Cover illustration for Uncovering hidden intensity creep in a longevity running routine

What the AI found

“Your Tuesday recovery run averaged 151 bpm on hills, meaning 41% of your supposed zone 2 volume was spent in zone 3, driving your Thursday fatigue.”

Before

9 hours logged, resting heart rate drifting upward

After

One 4-minute weekly audit of true zone 2 volume

The same system, three states — real screens, not a screenshot

1Starting
Apple Notes
  • Tue: 11km road run (felt sluggish, legs heavy)
  • Thu: 60m indoor spin (RHR elevated +4bpm at wake)
  • Sat: 18km long run (avg 146 bpm, planned <138 bpm)
  • Note: Why is recovery score dropping despite 8h sleep?
2Working
Claude

Prompt

Here is a CSV export of my last 12 aerobic runs with 10-second heart rate samples and elevation. My ceiling for Zone 2 is 138 bpm. Identify every instance where I exceeded this ceiling for more than 90 consecutive seconds, calculate the percentage of total time spent above it, and flag any geographic or pacing commonalities.

Here is a CSV export of my last 12 aerobic runs with 10-second heart rate samples and elevation. My ceiling for Zone 2 is 138 bpm. Identify every instance where I exceeded this ceiling for more than 90 consecutive seconds, calculate the percentage of total time spent above it, and flag any geographic or pacing commonalities.

AI

Across 12 sessions, you logged 9h 14m. However, 3h 48m (41.2% of total time) was spent above 138 bpm. This was not random cardiovascular drift: in 11 out of 12 runs, heart rate exceeded 148 bpm within 180 seconds of hitting any grade steeper than 3.5%, primarily on the Tuesday river loop. Your body was performing mid-tempo intervals during designated active recovery.
3Implemented
Apple Health

3h 40m / wk

True Zone 2 Volume

51 bpm

Avg RHR (7-day)

< 4%

Threshold Leakage

IndividualHacks Pass in use

Uncovering hidden intensity creep in a longevity running routine

How a runner used a targeted prompt pass to audit hidden threshold leakage in her aerobic longevity routine.

A 44-year-old architectural consultant and masters runner, Northern Europe

Tools used

The real tools used here — swap any for your own equivalent. Each links to how we’d set it up.

4 min readWellness & AI editorial
1

Before anything was set up

Nine hours of endurance work were logged faithfully each week into Apple Health, intended strictly for mitochondrial efficiency and lifelong cardiovascular fitness. Yet waking heart rate was creeping upward, from 52 to 56 bpm over two months. Workouts felt heavier instead of lighter. She suspected insufficient sleep or dietary protein, but her spreadsheet notes offered only vague impressions and conflicting subjective ratings. The raw workout files sat unexamined in her phone.

Apple Notes
  • Tue: 11km road run (felt sluggish, legs heavy)
  • Thu: 60m indoor spin (RHR elevated +4bpm at wake)
  • Sat: 18km long run (avg 146 bpm, planned <138 bpm)
  • Note: Why is recovery score dropping despite 8h sleep?
2

Hacks Pass, doing its job

Using the structured prompt provided in the daily pass, she exported her raw interval logs from Apple Health as a structured text table. She dropped the data into Claude on a Sunday afternoon. Instead of offering generic advice on sleep hygiene or supplements, the model correlated heart rate directly against the route profile, pinpointing an unperceived 41% leakage into zone 3 and 4 caused by keeping pace constant on slight hills.

Claude

Prompt

Here is a CSV export of my last 12 aerobic runs with 10-second heart rate samples and elevation. My ceiling for Zone 2 is 138 bpm. Identify every instance where I exceeded this ceiling for more than 90 consecutive seconds, calculate the percentage of total time spent above it, and flag any geographic or pacing commonalities.

Here is a CSV export of my last 12 aerobic runs with 10-second heart rate samples and elevation. My ceiling for Zone 2 is 138 bpm. Identify every instance where I exceeded this ceiling for more than 90 consecutive seconds, calculate the percentage of total time spent above it, and flag any geographic or pacing commonalities.

AI

Across 12 sessions, you logged 9h 14m. However, 3h 48m (41.2% of total time) was spent above 138 bpm. This was not random cardiovascular drift: in 11 out of 12 runs, heart rate exceeded 148 bpm within 180 seconds of hitting any grade steeper than 3.5%, primarily on the Tuesday river loop. Your body was performing mid-tempo intervals during designated active recovery.
3

The finished system, running on its own

The routine now runs on automatic pilot every Sunday evening. A simple export from Apple Health passes into Claude via an automated shortcut. Within seconds, a four-line summary isolates any heart-rate creep across her endurance sessions. She no longer second-guesses her recovery days or adds junk volume. Her baseline resting heart rate settled back down to 51 bpm over six weeks, simply because her easy training sessions finally stayed strictly easy.

Apple Health

3h 40m / wk

True Zone 2 Volume

51 bpm

Avg RHR (7-day)

< 4%

Threshold Leakage

-88%

Unintended Zone 3 time

4 mins

Weekly review time

-4 bpm

Resting heart rate delta

Apple HealthData source

Consolidates continuous optical heart rate and GPS elevation from daily road sessions without manual logging.

ClaudeWeekly pattern audit

Handles large tabular text dumps reliably without hallucinating timestamps or miscalculating zone percentages.

These are the tools used in this story. Any can be swapped for an equivalent you already trust.

Read the daily protocol pass

This story runs on Hacks Pass. The tools and prompts above are the real build — swap any tool for your own equivalent and follow the same steps.

Suggested for you

Based on what you've been reading — always learning.

See all →