AI assistants are becoming increasingly capable of analyzing complex datasets, identifying patterns, and answering questions in natural language. For athletes, this opens up exciting possibilities. Instead of manually digging through charts and dashboards, you can simply ask questions about your training, recovery, health and performance.
That’s exactly why we’ve launched the RUNALYZE MCP Server.
Using the Model Context Protocol (MCP), you can connect RUNALYZE directly to AI assistants such as Gemini, Mistral, Claude Desktop and many other MCP-compatible clients.
What is MCP?
The Model Context Protocol (MCP) is an open standard that allows AI assistants to securely access external tools and data sources. Instead of uploading screenshots or manually exporting data, MCP allows an AI assistant to retrieve information directly from your RUNALYZE account – with your authorization.
Once connected, you can ask questions such as:
- How has my HRV changed during marathon training?
- What are the main factors affecting my recovery?
- How has my resting heart rate evolved this season?
- Am I currently accumulating fatigue faster than fitness?
- How does sleep quality affect my performance?
- What race result does my current fitness indicate?
More than just activity data
Many MCP integrations in the endurance sports ecosystem focus primarily on activity data such as workouts, distances, heart rate recordings and activity streams. While this is sufficient for analyzing what an athlete did, it provides only limited insight into how the athlete responded to training. Questions about recovery, readiness, fatigue, sleep, health trends or long-term performance trends require a much broader dataset.
The RUNALYZE MCP Server combines activity data with training metrics, race history, recovery indicators, health data and subjective athlete feedback, giving AI assistants the context needed to provide deeper and more personalized insights.
A complete athlete model
The RUNALYZE MCP Server provides access to:
Activities
- Activity history
- Distance, duration and heart rate
- Elevation gain
- Splits
- RPE
- Effective VO2max estimates
Training Metrics
- CTL, ATL and TSB (Fitness, Fatigue and Performance)
- Effective VO2max
- Acute:Chronic Ratio
- Training Monotony and Strain
- Marathon Shape
- HRV Baseline and Normal Range
- Recommended training load
- Recommended recovery days
Performance Data
- Historical race results
- Performance predictions
Health & Recovery Data
- Sleep duration, stages and quality
- Overnight heart rate
- Respiratory rate
- HRV trends
- Resting heart rate trends
- Weight and body composition
- Blood glucose
- Blood pressure
- Body temperature
- Stress, mood and fatigue tracking
- Daily notes
- Injuries and illnesses
This allows AI assistants to connect training outcomes with recovery, health and long-term performance trends.
What can you ask?
Once connected, you can ask questions that go far beyond workout summaries:
Training Analysis
- “How has my fitness changed over the last three months?”
- “What training blocks produced my biggest fitness gains?”
- “Am I currently accumulating fatigue faster than fitness?”
- “Is my training load increasing too quickly?”
- “What signs of overtraining can you find in my data?”
Recovery & Health
- “How does sleep quality affect my running performance?”
- “How did my HRV react during my highest training load weeks?”
- “What happened to my resting heart rate before my last race?”
- “How long did recovery take after my last marathon?”
- “Are there patterns between stress levels and performance?”
Race Preparation
- “How ready am I for my marathon?”
- “What race result does my current fitness indicate?”
- “How does my current marathon shape compare to previous race preparations?”
- “What changed before my personal best performance?”
Instead of simply reporting metrics, AI assistants can identify relationships between fitness, fatigue, recovery and performance.
Getting Started
The RUNALYZE MCP Server is available to all RUNALYZE Premium Members.
Setup Guides
Gemini CLI: https://runalyze.com/help/article/gemini-with-runalyze-mcp
Mistral Le Chat: https://runalyze.com/help/article/mistral-with-runalyze-mcp-server
Any MCP Client Supporting Personal API Tokens (e.g. Claude Desktop, Cursor, Windsurf, Open WebUI, and other MCP-compatible clients): https://runalyze.com/help/article/mcp-with-personal-api-token
Privacy and Security
Your data remains under your control at all times.
Authentication is handled through either:
- Personal API Tokens
- OAuth 2.1 credentials
Only clients you explicitly authorize can access your RUNALYZE data. Credentials can be revoked and regenerated at any time from your RUNALYZE account settings.
Start Exploring Your Data
The combination of RUNALYZE and MCP creates a new way to interact with your training data.
Instead of searching through dashboards, filtering reports, and comparing charts manually, you can simply ask questions and receive personalized answers based on your complete training and health history.
Whether you’re preparing for your next race, optimizing recovery, or exploring long-term performance trends, the RUNALYZE MCP Server gives AI assistants the context they need to help you make better training decisions.
