Strava MCP

Fetches the activity statistics (recent, YTD, all-time) for an athlete. Defaults to the authenticated athlete; pass athleteId to target a specific athlete.

Aggregates personal best efforts across all running activities. This tool retrieves and ranks your best times at standard distances: - 400m, 1/2 mile, 1K, 1 mile, 2 mile - 5K, 10K, 15K, 10 mile - Half Marathon, Marathon, 50K Use Cases: - Track PRs across all distances - Find your best performances at specific distances - Analyze progress over time at key distances Parameters: - distance (optional): Filter to a specific distance (e.g., "5K", "1 mile") - limit (optional): Maximum number of efforts to return per distance (default: 3, max: 50) - maxActivities (optional): Maximum number of acti…

Internal data feed for the cadence-trends UI: returns per-run summary cadence and pace for recent running activities as JSON. The view-cadence-trends app calls this; not intended for direct model use.

Internal data feed for the compare-activities UI: returns the aggregate comparison (per-activity summaries, activity2−activity1 differences, efficiency analysis) as JSON. The view-compare-activities app calls this alongside get-activity-streams-raw; not intended for direct model use.

Computes the fitness/fatigue/form trend (CTL, ATL, TSB) from Strava relative effort. This tool builds the classic performance-management chart from the relative effort (suffer score) already present on the activity list — no per-activity fetches: - CTL ("fitness"): 42-day exponentially weighted average of daily load - ATL ("fatigue"): 7-day exponentially weighted average of daily load - TSB ("form"): CTL − ATL. Negative = carrying fatigue, positive = fresh Use Cases: - "When does my form (TSB) return positive, and does it align with my next quality day?" - Judge whether a training block is …

Internal data feed for the fitness-trend UI: returns the per-day CTL/ATL/TSB series, the forward projection, any solved taper plan (weekly loads and the days they produce), and the dated deep-fatigue / freshness / steep-ramp bands as JSON. The view-fitness-trend app calls this; not intended for direct model use.

Analyses climbing and descending performance within one activity from its elevation, grade, and pace streams. This tool detects sustained climbs (grade ≥ 2% for ≥ 200 m, dip-tolerant) and descents, and reports per segment: - Start km, length, average grade, elevation gain - Moving pace and grade-adjusted (GAP, flat-equivalent) pace - Average HR, cadence, and power where recorded The headline output is early-vs-late climb drift: climb effort is normalised as HR per unit of grade-adjusted speed, then climbs starting in the first half of the run are compared with climbs in the second half. Pos…

Detects and analyses interval structure in one activity, with urban-stop-aware rest classification. Naive rest-based interval detection false-positives on urban runs: traffic-light stops read as recovery intervals. This tool classifies every stopped segment (from the moving stream) before trusting it: - Stop under 60 s with no fast effort before it → traffic light, excluded - Stop up to 3 min after a fast effort → genuine interval recovery - Stop over 5 min → café/regroup/kit stop, noted but excluded - Anything else → unclassified, excluded (lowers confidence) Work reps are reconstructed be…

Predicts race times from your recorded best efforts and builds a goal-pace split table. Uses Riegel's equivalent-performance formula (T2 = T1 × (D2/D1)^1.06) over the best efforts Strava records inside each run, combining them into one estimate per distance weighted by how recent each effort is and how far it has to be extrapolated. Use Cases: - "I am racing a half in six weeks — what should I target, and what is my km split?" - Sanity-check a goal time against what your training actually supports - See which effort is driving a prediction, and how much your distances disagree Parameters: …

Fetch full detail for one saved route by id: name, distance, elevation gain, estimated moving time, and segment count. Use when the user wants details of a route from list-athlete-routes, or before exporting it with export-route-gpx or export-route-tcx.

Internal data feed for the route-map UI: returns decoded [lat, lng] coordinates plus start/end points, distance, elevation gain, and (for activities with GPS streams) index-aligned metric streams (time, distance, altitude, heartrate, watts, velocity_smooth, grade_smooth) and annotation anchors (lap boundaries, segment-effort spans with PR/top-10 flags, geotagged photos, caller-supplied distance-anchored waypoints) for one activity or route as JSON. The view-route-map app calls this; not intended for direct model use.

Previews a saved route's climbs before you ride or run it, from the route's stored elevation profile. list-athlete-routes and get-route report a route's total distance and elevation gain, which says nothing about where the hard parts are. This tool names every sustained climb on the course and reports: - Each climb's position along the route, length, average grade, and gain - The crux: the steepest sustained ~200 m stretch and how far in it sits - A gradient band breakdown along the whole course - A shape verdict — steady, front-loaded, back-loaded, rolling, or flat Use Cases: - "Where are …

Retrieves a comprehensive running-focused summary of a Strava activity. This tool combines data from multiple sources and computes derived metrics specifically for running analysis: - Pace in min/km and min/mile (not just speed) - Cadence as steps per minute (doubled from Strava's strides) - Heart rate zone distribution computed from HR stream - Per-lap breakdown with running-specific metrics Use Cases: - Get a complete picture of a run in a single call - Analyze heart rate zone distribution - Review lap-by-lap performance with pace data - Assess cadence efficiency Parameters: - activityId…

Fetch full detail for one segment by id: distance, average and maximum grade, total elevation gain, climb category, location, and starred state. Use after finding a segment via explore-segments or list-starred-segments when the user wants its difficulty or stats. Obtain the id from those tools.

Fetch one specific segment effort by its effort id, including elapsed and moving time, PR rank, and KOM/QOM rank. Use when the user asks about a single recorded attempt on a segment. For all of an athlete's efforts on a segment use list-segment-efforts; for the segment's static stats use get-segment.

Breaks one segment's gradient down along its length, from its stored elevation streams. get-segment reports a single average grade, which hides the difference between a steady ramp and a flat kilometre followed by a wall — two segments you would pace completely differently. This tool reports: - A gradient band every 100 m (wider on long segments) with each band's own grade - Every sustained climb inside the segment (grade ≥ 2% for ≥ 200 m, dip-tolerant) - The crux: the steepest sustained ~200 m stretch and how far into the segment it sits - A shape verdict — steady, front-loaded, back-loaded…

Internal data feed for the segment-progress UI: returns the segment's details plus the athlete's efforts on it (date, elapsed and moving time, pace, heart rate, power, cadence, PR/KOM ranks, rank within the history) and a derived summary as JSON. The view-segment-progress app calls this; not intended for direct model use.

Breaks one activity into even distance splits and says whether it was positive-, negative-, or evenly split — corrected for terrain. Device laps are whatever the athlete pressed the button for; this tool ignores them and bins the streams into fixed 1 km (or 1 mile) splits, reporting per split: - Moving pace and grade-adjusted (GAP, flat-equivalent) pace - Elevation change and average grade - Average HR, cadence, and power where recorded The headline is the two-halves verdict, stated twice: once on the clock and once grade-adjusted. A hilly back half slows raw pace with no fade at all, and a…

Retrieves training load summary for a specified time period. This tool aggregates running activities to provide: - Total and weekly volume (distance, time, elevation) - Training trend analysis (increasing, stable, decreasing) - Warnings for sudden volume increases (injury risk) Use Cases: - Monitor weekly training volume - Track training consistency over time - Identify potential overtraining risks - Review training history for periodization planning Parameters: - days (optional): Number of days to analyze (default: 28, i.e., 4 weeks) - activityTypes (optional): Activity types to include (…

Internal data feed for the training-load UI: returns per-week running volume (distance, runs, time, elevation), a rolling trend value, and injury-risk warning flags with reasons as JSON. The view-training-load app calls this; not intended for direct model use.

Lists the routes created by the authenticated athlete, with pagination.

Lists the authenticated athlete's efforts on a specific segment, optionally filtering by date.

List the segments the athlete has starred. Returns each segment's id, name, distance, average grade, and climb category so the model can pick one to inspect with get-segment or list efforts with list-segment-efforts. Results are paged: when the response says more are available, call again with the next page. Use when the user refers to their saved or favorite segments.

Zoom the activity chart's x-axis to one window of the activity, given either in kilometres from the start or in seconds of elapsed time. Use it to put the part of the run being discussed on screen — a surge, a climb, an interval — rather than describing where to look. Pass reset to show the whole activity again.

Zoom the route map to a stretch of the course, given in kilometres from the start. Use it to show the user where on the route something happens — a climb, a split, a segment — instead of only describing it. Omit both bounds and pass reset to show the whole route again.

Stars or unstars a specific segment for the authenticated athlete.

Updates an activity's title, description, sport type, gear, or flags (commute/trainer/hidden). Description defaults to append (adds to existing notes); pass descriptionMode 'replace' to overwrite. Requires the activity:write scope. Assign gear by passing a Strava gear id.

Open an interactive chart of one activity with selectable heart rate, power, pace, altitude, cadence, and grade overlays. Prefer this over a text summary when the user wants to see or explore how metrics change over the course of an activity. Takes the activity id.

Open a prioritised, scrollable list of the segments run in one activity: your PRs and top-10s pinned on top, then every segment in run order, each with pace, grade, and expandable heart-rate, cadence, and power detail. Prefer this over text when the user wants to review the segments in a workout. Takes the activity id.

Open an interactive time-in-zone chart for one activity: bars for the time spent in each heart rate and power zone, with percentages and an easy/moderate/hard split. Prefer this over the text-only get-activity-zones when the user wants to see how a workout's effort was distributed. Takes the activity id.

Open an interactive cadence dashboard across recent runs: trend timeline, cadence-versus-pace scatter, pace-zone breakdown, and per-run overlay comparison. Prefer this over text when the user wants to explore cadence patterns over time. Takes a number of weeks of history.

Open an interactive side-by-side overlay of two activities: their pace, heart rate, power, cadence, or altitude streams aligned on a shared distance or time axis, with an aggregate delta summary. Prefer this over the text-only compare-activities when the user wants to see WHERE in the activities the difference happened. Takes both activity ids.

Open an interactive fitness/fatigue/form chart (the performance-management chart): fitness (CTL) and fatigue (ATL) over the lookback window with form (TSB) on its own axis, deep-fatigue, freshness, and steep-ramp periods shaded, and a dashed continuation past today. Pass targetDate to chart a solved taper landing on targetTsb on that day, week by week; omit it for a rest projection. Prefer this over the text-only get-fitness-trend when the user wants to see whether they are peaking or digging a hole.

Open an interactive map of one activity's or saved route's GPS track, fit to bounds with start and finish markers and a distance/elevation summary. Prefer this over a text summary when the user wants to see where an activity or route went. Takes either an activity_id or a route_id (provide exactly one). Optionally pin distance-anchored waypoints (fueling points, climb warnings, …) along the track via the waypoints array — useful when discussing a race plan or course guide.

Open an interactive history of the athlete's own efforts on one segment: effort time over date with the personal best and top three highlighted, an optional average heart rate series, and a per-effort list. Prefer this over the text-only list-segment-efforts when the user wants to see whether they are getting faster on a climb or course segment, or whether the same time is now costing less heart rate. Takes the segment id (from list-starred-segments, explore-segments, or get-segment) and an optional date range.

Open an interactive training-load chart: weekly running volume bars with a rolling trend line, and injury-risk warning weeks highlighted with their reason on hover. Prefer this over text when the user wants to see how their training volume is trending. Takes a number of days of history.