Strava Training MCP
Connect Strava training data to Claude for training load analysis and activity insights
Strava MCP Server
Connect your Strava training data to Claude. This MCP server turns Claude into your personal training coach — analyzing your training load, planning workouts, and tracking your progress. Works with cycling, running, and any sport you track on Strava.
Install
Works with Claude Desktop and claude.ai on any plan. Free plans can add 1 custom connector (remove any existing connector first), Pro and Max are unlimited.
1. Create a Strava API app
This takes about 2 minutes. You need a free Strava API app so Claude can access your data.
- Go to strava.com/settings/api
- Fill in the form — Application Name, Category, and Description can be anything (e.g. "My Claude connector")
- Set Authorization Callback Domain to
strava-mcp-web.vercel.app - Click Create and note your Client ID and Client Secret
2. Add the connector
- Open Claude Desktop or claude.ai
- Go to Settings → Connectors → Add custom connector
- Paste this URL and save:
https://strava-mcp-web.vercel.app/mcp
3. Connect
Click Connect on the newly added connector. This opens a page where you enter your Strava Client ID and Client Secret, then authorize with Strava.
That's it! Ask Claude about your training.
Tools
The server exposes 10 tools:
| Tool | Description |
|---|---|
get_recent_activities | Your last activities with distance, duration, and heart rate |
get_activity_details | Deep dive into a specific activity — power, HR, speed, suffer score |
get_training_load_analysis | ATL, CTL, TSB with training advice and 8-week trends |
get_weekly_stats | Weekly volume — activities, kilometers, hours |
get_weekly_training_plan | Weekly plan based on current fitness and fatigue |
get_gear_maintenance | Bike/shoe km totals with chain, cassette, tire warnings |
get_power_curve | Best power outputs (5s–60min), FTP estimate, monthly comparison |
get_hr_zone_distribution | Time in each HR zone with polarized training advice |
get_hr_drift_analysis | Cardiac drift in steady rides — aerobic efficiency indicator |
check_workout_quality | Interval consistency, power decoupling, recovery scoring |
Example conversations
"I want to ride tonight, what should I do?" "How is my training load looking?" "Check the quality of my last interval workout" "When do I need to replace my chain?" "What's my FTP and how does it compare to last month?" "Am I spending enough time in zone 2?" "I'm training for a 150km race in April — am I on track?"
How it works
Claude pulls your Strava data in real-time and reasons about it in context — your recent rides, fitness trend, fatigue level, and goals.
Training load analysis uses the standard PMC model:
- ATL (7-day) — how tired you are now
- CTL (42-day) — your fitness level
- TSB — CTL minus ATL. Positive = fresh, negative = fatigued
- Ramp rate — week-over-week load change, with injury risk warnings
Gear maintenance tracks km on each bike and shoe, warning when chain (3000km), cassette (6000km), tires (5000km), or cables (8000km) need attention.
Power curve finds your best efforts across durations and estimates FTP at 95% of your 20-minute power.
HR analysis covers zone distribution (are you polarizing enough?) and cardiac drift (is your aerobic base improving?).
Limitations
- Training load uses suffer score / HR — it underestimates fatigue from strength training or activities where HR stays low
- Power features require a power meter or smart trainer
- The weekly plan is a heuristic, not a replacement for a certified coach
Troubleshooting
Connector not working? Remove and re-add the connector in Settings → Connectors.
Authentication error?
Check that your Strava API app has strava-mcp-web.vercel.app as the callback domain.
Website
Built with
- Model Context Protocol (MCP) — Anthropic's open standard for connecting AI to external tools
- Strava API v3 — Access to athlete activities and metrics
License
MIT
Installation
Source-derived launch command. Check the maintainer’s required arguments and credentials before running:
uvx strava-training-mcpSet up in your AI client
Merge this template into ~/Library/Application Support/Claude/claude_desktop_config.json. Keep existing servers. Add any arguments, credentials, and permissions required by the maintainer; this template has not been install-tested.
{
"mcpServers": {
"io-github-arjanlig-strava-mcp": {
"command": "uvx",
"args": [
"strava-training-mcp"
]
}
}
}Restart Claude Desktop completely for changes to take effect. Confirm the server appears connected in the client’s tool list, then try a read-only example from its documentation.
Claude Desktop setup referencePackage
strava-training-mcppypiCompatible MCP Clients
Strava Training MCP works with any MCP-compatible client. Copy the config snippet from the Configuration section above and add it to the file shown for your client, then restart the application.
- Claude Desktop
~/Library/Application Support/Claude/claude_desktop_config.jsonRestart Claude Desktop completely for changes to take effect. - Cursor
~/.cursor/mcp.jsonRestart Cursor for changes to take effect. - VS Code
.vscode/mcp.jsonReload VS Code window for changes to take effect. - Windsurf
~/.codeium/windsurf/mcp_config.jsonRestart Windsurf for changes to take effect. - Claude Code
.mcp.jsonSave at the project root, then start Claude Code in that project and review the MCP server approval prompt. Keep real credentials out of shared files.