Labsco
MCP SERVER

RunDida

by XWeaponX7

Running maths an assistant can actually do — paces, race predictions, heart-rate zones and marathon countdowns.

Health, Fitness & WearablesVerified
Summary
The formulas are named, which is the whole reason to use it.

Ask a model to predict a marathon time from a 10K and it will produce a plausible number. Here the answer comes from Riegel, the zones come from Karvonen, and the tool says so — so a runner can check the method rather than trust the arithmetic.

What it is

A running toolkit for agents. Four tools compute locally with no API call — pace, race prediction, heart-rate zones and a countdown — and six query RunDida's public library of calculators, training guides and marathon events.

What you get
  • calculate_pace works out pace, finish time or distance from any two of the three
  • predict_race applies the Riegel formula with VO2max estimation, from a known race result
  • heart_rate_zones computes training zones by the Karvonen method from age and resting heart rate
  • marathon_countdown returns the countdown to a specific event
  • list_tools and get_tool browse the running calculators with their FAQs and sources
  • list_guides and get_guide browse the training guides
  • list_marathons and get_marathon cover events with dates, locations, course profile and race-day weather
  • The compute tools run locally with no latency; the data tools fetch from the RunDida API with 30-minute caching
Requirements

The npm package @rundida/mcp-server, run with npx over stdio. Node.js 18 or newer. No API key and no account. The compute tools work without network access; the data tools need to reach rundida.com.

Setup effort

One command — claude mcp add rundida -- npx -y @rundida/mcp-server