AI Endurance

Retrieves the user nutrition model with daily calorie and macronutrient requirements (protein, fat, carbohydrates) including lower and upper bounds. Returns data for the past day plus 5 future days based on planned workouts and user physiology.

getNutritionModel

Retrieves a list of the user's "other" activities with summary data - any sport outside running, cycling, and swimming (e.g. strength training, cross-country skiing, yoga, hiking, walking). Use this when the activity is not a run, ride, or swim. Returns at most the 40 most recent activities when a date range is given, otherwise the 20 most recent. Other activities are duration-only: there is NO time-series/stream data and NO detail tool for them, so do not expect power, pace, HRV, or per-second metrics.

getOtherActivity

Retrieves performance predictions from AI Endurance machine learning models including future forecasts, historical data, and model validation metrics.

getPrediction

Retrieves user training plan progress showing how well they are adhering to their prescribed training zones. Returns progress data comparing actual training completed versus what was prescribed in the plan, broken down by training zones (Endurance, Tempo, Threshold, VO2Max, Anaerobic). For triathletes, includes activity-specific progress for Ride, Run, and Swim. Always includes weekly compliance summary. Pass fullDetails=true to also get per-workout compliance breakdown.

getPlanProgress

Retrieves a list of planned workouts for a user between two dates. Default range is next 14 days from today. Returns a human-readable summary per workout; use summaryMode=true for a lightweight overview of the entire plan, or fullDetails=true for the machine-readable step structure.

getPlannedWorkouts

Retrieves primary and secondary race goal events with performance predictions and priorities.

getRaceGoalEvent

Retrieves user recovery model data including cardio recovery, DFA alpha 1, HRV (rMSSD), resting heart rate, external stress score, and orthopedic (joint/muscle) recovery for cycling, running, and swimming. Returns daily time-series data for the last 14 days by default; pass days_back for a longer recovery trend.

getRecoveryModel

Retrieves a list of completed running activities with summary data (date, duration, distance, gradient_adjusted_pace, heart rate, stress scores, weather). gradient_adjusted_pace is GAP - the pace normalized for elevation change and deliberately the only pace reported; on hilly terrain it differs from the pace on the athlete's watch by design (typically faster on climb-heavy runs, possibly slower on descent-dominant ones - cite the terrain when they differ), and it must never be used to re-derive distance. Returns the 20 most recent activities if no date range is specified, up to 40 with a date range. Each activity carries its id in the "id" field: pass it as activityId to getRunningActivityDetail for the full session, or to setActivityFlags to correct how the activity is analyzed.

getRunningActivity

Retrieves detailed information about a specific running activity by its id. Returns the basic activity data plus device laps by default. Pace comes in two distinct channels that must never be compared: gradient_adjusted_pace and the gap time-series are GAP (elevation-normalized, the pace we report), while each lap's avg_pace_device is the RAW pace from the athlete's watch - on hilly terrain they differ by design (typically GAP faster on climbs, possibly slower on descents). Use this when you need in-depth analysis of a single run, not when fetching a list of activities. Durability and the pace/power curve are opt-in: set with_dfa_alpha1 for the internal/HRV drift view (durability_drift) and with_power_curve for the external curve and within-run fade view; raw per-sample arrays need with_time_series_metrics. For computed numbers (normalized power, intensity factor, time-in-zone, pacing/fade, channel extremes) prefer analyzeActivityStream and keep with_time_series_metrics off.

getRunningActivityDetail

Retrieves a list of completed swimming activities. Returns at most the 40 most recent activities if no date range is specified. Each activity carries its id in the "id" field: pass it as activityId to getSwimmingActivityDetail for the full session.

getSwimmingActivity

Retrieves detailed information about a specific swimming activity by its id. Returns the basic activity data plus device laps by default: one lap per lap-button press, structured-workout step or distance auto-lap, with rest steps included as zero-distance laps (distance_m 0, elapsed_s = rest seconds), and pool laps nesting their per-length splits in "lengths". Lap start_s/end_s are on the STORED stream axis (active swimming only, rest excluded, the same axis as time_series_metrics and analyzeActivityStream), so pass them directly as an analyzeActivityStream range {type: "time_seconds", from: start_s, to: end_s}; wall_start_s is the watch's elapsed time incl. rest. Lap and length avg_pace come pre-formatted as "mm:ss" per 100 m or 100 yd (see pace_unit) - quote them, never recompute. Use this for in-depth analysis of a single swim, not when fetching a list of activities. Raw per-sample arrays (time, pace, cumulative distance_m on the same active-only axis) need with_time_series_metrics.

getSwimmingActivityDetail

Retrieves user training availability including weekly hours breakdown and daily schedule for each activity type.

getAvailability

Retrieves comprehensive user data from AI Endurance including training zones, preferences, and physiological metrics. Zone values are zone UPPER bounds, not thresholds (Ride Threshold bound = 106% of FTP, run power Threshold bound = 105% of critical running power); never quote a bound as the user's FTP or CP.

getUser

Sets per-activity flags on a cycling or running activity: indoor, virtual (Zwift etc.), erg mode, and read-time analysis exclusions (power/pace curves, digital twin model training, durability, unreliable heart rate data). Only pass the flags to change; others stay untouched. Use when the user says an activity was misdetected (e.g. an indoor ride treated as outdoor) or wants an activity with bad sensor data excluded from analysis. Setting isIndoor also updates the stored activity weather (indoor marker, or a re-fetch when flipped back to outdoor). Get the activity id from getCyclingActivity/getRunningActivity results.

setActivityFlags

Sets training zones for a specific activity type (Run or Ride). For runners and triathletes using running power, automatically manages both pace and power zones. IMPORTANT: Check existing zones with getUser before setting new zones.

setZones

Skips (removes) a specific workout from the training plan. The workout will be marked as skipped.

skipWorkout