WholisticHuman

Everyone's guessing.

Sensitive people — those dealing with histamine intolerance, MCAS, eczema, IBS, anxiety, chronic fatigue, or general reactivity — live in a constant guessing game. A symptom appears and they try to mentally trace back 24–48 hours of inputs: what they ate, what they put on their skin, what they were exposed to.

Memory is unreliable, the lag between cause and effect varies by reaction type, and multiple inputs interact in ways that aren't obvious in the moment.

Existing tools don't solve this. Food diaries, symptom trackers, and skincare logs are all siloed. None of them connect inputs to outputs across categories, and none surface patterns over time.

One log. Everything in it.

WholisticHuman captures both inputs — food, drinks, supplements, medications, topicals, environmental exposures — and outputs — mood, energy, skin state, congestion, pain, digestion — in one unified timeline.

The person who suspects something.

Someone who believes that things they consume, apply, or encounter are affecting how they feel — but can't reliably identify what, when, or why. They may or may not have a formal diagnosis. They are motivated to track but have likely tried other apps and quit due to friction or lack of insight.

How it has to feel.

What ships first.

• Voice Logging Speak a log entry naturally. Claude parses it into structured data. User confirms before saving.
• Quick-Add Personal item library built over time. Frequent items surface for one-tap entry.
• Unified Timeline All entries chronological, with inputs and outputs visually distinct from each other.
• Library Add, edit, and categorize personal items: food, drink, supplement, medication, topical, environment.
• Basic Insights Symptom heatmap by day. A foundation for the correlation engine to come.
• Time-Aware Voice Voice entries support retroactive logging — "breakfast earlier" or "dinner yesterday" are parsed into inferred timestamps relative to now. Low-confidence times are flagged for user confirmation before saving. Inferred timestamps are visually marked in the timeline.
• Ambient Environment Daily pollen and air quality data auto-pulled by location (Open-Meteo pollen API + AirNow AQI — both free, no key required). Appears as a passive context card at the top of each day's timeline. Not manually logged — always there. Stored in the data model for future correlation use.

Intentionally deferred.

Kept out to keep V1 buildable and the habit testable before adding complexity.

How it's built.

What happens from here.

1. Create Supabase project and define database schema
2. Write Cursor build prompt from this spec
3. Build and deploy V1 to Vercel
4. Hand to real user — observe logging behavior for 30 days
5. Evaluate whether correlations surface anything meaningful before scoping V2