Stride MCP

Chat with your Strava data in Claude and ChatGPT

Stride connects your Strava account to AI so you can ask real questions about your training and get answers grounded in your actual runs, not generic advice.

Get started free →Download stride-skill.md

Works with Claude.ai and ChatGPT. Free to connect. Use your own Strava API application.

Your runs. Your AI coach.

Strava tracks splits, heart rate, shoe mileage, and how your volume shifts week to week. It does not tell you what any of it means for your next month of training.

Stride bridges that gap. It connects Strava to Claude or ChatGPT using an open standard called MCP (Model Context Protocol). Once connected, your AI can pull your data for any activity or timeframe and talk it through with you.

Not canned summaries. Not generic plans. Your trends and your questions.

Live data when you ask

Each question drives fresh calls into Strava so answers reflect recent runs and totals, not a stale export from last month.

Works in Claude and ChatGPT

Add Stride as a connector, then chat in the assistant you already pay for. No extra athlete app required for MCP setup.

Your API credentials

You create your own Strava API application. That avoids shared developer quotas and keeps usage under your Strava app.

Set up in three steps

1

Create your Strava API app

In Strava settings, create a free API application. You receive a Client ID and Client Secret. Set the authorization callback domain to match app.runstri.de as documented in setup.

2

Connect at app.runstri.de

Sign in with Google, paste your Strava credentials, and complete OAuth. You receive a personal MCP URL for Claude or ChatGPT.

3

Add to Claude or ChatGPT

In Claude, open your Project, add the MCP URL as a connector, and add the Stride skill to project instructions. Start chatting.

Get your MCP URL →

What you can actually ask

Real runners ask messy questions. Here is what it looks like when the model can read Strava.

Understanding one long run
You
Can you break down my long run from last Sunday?
Stride Your Sunday run was 22.4 km in 2:08:34. Average pace 5:44/km, average HR 148 bpm. Early kilometres held aerobic effort; pace lifted in the second half while HR crept toward Zone 3. Worth checking how your legs felt the next morning before stacking another hard week.
Monthly comparison
You
How does my training this month compare to last month?
Stride Pulls month-to-date distance, time, elevation, and session count, then compares to the full prior month with percentage deltas and flags when easy-run HR drifts up at similar pace.
Flat legs
You
I have felt really flat. Am I overtraining?
Stride Reviews rolling weekly volume and HR at easy effort across several weeks to spot progressive overload and early fatigue patterns, then suggests conservative easy days before you miss a goal race.
Gear
You
Should I replace my running shoes?
Stride Reads shoe entries from Strava with logged distance, compares rotation pairs, and ties recommendations to your upcoming race or volume block.
Segment progression
You
How have I been going on the hill near my house?
Stride Lists recent segment efforts with times and ranks so you can see whether pace on that climb is trending up across weeks.
Race prep
You
My marathon is in five weeks. Where am I at?
Stride Combines peak weeks, long-run progression, shoe mileage, and recent intensity so you get a grounded snapshot instead of a guess from generic templates.

Prompt ideas

Training load

  • "How is my mileage compared to last month?"
  • "Am I building volume too fast?"
  • "What did the last eight weeks look like?"
  • "Compare this year to last year at this date."

Single sessions

  • "Break down Saturday's long run."
  • "Was yesterday actually easy? Show HR zones."
  • "Did my interval pace hold on the last rep?"
  • "Find my fastest 10K from last year."

Races

  • "Half marathon in four weeks. How is fitness?"
  • "Have I logged enough long runs for this marathon?"
  • "Shakeout tomorrow or full rest?"
  • "Am I hitting race pace in training?"

Trends

  • "Am I getting fitter?"
  • "Why do I feel flat lately?"
  • "Is my easy pace improving?"
  • "How has easy HR changed over three months?"

Gear and segments

  • "How many km on my shoes?"
  • "Time to retire my daily trainer?"
  • "Progress on my favourite segment this spring."
  • "Any segment PRs recently?"

Social

  • "Who gave kudos on yesterday's run?"
  • "Comments on my race activity."

Eighteen tools across your training

Stride exposes specialised tools so Claude or ChatGPT can fetch profile, activities, trends, gear, segments, and social context on demand.

Your profile

Athlete profile

Name, locale, sport prefs, and basics so answers stay personal.

Lifetime stats

Totals and standout efforts for historical context.

Activities

Recent activities

Feed of runs and rides with pace, distance, and elevation.

Single activity detail

Deep dive on one session including HR and gear when present.

Streams and laps

Second-by-second or per-lap views for structured workouts.

Zone distribution

Time in HR or power zones for a session.

Trends

Weekly summary

Rolling weeks compared for load and intensity.

Period comparison

This week vs last week, month vs month, year vs year.

Personal records

Best efforts surfaced from history.

Search, gear, segments, social

Search activities

Find sessions by keyword, type, or date range.

Gear roster

Shoes and bikes with accumulated distance.

Segments and routes

Starred segments, efforts over time, saved routes.

Kudos, comments, photos

Social context tied to an activity when available.

Stride skill for Claude Projects

Paste the Stride skill into Project instructions so Claude fetches live data before guessing, chains tools when helpful, and answers like a coach: lead with the insight, cite real dates and activity names, finish with one sharp follow-up question.

Open or download the skill file and add it alongside your MCP connector in the same Project.

View stride-skill.mdDownload stride-skill.md

Common questions

Do I need a paid Claude or ChatGPT plan?

Claude.ai typically needs a paid plan (for example Claude Pro) to use MCP connectors in Projects. ChatGPT often needs Plus or higher for similar integrations. Check current Anthropic and OpenAI pricing. Claude is usually the smoother fit for Stride MCP today.

Does Stride store my Strava activity data?

Stride stores your Strava API credentials and tokens (encrypted) so the connector can run. Live activity details flow through to your AI session when you ask questions; we do not build a separate workout archive like the full Stride coaching app.

Why do I need my own Strava API application?

Strava limits how many athletes can connect to a single API application. With your own Client ID and Secret, your quota is isolated from everyone else. Stride MCP is built for that model.

Does it work with Garmin, Whoop, or Apple Watch?

If your device syncs activities into Strava, that data is available to Stride MCP wherever Strava exposes it through the API.

What if I am a cyclist, not a runner?

The same tools cover rides, segments, gear, and training load. Coaching tone can span running and cycling.

Is Stride MCP free?

Connecting through app.runstri.de is free on our side. You pay Claude or ChatGPT according to your subscription.

Can I use both Claude and ChatGPT?

Yes. Your MCP URL can be added wherever the client supports it. Exact setup steps differ by product.

Ready to understand your training?

Budget about ten minutes: Strava account, free Strava API app, then Claude or ChatGPT with your MCP URL.

Connect your Strava →
Stride MCP · Chat with your Strava data in Claude and ChatGPT | Stride