Strava Mcp
Expose Strava activities, streams, routes, and training context to your coding agent via OAuth with privacy controls and local token storage.
Overview
Strava MCP is a Build-phase MCP server that gives AI agents OAuth-backed access to Strava activities, streams, routes, and training data with privacy-first local tokens.
What is this MCP server?
- Privacy-first unofficial Strava MCP (strava-mcp-unofficial v0.4.6) over stdio
- Activities, streams, routes, and training-oriented agent tools
- strava-mcp-server setup stores STRAVA_CLIENT_SECRET in ~/.strava-mcp/config.json instead of MCP JSON
- STRAVA_PRIVACY_MODE and local tokens default ~/.strava-mcp/tokens.json
- OAuth fields STRAVA_CLIENT_ID, STRAVA_REDIRECT_URI align with Strava Developer Dashboard
- Server version 0.4.6
- stdio npm identifier strava-mcp-unofficial
- Documented tool surface includes activities, streams, routes, and training context
Community signal: 1 GitHub stars.
What problem does it solve?
Hand-rolling Strava OAuth, rate limits, and stream decoding blocks you from shipping the coaching or analytics feature you actually care about.
Who is it for?
Developers with a Strava API application who want agent-ready workout and route data on localhost.
Skip if: Enterprise multi-athlete hosting without your own backend, official Strava partnerships, or teams unwilling to manage OAuth compliance.
What do I get? / Deliverables
Once OAuth is configured, your agent queries Strava-backed MCP tools using stored tokens and privacy settings while you build the product layer.
- Agent-callable Strava activity, stream, route, and training MCP tools
- Local OAuth token storage (default ~/.strava-mcp/tokens.json)
- Privacy-scoped responses via STRAVA_PRIVACY_MODE
Recommended MCP Servers
Journey fit
Strava connectivity is built when integrating external fitness APIs into agents or apps, not during initial idea research alone. Strava Developer OAuth fits the integrations subphase alongside other wearable and endurance data MCP servers.
How it compares
Strava API MCP connector, not a marathon training skill or route-map UI kit.
Common Questions / FAQ
Who is Strava MCP for?
Solo builders creating AI coaching, club tools, or personal endurance dashboards that need live Strava athlete data.
When should I use Strava MCP?
Use it during build when agents must read activities, streams, or routes to personalize feedback, summaries, or internal analytics.
How do I add Strava MCP to my agent?
Install stdio package strava-mcp-unofficial, run strava-mcp-server setup for OAuth, set STRAVA_PRIVACY_MODE, and register the server in your MCP client.