Beyond Strava

A personal tool to fetch and analyze my Strava activities using the Strava API with OAuth2 authentication.

Why I Built This

I wanted to get the most out of my training data. Strava gives you the basics, but there's so much more buried in those activity files. By pulling all my data locally, I can dig deeper, run custom analysis, build visualizations that actually matter to me, and discover insights that the standard apps just don't show.

Plus, let's be honest, this was a perfect excuse to write SQL queries again!

Setup

1. Create a Strava app: https://www.strava.com/settings/api
2. Get your client ID and secret
3. Get your refresh token
4. Create a .env file with your Strava app credentials:

STRAVA_CLIENT_ID=your_client_id
STRAVA_CLIENT_SECRET=your_client_secret
STRAVA_REFRESH_TOKEN=your_refresh_token  # Will be updated after first auth

How to Run

The workflow is the following:

1. uv run fetch.py → fetches activities from Strava API
2. uv run analyze.py → processes the JSON data into DuckDB database
3. duckdb strava_activities.duckdb -f queries/<query-name>.sql → runs analysis queries

Note: Re-run uv run analyze.py whenever you fetch new activities to update your DuckDB database with the latest data.

What it does:

• Launches a local auth server on http://localhost:8000
• Opens your browser automatically for Strava authorization
• Handles the OAuth callback seamlessly
• Fetches ALL your historical activities (using pagination)
• Saves everything to activities/<username>_yyyy-mm-dd_export.json with a nice summary breakdown
• Stores activities in a DuckDB database for fast querying

What You Need

• uv (recommended) or Python 3.11+ installed on your machine
• DuckDB for database queries and UI
• A Strava API app (free to create)
• Your Strava credentials in a .env file

What You Get

Your activities get saved to activities/<username>_yyyy-mm-dd_export.json with all the juicy details:

• Performance metrics: Distance, time, elevation, pace/speed
• Physiological data: Heart rate zones, power output, cadence
• Route data: GPS coordinates, polyline for mapping
• Activity metadata: Kudos, photos, gear used, weather conditions

Perfect for building your own analysis dashboards, tracking progress, or feeding into ML models for performance insights.

Database Analysis

Your activities are stored in a DuckDB database (strava_activities.duckdb) for fast querying. Pre-built queries are organized by sport for personalized analysis:

• queries/global/ - Overall activity analysis
• queries/running/ - Running and trail running specific analysis
• queries/strength/ - CrossFit and weight training analysis
• queries/dashboard/ - Comprehensive training dashboard with all metrics

Quick Dashboard View

Get a complete overview of your training with the comprehensive dashboard:

duckdb strava_activities.duckdb -f queries/dashboard/dashboard.sql

This shows:

• Weekly & Monthly Trends: Training volume, intensity, and progression
• Performance Analysis: Pace trends, heart rate zones, personal records
• Training Balance: Sport distribution, consistency scores, workout scheduling
• Recent Activity Log: Latest workouts with key metrics
• Strength Training: Session frequency, intensity patterns, optimal training days

Custom Analysis

Run any specific query with:

duckdb strava_activities.duckdb -f queries/<folder>/<query-name>.sql

Or launch the DuckDB web UI for interactive exploration:

duckdb -ui strava_activities.duckdb

Common Issues

• Port 8000 busy? The app will tell you - just kill whatever's using it
• Browser won't open? Copy the authorization URL manually
• Empty activities? Check your Strava privacy settings and API permissions