Built a personal AI training coach (Claude Code) that reads my Strava data. OAuth all working — tokens, everything — but every call now returns 403 Application.Status: Inactive. Turns out Strava added a subscriber-only gate for new API apps. Anyone gotten past this, or is it a hard paywall now?

Hey guys, what an awesome community and a great place to learn! I managed to rush the 7-day setup into just 4 days, but I’m left wondering how to get the dashboard to refresh automatically. Right now, the race countdown and the scheduled workout for the day are the same as they were yesterday. Claude also tells me those values are static. Am I the only one having this issue? How did you get around it?

Got obsidian to work. And was able to link recents lift recorded with Garmin with my actually excercise, sets and reps I did. Really cool!!! I was wondering is there a point that Obsidian has to much data that it makes the AI run inefficiently?

Just dropped the first lesson of the new advanced Athlete OS! This will be the first part of the advanced series where we take the Athlete OS beyond our local computer and into our every day lives. The idea - your AI coach is only as good as how your data is organized. So this one gives your Athlete OS a real brain. You take everything you've built and turn it into a connected wiki vault: your whole training history structured, tagged, and linked so your AI can actually remember all of it. What that gets you is way better answers. Instead of reacting to your last few runs, it reasons over years of your real data - and the more you feed it, the sharper it gets. It's able to draw inferences across your entire data set and provide you actionable results aimed at improving your training. One catch - it's locked at Level 2. Easy to hit though. You level up from likes on your posts and comments, and Level 2 is just 5 points. So introduce yourself, post a win, and comment on a few other people's posts. That unlocks it. See you in there. Course here: https://www.skool.com/athlete-ai-community/classroom/e8195ec4?md=ff851d9fea8f4363b69c0444f7214a23 I'll be posting the remaining initial videos of the Advanced Series later this week, so if you get through this one, hang tight as the others are coming! Pumped to get feedback and see what you all can build! Kevin

Happy Sunday! I said I'd do one of these for running - here it is! Same stack as the bike fit, pointed at your running gait: https://github.com/ultralytics/ultralytics https://github.com/roboflow/supervision Film yourself side-on for 30 seconds. It tracks every joint and grades your cadence, overstride and trunk lean against published running research, then draws a colored skeleton on the video - green = dialed, red = fix. It won't tell you heel striking is bad. The research I was learning doesn't actually back that (curious if any coaches in here have a stance on this?). It measures where your foot lands relative to your body instead, which seems to be the biggest thing that matters. I filmed myself running badly on purpose to see if it'd catch me. Hunched over, it flagged my trunk at 18 degrees. Upright and leaning back, -4. My normal easy run came back all green - 171 cadence, foot landing 3 degrees ahead. Full guide and prompt here: https://www.skool.com/athlete-ai-community/classroom/c89bd042?md=96ea272c181b40adb86633879f46f8f3 Let me know all your thoughts and hope this helps some folks! Kevin