Strava MCP Broker
An HTTP broker that exposes Strava data and operations (activities, routes, segments, etc.) as MCP tools, allowing users to interact with their own Strava account via natural language using per-user OAuth2 authentication.
README
Strava MCP Broker
An HTTP broker that exposes the r-huijts/strava-mcp server over MCP Streamable HTTP, with per-user Strava OAuth2. Each user brings their own Strava API credentials — no shared app secrets needed.
Quick Start (Hosted Version)
A public instance is running at strava-mcp.julian-bruegger.ch. No setup required.
- Create a Strava API app at strava.com/settings/api
- Set Authorization Callback Domain to
strava-mcp.julian-bruegger.ch
- Set Authorization Callback Domain to
- Visit strava-mcp.julian-bruegger.ch
- Enter your Strava Client ID and Client Secret, then click Connect with Strava
- Authorize on Strava — you'll receive your personal MCP endpoint URL
- Add the endpoint to Claude:
- Claude.ai: Settings → Integrations → Add MCP server → paste the URL
- Claude Desktop: Add to your
claude_desktop_config.json(see below)
- Manage your sessions anytime at strava-mcp.julian-bruegger.ch/dashboard
Self-Hosted Version
Host your own instance with Docker and Cloudflare Tunnel.
Architecture
Claude.ai / Claude Desktop
│ HTTPS (MCP Streamable HTTP)
▼
Cloudflare Tunnel
│
▼
Docker: strava-mcp-broker (Express server, port 3000)
│ stdio JSON-RPC
▼
Docker: @r-huijts/strava-mcp-server (child process per session)
│ HTTPS
▼
Strava API v3
Prerequisites
- A server with Docker + Docker Compose (e.g. Proxmox, VPS, Raspberry Pi)
- A domain managed by Cloudflare (free plan is fine)
- A free Cloudflare Zero Trust account (one.dash.cloudflare.com)
Step 1 — Set Up Cloudflare Tunnel
- Log in to one.dash.cloudflare.com
- Go to Networks → Tunnels → Create a tunnel
- Choose Cloudflared as the connector type
- Name your tunnel (e.g.
strava-mcp) - On the "Install connector" page, choose Docker — copy the tunnel token
- In Public Hostname, add a route:
- Subdomain:
strava-mcp - Domain:
yourdomain.com - Service:
http://strava-mcp:3000(Docker service name on the internal network)
- Subdomain:
- Save the tunnel
Step 2 — Configure and Deploy
git clone https://github.com/julianbruegger/strava-mcp-broker
cd strava-mcp-broker
cp env.example .env
Edit .env:
PUBLIC_URL=https://strava-mcp.yourdomain.com
CLOUDFLARE_TUNNEL_TOKEN=eyJhIjoiM... # from Step 1
SESSION_SECRET=<run: openssl rand -hex 32>
Deploy:
docker compose up -d --build
Check logs:
docker compose logs -f
You should see:
strava-mcp-broker | Strava MCP broker running on http://0.0.0.0:3000
strava-mcp-broker | Public URL: https://strava-mcp.yourdomain.com
strava-mcp-tunnel | Registered tunnel connection
Step 3 — Connect Your Strava Account
- Go to strava.com/settings/api and create an app
- Set Authorization Callback Domain to your tunnel domain (e.g.
strava-mcp.yourdomain.com)
- Set Authorization Callback Domain to your tunnel domain (e.g.
- Open your broker URL in a browser
- Enter your Client ID and Client Secret, then click Connect with Strava
- Authorize on Strava — you'll receive your personal MCP endpoint URL
Connect to Claude
Option A — Claude.ai (recommended)
Go to Settings → Integrations → Add MCP server and paste your endpoint URL:
https://strava-mcp.yourdomain.com/mcp?session=YOUR_SESSION_TOKEN
Option B — Claude Desktop
Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"strava": {
"command": "npx",
"args": [
"mcp-remote",
"https://strava-mcp.yourdomain.com/mcp?session=YOUR_SESSION_TOKEN"
]
}
}
}
Restart Claude Desktop. You'll see the Strava tools available.
Available Tools (25+)
| Category | Tools |
|---|---|
| Athlete | get-athlete, get-athlete-stats, get-athlete-zones |
| Activities | get-activities, get-activity, get-activity-laps, get-activity-streams, get-activity-photos, get-activity-comments, get-activity-kudoers, create-activity, update-activity, delete-activity |
| Routes | get-routes, get-route, list-athlete-routes, export-route-gpx, export-route-tcx |
| Segments | get-starred-segments, get-segment, get-segment-effort, explore-segments, star-segment |
| Clubs | get-clubs, get-club, get-club-activities |
| Gear | get-gear |
| Connection | connect-strava, disconnect-strava, check-connection |
See SKILL.md for detailed tool descriptions and usage guidelines.
Dashboard
Each user gets a dashboard at /dashboard to manage their sessions:
- View all connected sessions
- See MCP endpoint URLs
- Delete individual sessions
- Delete all data at once
The dashboard is protected by a signed cookie set during Strava OAuth — no extra login needed.
Maintenance (Self-Hosted)
Update to latest strava-mcp-server version
docker compose build --no-cache
docker compose up -d
Backup tokens
docker cp strava-mcp-broker:/data/tokens.db ./tokens-backup.db
Security Notes
- Each user provides their own Strava API credentials — the broker stores no shared secrets
- Session tokens are random 256-bit hex strings — treat them like passwords
- Tokens are stored in SQLite at
/data/tokens.db(mounted volume) — back it up - Access tokens are refreshed automatically (Strava tokens expire every 6 hours)
- The
/mcpendpoint is rate-limited (60 req/min per IP) and requires a valid session token - Dashboard auth uses HMAC-signed cookies (30-day expiry, HttpOnly, SameSite)
- Cloudflare Tunnel handles TLS — no certificates needed on the server
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.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
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.