claudex-strava-mcp
Connects Claude to your Strava fitness data, enabling natural language queries about your training history such as activities, stats, and routes.
README
claudex-strava-mcp
A Model Context Protocol (MCP) server that connects Claude to your Strava fitness data. Ask Claude questions about your real training history in plain English — no dashboards, no manual exports.
Built with TypeScript, the MCP SDK, and Zod schema validation. Automatic OAuth token refresh included.
What you can ask Claude
- "How many walks did I do this month?"
- "Where were my last 3 activities?"
- "How do my 2025 and 2026 activity counts compare?"
- "Which of my activities had the most stop time?"
- "Am I hitting my Zone 2 training targets?"
Prerequisites
- Node.js 18+
- A Strava account with activities
- A Strava API app (create one here)
- Claude Desktop (or any MCP-compatible client)
Setup
1. Clone and install
git clone https://github.com/your-username/claudex-strava-mcp
cd claudex-strava-mcp
npm install
2. Create your Strava API app
Go to strava.com/settings/api and create an app. Set the Authorization Callback Domain to localhost.
You'll receive a Client ID and Client Secret — keep these safe.
3. Get your refresh token
Copy the example env file and fill in your Client ID and Secret:
cp .env.example .env
# Edit .env and add your STRAVA_CLIENT_ID and STRAVA_CLIENT_SECRET
Then run the OAuth helper:
npx tsx scripts/get-token.ts
This will:
- Print an authorization URL — open it in your browser
- Ask you to authorize the app on Strava
- Redirect to localhost and exchange the code for tokens
- Print your
STRAVA_REFRESH_TOKENin the terminal
Copy the printed refresh token into your .env:
STRAVA_CLIENT_ID=your_client_id
STRAVA_CLIENT_SECRET=your_client_secret
STRAVA_REFRESH_TOKEN=your_refresh_token
4. Build the server
npm run build
This compiles TypeScript to dist/.
5. Configure Claude Desktop
Add the following to your Claude Desktop config file:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"strava": {
"command": "node",
"args": ["/absolute/path/to/claudex-strava-mcp/dist/index.js"],
"env": {
"STRAVA_CLIENT_ID": "your_client_id",
"STRAVA_CLIENT_SECRET": "your_client_secret",
"STRAVA_REFRESH_TOKEN": "your_refresh_token"
}
}
}
}
Restart Claude Desktop. The Strava tools will appear in Claude's tool list.
Project structure
claudex-strava-mcp/
├── src/
│ ├── index.ts # MCP server entry point
│ ├── strava-client.ts # Strava API client with auto token refresh
│ └── tools.ts # 8 MCP tool definitions (Zod-validated)
├── scripts/
│ └── get-token.ts # One-time OAuth token helper
├── .env.example # Template — copy to .env and fill in values
├── .gitignore
├── package.json
└── tsconfig.json
Available tools
| Tool | Description |
|---|---|
get_athlete_profile |
Your Strava profile and preferences |
get_athlete_stats |
All-time, YTD, and recent totals by sport |
list_activities |
Paginated activity list with date filters |
get_activity |
Detailed stats for a single activity |
get_activity_laps |
Lap-by-lap breakdown |
get_activity_zones |
Heart rate zone distribution |
list_routes |
Your saved routes |
get_route |
Detailed route info |
How it works
The server uses the MCP SDK's McpServer class and registers tools using Zod schemas for runtime validation. The Strava client handles OAuth token refresh automatically — if your access token is within 60 seconds of expiring, it refreshes using your stored refresh token before making any API call.
Claude prompt
│
▼
MCP Tool Layer (TypeScript)
│
├── list_activities ──► GET /athlete/activities
├── get_activity ──► GET /activities/{id}
├── get_activity_zones ► GET /activities/{id}/zones
└── ...
│
▼
Strava API
│
▼
Structured JSON → Claude reasons → Natural language response
Development
npm run dev # Watch mode — recompiles on save
npm run build # Production build
npm start # Run compiled server
Security notes
- Never commit your
.envfile — it's in.gitignore - Rotate your Strava client secret at strava.com/settings/api if it's ever exposed
- The OAuth scopes requested are
read,activity:read_all, andprofile:read_all— read-only access only
License
MIT
Built by Mostafa Didar Mahdi — Data Scientist, Anthropic-certified AI practitioner, Adelaide.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.