strava mcp

★ 22

from MariyaFilippova

A Model Context Protocol (MCP) server that integrates Strava with Claude for Desktop, enabling AI-powered analysis of your fitness activities.

🔥🔥🔥✓ VerifiedFreeAdvanced setup

#automation #git #maps

MCP Strava Server

A Model Context Protocol (MCP) server that integrates Strava with Claude for Desktop, enabling AI-powered analysis of your fitness activities.

Features

⭐ Route Generation - suggest round-trip routes with a Google Maps link (no API key needed)
⭐ Popular Routes - discover popular Strava segments nearby and get a Google Maps loop through them
OAuth Authentication with automatic token refresh and persistence
Activity Analysis - get details on any activity by ID, or browse recent ones
Data Streams - heart rate, pace, altitude, cadence, power, GPS for any activity
Lap Splits - per-lap distance, time, speed, elevation, and heart rate
Paginated Search - find activities by date range with full pagination
Historical Comparisons - compare months year-over-year
Statistics - all-time stats, weekly/monthly summaries
Filtering - by activity type (Run, Ride, Swim, etc.)

Available Tools

Tool	Description
`auth_strava`	Authorize with Strava (opens browser)
`logout`	Clear stored tokens to switch accounts
`last_activity`	Get details of your most recent activity
`get_activity`	Fetch any activity by its ID
`recent_activities`	List last 10 activities
`search_activities`	Paginated search with optional date range (`before`/`after` epoch, `page`, `per_page`)
`athlete_stats`	All-time statistics (rides, runs, swims)
`activities_by_type`	Filter activities by sport type
`weekly_summary`	Summary of the past 7 days
`monthly_summary`	Summary of the past 30 days
`month_summary`	Summary for a specific month/year
`compare_months`	Compare two months (e.g., Jan 2025 vs Jan 2026)
`get_streams`	Heart rate data for last activity
`get_activity_streams`	Full data streams (HR, pace, altitude, cadence, power, GPS, etc.) for any activity
`get_laps`	Lap splits for any activity (distance, time, speed, elevation, HR)
`suggest_route`	Generate a round-trip route and get a Google Maps link for navigation
`popular_routes`	Find popular Strava segments nearby and build a Google Maps route through them

Authentication

On first use, the server will open your browser for Strava authorization. Tokens are persisted to ~/.strava-mcp-token.json and automatically refreshed when expired.

Use the logout tool to clear stored tokens if you need to switch accounts.

License

MIT

Setup

1. Clone the Repository

git clone https://github.com/MariyaFilippova/mcp-strava.git
cd mcp-strava

2. Configure Strava API Credentials

Get your credentials from Strava API settings.

Create/edit src/main/resources/.env:

CLIENT_ID="your-client-id"
CLIENT_SECRET="your-client-secret"

3. Build the Project

./gradlew shadowJar

The JAR will be at: build/libs/strava-mcp-server-2.1.0-all.jar

4. Configure Claude for Desktop

Edit Claude's configuration file:

# macOS
code ~/Library/Application\ Support/Claude/claude_desktop_config.json

# Windows
code %APPDATA%\Claude\claude_desktop_config.json

Add the MCP server:

{
  "mcpServers": {
    "strava": {
      "command": "java",
      "args": [
        "-jar",
        "/absolute/path/to/mcp-strava/build/libs/strava-mcp-server-2.1.0-all.jar"
      ]
    }
  }
}

Restart Claude for Desktop.

Usage Examples

Once configured, you can ask Claude things like:

"Show me my recent Strava activities"
"What are my all-time running stats?"
"Compare my January 2025 with January 2026"
"Give me a summary of my cycling this week"
"How did my training this month compare to last month?"
"Show me the lap splits for activity 123456789"
"Analyze the heart rate and elevation data for my last run"
"Find all my activities from March 2025"
"Suggest a 5km running route starting from Central Park, NYC"
"Generate a 10km cycling route from 48.8566, 2.3522"
"Find popular running segments near Central Park, NYC"
"Show me popular cycling routes within 10km of Paris"