borz.io about log blog projects

Dev Update June 9

MCPs

A couple weeks ago my friend and college roommate Cody Pinto shot me a message that the Strava MCP was available. I’ve been doing a bunch of training for work recently, and it seemed like a good excuse to play around with a large, personal dataset.

For those unfamiliar with Strava, it’s an app you use to track all sorts of physical activities, but for me it’s been the single place I’ve logged every run or bike ride I’ve done since 2016 when I decided to start training for a marathon. Years since I stopped doing that, I still use it consistently.

Strava’s also notable for introducing some extremely non-useful AI features like it’s extremely odd planning system that has you train for a marathon by simply running the distance faster than you would on race day (https://x.com/imgeeksonfeet/status/2015672649012355165) or its post workout analysis that loves to tell you that your effort was solid. I didn’t have a ton of faith in their latest offering, but given that I was able to call it myself maybe there’d be something interesting…

Time to make a lot of calls

Okay, time to just pull it by hand.


Graphs

Average Miles

Avg miles per week (quarterly)

Loading chart…


Pace Over Time

Avg pace (min/mile, quarterly)

Loading chart…


Weather Conditions

Workouts by weather condition

Loading chart…


Where I’ve Run

US States

States with at least one run

Loading map…

World

Countries with at least one run

Loading map…