Intervals.icu MCP Server

The Intervals.icu MCP Server bridges the gap between Claude and the rich, time‑stamped activity data stored in Intervals.icu. By exposing a set of authenticated tools over the Model Context Protocol, it lets developers treat an athlete’s training log as a first‑class data source in AI workflows. Instead of manually exporting CSVs or building custom API wrappers, a Claude assistant can now query for events, activities, and wellness metrics with simple tool calls, instantly incorporating the latest performance insights into coaching conversations, injury prevention analysis, or personalized training plans.

At its core, the server offers three primary capabilities: authentication, activity retrieval, and wellness data access. The authentication tool securely exchanges an API key and athlete ID for a short‑lived bearer token, ensuring that each request respects Intervals.icu’s rate limits and privacy settings. Once authenticated, the activity tool can fetch detailed summaries of runs, rides, or swims—complete with distance, pace, heart‑rate zones, and GPS tracks. The wellness tool pulls subjective metrics such as sleep quality, recovery scores, or mood ratings that athletes log daily. Together, these tools provide a holistic view of an athlete’s performance ecosystem, enabling Claude to answer nuanced questions like “How did my recent rides affect my upcoming marathon pace?” or “Which training block yielded the best recovery?”

Developers benefit from this server in several concrete ways. In a coaching platform, the MCP can automatically surface an athlete’s latest workouts whenever a client asks for progress updates. In health analytics, the server allows researchers to query longitudinal wellness data without writing bespoke ingestion pipelines. For personal training apps, developers can embed Claude‑powered chatbots that recommend next workouts based on real‑time activity trends. Because the server follows MCP’s declarative schema, it plugs seamlessly into any AI assistant that supports the protocol—Claude Desktop, web agents, or custom bots built on top of the MCP SDK.

What sets this server apart is its focus on real‑time, authenticated access. Unlike static data exports, the MCP tools honor Intervals.icu’s OAuth‑style token flow and respect user privacy by scoping requests to the authenticated athlete only. The server also handles pagination, caching, and error handling behind the scenes, so developers can concentrate on building conversational logic rather than plumbing. Its lightweight Python implementation means it can run locally or in a cloud function with minimal overhead, making it ideal for both hobbyists and production‑grade deployments.