Council of Medical Research', 'ICMR', 'WHO', 'American Heart Association')
• include_india_context (optional, boolean): Include Indian medical organizations and context in search (default: true)

Output:

• Clinical guidelines with organization, year, category, evidence level, and URLs

Supported Indian Organizations:

• Indian Council of Medical Research (ICMR)
• Ministry of Health and Family Welfare India
• Indian Medical Association (IMA)
• Cardiological Society of India
• Indian Academy of Pediatrics
• Association of Physicians of India
• Endocrine Society of India
• Indian Psychiatric Society

Note: By default, searches prioritize Indian medical guidelines and organizations. Set include_india_context to false for global-only results.

💊 Drug Interaction Tools

check-drug-interactions

Check for potential drug-drug interactions between two medications.

Input:

• drug1 (string): First drug name
• drug2 (string): Second drug name

Output:

• Potential drug interactions, severity levels, and recommendations

Installation

1. Clone this repository:

git clone <repository-url>
cd medical-mcp

2. Install dependencies:

npm install

3. Build the project:

npm run build

Usage

Running the Server

Start the MCP server:

npm start

The server runs on stdio and can be connected to any MCP-compatible client.

Example Queries

Here are some example queries you can make with this MCP server:

Search for Drug Information

{
  "tool": "search-drugs",
  "arguments": {
    "query": "Tylenol",
    "limit": 5
  }
}

Get Drug Details by NDC

{
  "tool": "get-drug-details",
  "arguments": {
    "ndc": "00071015527"
  }
}

Get Health Statistics

{
  "tool": "get-health-statistics",
  "arguments": {
    "indicator": "Life expectancy at birth (years)",
    "country": "IND",
    "limit": 5
  }
}

Note: The country parameter defaults to 'IND' (India), so you can omit it to get India's statistics:

{
  "tool": "get-health-statistics",
  "arguments": {
    "indicator": "Life expectancy at birth (years)",
    "limit": 5
  }
}

Search Medical Literature

{
  "tool": "search-medical-literature",
  "arguments": {
    "query": "COVID-19 treatment",
    "max_results": 10
  }
}

Search Drug Nomenclature

{
  "tool": "search-drug-nomenclature",
  "arguments": {
    "query": "aspirin"
  }
}

API Endpoints

This MCP server integrates with the following medical APIs:

FDA API

• GET /drug/label.json - Drug labeling information
• Search by brand name, generic name, or NDC
• Provides safety information, warnings, and clinical data

WHO Global Health Observatory API

• GET /api/Indicator - Health statistics and indicators
• Global health data with country-specific information
• Temporal data for trend analysis

PubMed API

• GET /esearch.fcgi - Search for medical articles
• GET /efetch.fcgi - Retrieve article details
• Access to millions of medical research papers

RxNorm API

• GET /REST/drugs.json - Standardized drug nomenclature
• Drug name standardization and relationships
• Clinical drug information

Google Scholar (Web Scraping)

• Web scraping of Google Scholar search results
• Academic research article discovery
• Citation and publication information
• Note: Uses Puppeteer for browser automation with anti-detection measures

Data Sources

FDA (Food and Drug Administration)

• Source: Official FDA drug labeling database
• Coverage: All FDA-approved drugs in the United States
• Data: Drug safety, efficacy, dosage, warnings, and interactions
• Update Frequency: Real-time as drugs are approved or labeling changes

WHO (World Health Organization)

• Source: Global Health Observatory database
• Coverage: Global health statistics from 194 countries
• Data: Life expectancy, mortality rates, disease prevalence, and health indicators
• Update Frequency: Annual updates with historical data

PubMed (National Library of Medicine)

• Source: MEDLINE database of medical literature
• Coverage: Over 30 million citations from medical journals
• Data: Research articles, clinical studies, and medical reviews
• Update Frequency: Daily updates as new articles are published

RxNorm (National Library of Medicine)

• Source: Standardized drug nomenclature system
• Coverage: Clinical drugs available in the United States
• Data: Drug names, codes, relationships, and clinical information
• Update Frequency: Weekly updates

Google Scholar (Web Scraping)

• Source: Google Scholar academic search engine
• Coverage: Academic papers, theses, books, and abstracts across all disciplines
• Data: Research articles, citations, authors, journals, and publication dates
• Update Frequency: Real-time as new papers are indexed
• Note: Access via web scraping with rate limiting protection

Error Handling

The server includes comprehensive error handling:

• Network errors are caught and reported with descriptive messages
• Invalid queries return appropriate error messages
• Rate limiting and API errors are handled gracefully
• Fallback responses when specific APIs are unavailable

Web Scraping Implementation

The Google Scholar integration uses Puppeteer for web scraping with the following features:

Anti-Detection Measures

• Stealth Mode: Browser launched with multiple flags to avoid detection
• User Agent Spoofing: Realistic browser user agent strings
• Random Delays: Built-in delays between requests to avoid rate limiting
• Header Spoofing: Realistic HTTP headers to appear as a regular browser
• Viewport Settings: Standard desktop viewport dimensions

Robust Parsing

• Multiple Selectors: Uses various CSS selectors to handle different Google Scholar layouts
• Fallback Strategies: Multiple parsing approaches for different page structures
• Error Recovery: Graceful handling of missing elements or changed page structures
• Data Validation: Filters out incomplete or invalid results

Rate Limiting Protection

• Random Delays: 1-3 second random delays between requests
• Browser Management: Proper browser cleanup to prevent resource leaks
• Timeout Handling: Configurable timeouts for network requests
• Error Recovery: Automatic retry logic for failed requests

🚀 Usage

Stdio Mode (Default - Most Secure)

# Build and run in stdio mode (inherently localhost-only)
npm run build
npm start

# Or directly
node build/index.js

HTTP Mode (Localhost-Only)

# HTTP server on localhost only (port 3000)
npm run start:http

# Custom port
npm run start:http:port
# or
node build/index.js --http --port=8080

# Test localhost access
curl http://localhost:3000/info

Development Mode

# Build and run stdio
npm run dev

# Build and run HTTP
npm run dev:http

🤖 Claude Desktop Integration

Setup Instructions

1. Install and Build the Server

# Clone the repository
git clone https://github.com/JamesANZ/medical-mcp.git
cd medical-mcp

# Install dependencies
npm install

# Build the server
npm run build

2. Configure Claude Desktop

Create or edit your Claude Desktop configuration file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "medical-mcp": {
      "command": "node",
      "args": ["/path/to/medical-mcp/build/index.js"],
      "cwd": "/path/to/medical-mcp"
    }
  }
}

Replace the paths with your actual installation directory:

For users with NVM (Node Version Manager):

{
  "mcpServers": {
    "medical-mcp": {
      "command": "/Users/yourusername/.nvm/versions/node/v22.17.0/bin/node",
      "args": [
        "/Users/yourusername/Documents/projects/medical-mcp/build/index.js"
      ]
    }
  }
}

For users with system Node.js:

{
  "mcpServers": {
    "medical-mcp": {
      "command": "node",
      "args": ["/Users/yourusername/medical-mcp/build/index.js"],
      "cwd": "/Users/yourusername/medical-mcp"
    }
  }
}

To find your Node.js path:

# If using NVM
which node
# Example output: /Users/yourusername/.nvm/versions/node/v22.17.0/bin/node

# If using system Node.js
which node
# Example output: /usr/local/bin/node

3. Restart Claude Desktop

• Close Claude Desktop completely
• Reopen Claude Desktop
• The Medical MCP Server will start automatically

Available Medical Tools in Claude

Once connected, you'll have access to these medical tools:

🔍 Search Tools

• search-drugs - Search FDA drug database
• search-pubmed-articles - Search medical literature
• search-google-scholar - Search academic research
• search-medical-databases - Comprehensive multi-database search
• search-medical-journals - Search top medical journals

💊 Drug Information Tools

• get-drug-by-ndc - Get drug details by NDC code
• check-drug-interactions - Check drug interactions
• get-drug-safety-info - Get drug safety information

🏥 Clinical Tools

• generate-differential-diagnosis - Generate differential diagnoses
• get-diagnostic-criteria - Get diagnostic criteria for conditions
• get-risk-calculators - Get clinical risk calculators
• get-lab-values - Get normal lab value ranges

📊 Health Data Tools

• get-health-indicators - Get WHO health statistics
• search-rxnorm-drugs - Search RxNorm drug database
• search-clinical-guidelines - Search clinical guidelines

Example Claude Conversations

Drug Information Query

User: "What are the side effects of metformin and can it interact with lisinopril?"

Claude will use:
- search-drugs for metformin information
- check-drug-interactions for metformin + lisinopril
- get-drug-safety-info for detailed safety data

Clinical Decision Support

User: "A 45-year-old patient presents with chest pain, shortness of breath, and diaphoresis. What should I consider?"

Claude will use:
- generate-differential-diagnosis for possible conditions
- get-diagnostic-criteria for specific diagnostic criteria
- search-medical-databases for latest research

Research and Literature Review

User: "Find recent research on COVID-19 treatment protocols"

Claude will use:
- search-pubmed-articles for medical literature
- search-medical-journals for top-tier research
- search-google-scholar for additional academic sources

🔒 Security Features in Claude

• Localhost-Only: Server runs locally, no external access
• Process Isolation: Medical data stays on your machine
• No Data Storage: No medical data is stored locally
• Dynamic Data: All information retrieved in real-time
• Audit Logging: All queries are logged for transparency

Troubleshooting

Server Won't Start

# Check if the server builds correctly
npm run build

# Test the server directly
node build/index.js

# Check for port conflicts
lsof -i :3000

Claude Can't Connect

1. Verify the configuration file path is correct
2. Ensure the server executable path is absolute
3. Check that Node.js is in your PATH
4. Restart Claude Desktop after configuration changes

Permission Issues

# Make sure the executable has proper permissions
chmod 755 build/index.js

# On macOS, you might need to allow Node.js in Security & Privacy

Network Issues

• The server uses localhost-only binding for security
• No external network access required for the server itself
• Medical data is fetched from external APIs when needed

Best Practices

For Medical Professionals

• Always verify information through multiple sources
• Use as a starting point for research, not final decisions
• Follow established clinical guidelines
• Document your decision-making process

For Students and Researchers

• Use for literature reviews and research
• Cross-reference with primary sources
• Understand the limitations of AI-generated summaries
• Always cite original sources

For General Users

• Use for educational purposes only
• Never replace professional medical advice
• Consult healthcare providers for medical decisions
• Be aware of information limitations

Advanced Configuration

Custom Port (if needed)

{
  "mcpServers": {
    "medical-mcp": {
      "command": "node",
      "args": ["/path/to/medical-mcp/build/index.js", "--http", "--port=3001"],
      "cwd": "/path/to/medical-mcp"
    }
  }
}

Environment Variables

# Set custom API timeouts
export MCP_TIMEOUT=30000

# Set custom user agent
export MCP_USER_AGENT="Medical-Research-Tool/1.0"

🔒 Security Verification

Test Localhost Access

# Should work (localhost)
curl http://localhost:3000/info

# Should be blocked (external IP)
curl http://YOUR_EXTERNAL_IP:3000/info
# Returns: "Access denied: This server is restricted to localhost only"

Check Binding

# Verify server is bound to localhost only
netstat -an | grep :3000
# Should show: 127.0.0.1:3000 (not 0.0.0.0:3000)

Medical Disclaimer

Important: This MCP server provides information from authoritative medical sources but should not be used as a substitute for professional medical advice, diagnosis, or treatment. Always consult with qualified healthcare professionals for medical decisions.

• The information provided is for educational and informational purposes only
• Drug information may not be complete or up-to-date for all medications
• Health statistics are aggregated data and may not reflect individual circumstances
• Medical literature should be interpreted by qualified healthcare professionals

🌐 Access from Anywhere

This MCP server is published on npm and can be accessed globally by anyone:

npm Package URL

• Package: https://www.npmjs.com/package/india_mcp
• Install: npm install -g india_mcp
• Run: india_mcp or npx india_mcp

Using in Your Projects

Anyone worldwide can use this MCP server by adding it to their MCP client configuration. The server will be automatically downloaded from npm when first accessed.

Example for any MCP client:

{
  "mcpServers": {
    "india_mcp": {
      "command": "npx",
      "args": ["-y", "india_mcp"]
    }
  }
}

For Developers

If you want to integrate this server into your applications:

import { spawn } from 'child_process';

// Spawn the MCP server
const mcpServer = spawn('npx', ['-y', 'india_mcp']);

// Handle stdio communication
mcpServer.stdout.on('data', (data) => {
  console.log('Server output:', data.toString());
});

📊 Package Statistics

• Package Name: india_mcp
• Latest Version: Check npm
• Downloads: Updated weekly on npm
• Author: surendirakrishna

Dependencies

• @modelcontextprotocol/sdk - MCP SDK for server implementation
• superagent - HTTP client for API requests
• puppeteer - Browser automation for web scraping Google Scholar
• zod - Schema validation for tool parameters

Contributing

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

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

Support

If you encounter any issues or have questions:

• GitHub Issues: Report an issue
• npm Package: https://www.npmjs.com/package/india_mcp
• Email: Contact the maintainer through npm profile

License

This project is licensed under the ISC License - see the LICENSE.md file for details.