Agent Skills: Strava API

Universal Strava API integration for fitness data management. Use when working with Strava activities, athlete profiles, segments, routes, clubs, or any fitness tracking data. Triggers on requests to get/create/update activities, analyze training stats, export routes, explore segments, or interact with Strava data programmatically.

UncategorizedID: kdoronin/claude_code_skills/strava-api

Install this agent skill to your local

pnpm dlx add-skill https://github.com/kdoronin/claude_code_skills/tree/HEAD/strava-api

Skill Files

Browse the full folder contents for strava-api.

Download Skill

Loading file tree…

strava-api/SKILL.md

Skill Metadata

Name
strava-api
Description
Universal Strava API integration for fitness data management. Use when working with Strava activities, athlete profiles, segments, routes, clubs, or any fitness tracking data. Triggers on requests to get/create/update activities, analyze training stats, export routes, explore segments, or interact with Strava data programmatically.

Strava API

Direct integration with Strava API v3 for fitness data management.

Security Model

IMPORTANT: All secrets are stored in system keychain (macOS Keychain / Linux Secret Service).

  • AI agents CANNOT read secrets from keychain
  • Only the Python scripts can access credentials at runtime
  • Non-secret metadata (athlete name, token expiry) stored in ~/.strava/metadata.json

This design ensures your API credentials are never exposed to AI agents.


First-Time Setup Workflow

When user requests Strava data for the first time:

Step 1: Check if configured

python3 scripts/strava_api.py athlete

If error "Strava not configured" → proceed to Step 2. If success → skip to "Using the API" section.

Step 2: Create Strava Application

Guide user to create API application:

  1. Open https://www.strava.com/settings/api
  2. Fill application form:
    • Application Name: Any name (e.g., "My Fitness App")
    • Category: Choose any
    • Website: http://localhost
    • Authorization Callback Domain: localhost
  3. Note Client ID and Client Secret (will be entered in setup)

Step 3: Run Interactive Setup

python3 scripts/setup_oauth.py

Script will:

  1. Ask for Client ID and Client Secret (secret input hidden)
  2. Open browser for Strava authorization
  3. Guide user to copy redirect URL
  4. Exchange code for tokens automatically
  5. Store secrets in system keychain (secure)
  6. Store metadata in ~/.strava/metadata.json
  7. Verify connection

Step 4: Verify Setup

python3 scripts/strava_api.py athlete

Should display athlete profile JSON.


Using the API

CLI Commands

# Get athlete profile
python3 scripts/strava_api.py athlete

# Get athlete stats
python3 scripts/strava_api.py stats

# List recent activities
python3 scripts/strava_api.py activities --limit 20

# Get specific activity
python3 scripts/strava_api.py activity 12345678

# Explore segments in area
python3 scripts/strava_api.py segments "37.7,-122.5,37.8,-122.4" --type running

# Raw API request
python3 scripts/strava_api.py raw GET /athlete/zones
python3 scripts/strava_api.py raw POST /activities --data '{"name":"Test","sport_type":"Run","start_date_local":"2024-01-15T10:00:00Z","elapsed_time":1800}'

Python Import

from scripts.strava_api import StravaClient

client = StravaClient()  # Auto-loads from keychain, refreshes token if needed

# Get athlete
athlete = client.get_athlete()

# List activities
activities = client.list_activities(per_page=50)

# Create activity
new_activity = client.create_activity(
    name="Morning Run",
    sport_type="Run",
    start_date_local="2024-01-15T07:30:00Z",
    elapsed_time=1800,
    distance=5000
)

Token Management

Tokens refresh automatically. Check status:

# Check token status
python3 scripts/refresh_token.py --status

# Force refresh
python3 scripts/refresh_token.py --force

Decision Tree

User Request
├── First time / "setup strava" → Run setup_oauth.py
├── "Get my profile/stats" → strava_api.py athlete/stats
├── "List my activities" → strava_api.py activities
├── "Get activity details" → strava_api.py activity {id}
├── "Create/log activity" → client.create_activity()
├── "Update activity" → client.update_activity()
├── "Get training data/streams" → client.get_activity_streams()
├── "Find segments nearby" → strava_api.py segments {bounds}
├── "Export route" → raw GET /routes/{id}/export_gpx
└── "Token expired" → Auto-handled by client

Quick Reference

| Item | Value | |------|-------| | Secrets storage | System Keychain (secure) | | Metadata location | ~/.strava/metadata.json | | Base URL | https://www.strava.com/api/v3 | | Rate limits | 100 req/15min, 1000 req/day | | Token lifetime | 6 hours (auto-refresh) |

References:

Sport Types

  • Running: Run, TrailRun, VirtualRun
  • Cycling: Ride, MountainBikeRide, GravelRide, EBikeRide
  • Other: Swim, Walk, Hike, Workout, WeightTraining, Yoga

Error Handling

| Status | Meaning | Action | |--------|---------|--------| | 401 | Token expired | Auto-handled by client | | 403 | Insufficient scope | Re-run setup_oauth.py | | 404 | Resource not found | Check ID | | 429 | Rate limit | Wait 15 min |