gauravnadkarni/google-drive-xlsx-health-data-mcp-server

Author: Gaurav Nadkarni nadkarnigaurav@gmail.com
Repository: https://github.com/gauravnadkarni/google-drive-xlsx-health-data-mcp-server.git

Google Drive Health Data MCP Server

A Model Context Protocol (MCP) server for accessing and analyzing health metrics data stored in Google Drive spreadsheets.

📖 Related Article

📚 Read the accompanying Medium article: What I Learned Building My First MCP Server From Scratch

Learn about the architecture, implementation details, and real-world use cases of this MCP server in the comprehensive technical write-up.

Features

• Health Data Integration: Connects to Google Drive to access health metrics data
• Multiple Data Views: Supports daily, weekly, monthly, and seasonal health metrics
• Comprehensive Metrics: Tracks various health indicators including:
  • Activity data
  • Heart rate
  • Sleep patterns
  • Nutrition information
  • User profile data

Prerequisites

• Node.js (v16 or later)
• npm or yarn
• Google Cloud Platform project with Google Drive API enabled
• Service account credentials for Google Drive API
• Health data google sheet file (Sample file is in the ./uploads directory)

Installation

Clone the repository:

git clone https://github.com/gauravnadkarni/google-drive-xlsx-health-data-mcp-server.git
cd google-drive-xlsx-health-data-mcp-server

Install dependencies:

npm install

MCP server can be run in two modes:

1. STDIO Mode: Local process communication, ideal for single-user desktop applications.

2. HTTP Mode: Remote server access, supports multiple clients and cloud deployment.

Building the Project

npm run build

Running the Server

To run the server in STDIO mode:

MCP_TRANSPORT=stdio npm run start

To run the server in HTTP mode:

MCP_TRANSPORT=http npm run start

Running the Project in dev mode

npm run dev

Required environment configurations:

GOOGLE_DRIVE_FILE_ID=
GOOGLE_CREDENTIALS_PATH=
PORT="3002"
MCP_TRANSPORT="stdio" #http or stdio
DOTENV_CONFIG_QUIET="true" # Set to true to suppress dotenv warnings (optional for http mode but required for stdio mode)

Important Notes

• You should create a project in google console and enable google drive api.
• Create service account and download credentials json file and place it in the root directory of the project.
• Upload the health data file to drive and share it with the email id of the service account.
• For STDIO mode, you should set the GOOGLE_DRIVE_FILE_ID and GOOGLE_CREDENTIALS_PATH environment variables in the client application (example claude code mcp json file or the MCP inspector)
• For HTTP mode, you should set the GOOGLE_DRIVE_FILE_ID and GOOGLE_CREDENTIALS_PATH environment variables in the server application (example .env file)
• STDIO mode is the default mode and in case no MCP_TRANSPORT environment variable is set, it will run in STDIO mode.

Google Cloud Setup

• Create a project in Google Cloud Console
• Enable the Google Drive API
• Create a Service Account and download the credentials JSON file
• Place the credentials file in your project root as credentials.json

Spreadsheet Setup

• Upload the sample health data file (from ./uploads directory) to your Google Drive
• Share the file with your service account email (viewer permissions)
• Copy the File ID from the Google Drive URL

Claude Desktop (STDIO Mode) Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "health-data": {
      "command": "node",
      "args": ["/absolute/path/to/dist/server.js"],
      "env": {
        "GOOGLE_DRIVE_FILE_ID": "your_file_id",
        "GOOGLE_CREDENTIALS_PATH": "/path/to/credentials.json",
        "DOTENV_CONFIG_QUIET": "true"
      }
    }
  }
}

To run MCP Inspector

npx @modelcontextprotocol/inspector

🛠️ Available Tools

The server provides 12 data retrieval tools:

get_daily_metrics - Health data for specific dates

get_date_range_metrics - Data across date ranges

get_activity_data - Steps, workouts, calories burned

get_sleep_data - Sleep quality and patterns

get_heart_data - Heart rate and recovery metrics

get_nutrition_data - Nutritional intake tracking

get_metric_history - Historical trends for specific metrics

get_user_profile - User information and goals

get_weekly_data - Weekly aggregated health data

get_monthly_data - Monthly health overviews

get_seasonal_data - Seasonal pattern analysis

get_metric_range - Custom metric analysis across dates

Data Flow

Authentication: Service account credentials validate Google Drive access

Data Retrieval: Excel spreadsheet downloaded and parsed from Drive

Query Processing: Tools filter and structure health data based on LLM requests

Response Formatting: Structured JSON responses for LLM analysis