The Layer

The Body Next

I had six apps telling me things about my own body. None of them talked to each other. The data was real. The architecture was broken.

Jason MurphyMarch 31, 20266 min read

I didn't start this series planning to write about my body.

But here's the thing about a series on architecture and attention: the body is the machine everything runs on. You can optimize every system, streamline every workflow, eliminate every unnecessary cognitive load — and still crater your ability to think if you're not sleeping, not recovering, not fueling right. The body isn't a separate system. It's the substrate.

Which is why it's worth asking how much cognitive overhead you're currently spending to manage it.

The App Graveyard

At some point I counted the apps on my phone that existed to track or inform my body. I stopped at six.

A sleep tracker. A workout logger. A running app. A calorie/macro counter. The native Apple Health aggregator that everything sort of dumped into but nothing actually read from. And whatever Google's version was called, which had data in it from a phone I'd replaced two years ago and which I kept accidentally opening when I meant to open something else.

Six interfaces. Six separate mental models. Six dashboards that needed checking to construct a picture that should have been a single picture.

And none of them talked to each other in any useful way.

My sleep score from the ring didn't inform the workout app's recommendations. The workout app didn't know I'd run 30 miles the week before and might need a deload. The calorie app didn't know my activity level had dropped because I'd been traveling. Apple Health knew everything and surfaced almost nothing useful — it was a warehouse with no librarian.

The data was real. The architecture was broken.

The Hidden Computation

The invisible cost was the integration work.

Every morning I was doing unconscious arithmetic: sleep score plus soreness plus energy level plus what's on the calendar equals what kind of workout makes sense today. That calculation ran in the background of my first hour, pulling from three different interfaces, using data I'd internalized over time because no single app had all of it.

It wasn't a hard calculation. But it was a constant one.

And like all the other micro-decisions that had accumulated into a morning tax — what to eat, what meetings needed prep, what was sitting in email — it was drawing from the same resource pool I needed for actual thinking. The body stuff was supposed to make me sharper. The overhead of managing it was working against that.

Same pattern as the food. Same pattern as the social feeds. Different domain, identical architecture problem.

One Session Per Week

I moved body management to a single weekly session.

Friday evening — after the week is visible and before the weekend recovery — I open Claude and write a status brief. It takes about four minutes:

This week: four workouts (two strength, two runs). Sleep averaged 6.5 hours, quality felt medium, had two nights under six hours Tuesday/Wednesday. Mild left hip flexor tightness since Thursday's run. Caloric intake was lower than target Monday-Wednesday, roughly on target Thursday-Friday. Next week has three travel days with minimal equipment access. Goals: maintain strength base, prioritize recovery before the travel block.

What comes back is a complete plan: training schedule structured around the travel, adjusted intensity for the hip flexor, specific session templates I can run in a hotel gym with minimal equipment, recovery priorities, nutrition targets adjusted for lower activity days.

It's one interaction. It's informed by everything. And it produces something I can actually execute — not a dashboard I have to interpret every morning from scratch.

The daily app-checking mostly stopped. If something feels off, I note it and it goes into next Friday's brief. The calculation that used to run in the background every morning has been moved to a single intentional session where I have full cognitive capacity and full context.

What the Apps Miss

The apps are good at collection. They're mediocre at synthesis.

A sleep tracker knows your HRV and respiratory rate. It doesn't know you have a presentation at 9am tomorrow and need to decide whether tonight's hard interval session is actually a good idea. A calorie counter knows your intake. It doesn't know you're in a training block and running a mild deficit right now is going to cost you in next week's performance. A workout logger knows your lifts. It doesn't know your quad is subtly compensating for the hip flexor thing and you should probably address that before you add load.

All of that synthesis is possible. None of it happens automatically across fragmented apps.

Claude isn't a fitness app. It doesn't replace sensors or tracking. But it's the synthesis layer — the thing that can read all the inputs at once and produce coherent output. That's the gap the apps leave. And it turns out that gap is the most cognitively expensive part of the whole system.

What Carried Over

The pattern that held from meals, from email, from the calendar:

The goal isn't to stop caring about the thing. It's to decide when you'll care about it, and only then.

I'm not less attentive to my health than I was with six apps running. I'm more attentive — because the attention is focused, weekly, from a full state rather than scattered across dozens of micro-checks from a depleted one. The brief I write on Friday is more honest and more comprehensive than any individual app check ever was, because I'm synthesizing the whole week, not just the data point it happened to capture.

The body is the last domain I expected to be an architecture problem. It's been one of the most meaningful ones to fix.

What are you still managing reactively that could be moved to a single intentional session?

Want this for your business?

Tell us what you're building. We'll map out exactly what to build and what it costs.

Start Your Project →

Frequently Asked

What is the problem with using multiple fitness tracking apps?

The problem isn't the apps individually — it's the cognitive cost of managing information across fragmented systems. Each app is its own interface, its own mental model, its own dashboard to check. The data is siloed: your sleep is in one place, your workouts in another, your nutrition in a third. You become the integration layer between systems that don't talk to each other. That integration work is invisible but constant — a background cognitive process running at all times.

How do you use Claude for fitness and health planning?

The approach is to consolidate all health data into a single weekly briefing to Claude, rather than checking multiple apps daily. You describe what happened — workouts completed, sleep quality, energy levels, any soreness or constraints — and ask Claude to synthesize it and produce a plan for the next period. Claude acts as the analysis and planning layer across all your data at once, rather than you performing that synthesis manually across six separate interfaces.

Does this approach work without wearable devices or detailed tracking?

Yes. The architecture works whether your input is detailed biometric data from a Garmin watch or a rough qualitative summary you write in two sentences. Claude can work with whatever you actually have. The goal is to replace the fragmented app-checking behavior with a single consolidated interaction, regardless of how precise the underlying data is.

What is the difference between monitoring health data and making it actionable?

Monitoring is passive — checking a dashboard to see what it says. Making data actionable requires synthesis across multiple variables, pattern recognition over time, and translation into a specific decision. Most fitness apps are excellent at monitoring; they show you the data. Very few are good at synthesis across their own data, let alone data from other systems. Claude fills the synthesis and planning role that the apps were never designed to perform.

Jason Murphy

Written by

Murph

Jason Matthew Murphy. Twenty years building digital systems for businesses. Former CardinalCommerce (acquired by Visa). Now running VibeTokens — AI-built websites and content for small businesses.

The window is open.

It won't be forever.

Start Your Project →