training-load

TrainingPeaks MCP server for Claude Desktop, Code and Cowork. No API approval needed - works with any account. Query workouts, CTL/ATL/TSB fitness data, power PRs via natural language.

Evidence-based AI coach for endurance training. Protocol-driven. Deterministic guidance for any LLM, with Intervals.icu integration.

The open-source alternative to TrainingPeaks. Self-hostable endurance training platform. Your data stays yours.

🏃 An R package for advanced sports performance analysis and training load monitoring using Strava data.

Deterministic endurance coaching engine using Intervals.icu data, governed training logic, performance intelligence, adaptation progression and AI interfaces.

Own your training data! Open-source, self-hosted training log & analytics. Calendar, fitness/fatigue/form, dashboards, zones, peak curves & intervals.icu sync.

Local MCP server exposing your Intervals.icu training data to AI clients (Claude, etc.) — with server-side PMC / cardiac-decoupling math and a Stryd LBSS power extension.

Local-first running intelligence: import Strava, Garmin and OpenTracks, get injury-risk, 80/20 and VO2max analytics, and an AI coach that plans your next race.

🏋️♂️ Connect TrainingPeaks to AI assistants with ease. Query workouts, analyze data, and track fitness trends without API approval hassles.

Heartwork 健康画布 — a local-first health canvas measured against your own baseline, with an evidence-gated AI coach for Claude, ChatGPT, Gemini, Codex and DeepSeek.

Self-hosted cycling coach with a local LLM that never makes up a number — training load, overtraining detection & grounded AI advice.

Free, open-source training hub for cycling, swimming, running and lifting, on macOS, Windows or Linux, planned with your AI assistant. Your reports decide; your data stays on your computer.

Private, data-first training dashboard for your WHOOP data: sports-science load metrics, explainable daily call, and a full-history AI chat. Local-first, zero dependencies.

Local, self-hosted training log with Garmin sync, fitness/training-load analytics, and an AI coach: fart (speed) + lek (play).

Self-hosted MCP server on your own Garmin data: mirrors Garmin Connect into PostgreSQL, computes training load and muscle freshness, and serves it to Claude, ChatGPT and other clients over stdio or HTTP.

Self-hosted training log and analytics for your own FIT files — training load, fitness & form, power curves, records, 3D route flyover. A private Strava / Intervals.icu alternative that runs in Docker on a Pi.

Open-source, self-hosted endurance training analysis and AI coaching for runners, cyclists, and multisport athletes.

Evidence-based Claude Code skill: Garmin FR570 → 4-tier daily training decision + 8-KPI weekly fat-loss review. 206 citations, deterministic decision tree, anti-overreach red lines.

Este repositorio recoge el código para el proyecto Trail Analytics: Exploración visual y modelado predictivo de carreras por montaña.

Curated research on training periodization: linear, undulating, block, conjugate, and polarized models. Tapering science, training load monitoring, and individualization evidence.