
What the AI found
“Your afternoon fatigue spikes exclusively on cycle days 19-23 when lunch contains under 22g protein, not after nights with low REM.”
Before
Random 3pm crashes blamed on poor sleep hygiene
After
One calendar tag and three 15-minute adjustments
The same system, three states — real screens, not a screenshot
- 14 Oct: 3pm brain fog again, barely made the sprint retro
- 18 Oct: fine today, slept 6h?
- 22 Oct: exhaustion hit at 2:45pm. maybe iron? cut caffeine?
- 23 Oct: bought magnesium glycinate and blue-light glasses
Prompt
Analyze these three text logs: my cycle start dates, Apple Health sleep exports, and my lunch notes for the past 60 days. Look for correlations with the days marked 'fatigue crash'. Identify any recurring thresholds or timing patterns.
Here is my last 60 days of cycle start dates, Apple Health sleep scores, and my lunch notes. Why am I crashing at 3pm on random weekdays?
AI
Your crashes are not random. Across 60 days, all seven severe afternoon fatigue events occurred between cycle days 19 and 23. On those specific days, your logged lunch contained under 22g of protein (averaging 11g). On luteal days where protein exceeded 26g, your 3pm energy score remained above 7/10 regardless of sleep duration.0 of 14 days
Luteal 3pm dropoff
28g avg
Luteal lunch protein
4 min
Weekly log review
Finding the quiet threshold behind a monthly energy slump
How a ten-day email course isolated a specific hormonal pattern hiding behind daily lunch habits.
A 37-year-old product lead, Northern Europe
Tools used
The real tools used here — swap any for your own equivalent. Each links to how we’d set it up.
Starting state
Before anything was set up
For six months, every third week brought two or three days of disabling afternoon brain fog. She kept a chaotic Apple Notes file filled with contradictory hypotheses: too much oat milk, poor bedroom ventilation, low ferritin, or late-night screen exposure. She had bought three supplements, tried twenty-minute walks after lunch, and downloaded two separate cycle-tracking apps that gave vague horoscopic advice like 'honour your energy today'. The data existed, but stayed locked in separate siloes.
- 14 Oct: 3pm brain fog again, barely made the sprint retro
- 18 Oct: fine today, slept 6h?
- 22 Oct: exhaustion hit at 2:45pm. maybe iron? cut caffeine?
- 23 Oct: bought magnesium glycinate and blue-light glasses
Working state
10-Day Challenge, doing its job
On Day 4 of the 10-Day Challenge, she exported sixty days of cycle timestamps from Apple Health alongside her sleep data and pasted them into Claude with a simple prompt provided by the module. Instead of generic advice about rest and hydration, the model cross-referenced the cycle phases against her informal meal descriptions. It surfaced a quiet, exact threshold: the luteal phase demanded more midday protein than her routine grain bowls provided.
Prompt
Analyze these three text logs: my cycle start dates, Apple Health sleep exports, and my lunch notes for the past 60 days. Look for correlations with the days marked 'fatigue crash'. Identify any recurring thresholds or timing patterns.
Here is my last 60 days of cycle start dates, Apple Health sleep scores, and my lunch notes. Why am I crashing at 3pm on random weekdays?
AI
Your crashes are not random. Across 60 days, all seven severe afternoon fatigue events occurred between cycle days 19 and 23. On those specific days, your logged lunch contained under 22g of protein (averaging 11g). On luteal days where protein exceeded 26g, your 3pm energy score remained above 7/10 regardless of sleep duration.Use case implemented
The finished system, running on its own
Ten days in, the system runs quietly in the background. Her cycle phase sits as a small tag in Apple Calendar, cross-referenced against a simple two-column protein tally in Apple Notes. There are no fifty-dollar powders or complex regimens. On days nineteen through twenty-three, she brings leftover roast chicken or tempeh instead of a soup. The afternoon slump, once treated as an inevitable personal failing or a symptom of poor sleep discipline, failed to appear across her last two cycles.
0 of 14 days
Luteal 3pm dropoff
28g avg
Luteal lunch protein
4 min
Weekly log review
What an outside observer would notice
4 min
Weekly logging time
100%
3pm calendar blocks kept
The stack — build it yourself
Handles long, unstructured text transcripts and food logs with high fidelity and nuanced synthesis.
Keeps core biometric data locally stored while allowing clean, private CSV exports.
Zero friction for quick meal photos and two-sentence daily energy notes.
These are the tools used in this story. Any can be swapped for an equivalent you already trust.
Go deeper
Do this yourself
Start the 10-Day Challenge
This story runs on 10-Day Challenge. The tools and prompts above are the real build — swap any tool for your own equivalent and follow the same steps.