home / mcp / ai endurance mcp server
Provides access to training plans, activities, predictions, recovery metrics, and goals via MCP clients.
Configuration
View docs{
"mcpServers": {
"ai-endurance-mcp": {
"url": "https://aiendurance.com/mcp"
}
}
}Connect your training data to Claude and other MCP clients with the AI Endurance MCP Server. It exposes your training plans, activities, recovery metrics, and predictions so you can ask questions, modify workouts, and get insights through natural conversation.
You connect an MCP-compatible client to the AI Endurance MCP Server at the designated URL. Use a compatible chat or app to sign in with OAuth and choose read or write scopes. Once connected, you can ask your client to show upcoming workouts, check your recovery status, analyze recent activities, and create or adjust workouts. Typical workflows include reviewing your plan, moving workouts, requesting performance predictions, and asking for zone breakdowns or pacing guidance. Your client will present results and, where available, update connected platforms on applicable actions.
Prerequisites you need before connecting an MCP client: install a supported runtime and ensure you can run networked applications. Common options include Node.js with npm or Python with pip. Verify you can run a basic command and access the internet.
Step-by-step flow to enable MCP access from your client:
1) Ensure you have Node.js and npm installed, or Python and pip installed on your machine.
2) Open your MCP-enabled client application or environment.
3) Configure the client to connect to the MCP base URL:
- Base URL: https://aiendurance.com/mcp
4) Initiate the OAuth flow in the client to authorize access with the read and/or write scopes.
5) Complete the authorization in your browser and return to the client to finish setup.
6) Test by asking the client for your upcoming workouts or recovery status.The AI Endurance MCP Server is available via a remote server endpoint that supports Server-Sent Events (SSE) for streaming data. The connection is authenticated using OAuth 2.0. When you connect through a compatible client, you use the server URL below and follow the OAuth prompts to authorize access.
Server connection reference you can use in a compatible client:
{
"url": "https://aiendurance.com/mcp",
"transport": {"type": "sse"},
"auth": {
"type": "oauth",
"authorizationUrl": "https://aiendurance.com/authorize/",
"tokenUrl": "https://aiendurance.com/api/o/token/",
"scopes": ["read", "write"]
}
}The MCP server uses OAuth 2.0 for authentication. You grant your client read and/or write access to your data. You can revoke access at any time from the MCP client you use.
Ask for a plan review: “Show me my workouts for this week.” The client displays upcoming workouts with dates, types, durations, and zones. Ask to adjust: “Move tomorrow's threshold workout to Friday.” The client reschedules and syncs to connected platforms where supported.
The MCP server supports any client that adheres to the MCP 2025-03-26 specification and SSE transport. Examples include Claude-based tools and other MCP-compatible apps that can handle OAuth 2.0 flows and JSON-RPC communication.
If you encounter authentication errors, refresh the OAuth token or reauthorize the client. If you see missing data or failed actions, check that the client has the write scope for any operations that modify workouts or zones.
The MCP server provides access to training plans, activities, predictions, recovery metrics, and race goals. It cannot start plan generation, modify billing, delete historical data, or alter external connections without explicit client actions.
The server exposes a broad toolkit to view, create, modify, and analyze workouts and activities. See the Tools section for a complete list of capabilities and how to invoke them from your MCP client.
By connecting your training data to an MCP client, you gain conversational access to workouts, analytics, and training guidance, enabling you to make informed adjustments, stay on track with your plan, and visualize progress over time.
View user profile including training zones, units, and preferences.
Update cycling power zones or running pace zones with unit-specific formats and automatic dual-zone handling for runners.
View weekly training hours and daily availability for each activity type.
Retrieve planned workouts for a date range with details like date, type, duration, zones, and intervals.
Move a workout to a new date and synchronize the change across connected platforms.
Remove a workout from the plan and propagate the deletion to connected services.
Mark a workout as completed or incomplete in internal tracking.
Add or update coaching advice for a workout.
Create a custom cycling or running workout with structured steps and zones.
Create a custom swim workout with sections, sets, and pace/disciplines.
Create a workout for strength or other activities (e.g., yoga, hiking) with a description.
List recent cycling activities with summary metrics.
List recent running activities with summary metrics.
List recent swimming activities with summary metrics.
Provide detailed cycling activity with time-series data.
Provide detailed running activity with time-series data.
Provide detailed swimming activity with time-series data.
View primary and secondary race goals with predictions and priorities.
ML-based performance predictions including future trajectories and model validation.
Recovery model data including HRV, resting HR, and readiness metrics.
Training plan progress showing adherence by zone and sport.