Trail map
Detailed content
1.1~35 min
🫀 What is HealthOS
The core idea: a coach that thinks with your data. Problem, components, architecture, a day in the life, and disclaimers.
Progress0 of 7 · 0%
What is: The difference between the internet’s average and an answer grounded in you.
Why learn: It’s the project’s thesis: specific beats generic.
Key concepts: Specific · memory · mechanism.
What is: A personal coach blueprint for Telegram, with its own database.
Why learn: Defines what you're going to build.
Key concepts: Blueprint · Telegram · agent.
What is: The four inputs the coach cross-references.
Why learn: The value lies in the intersection, not in any one source.
Key concepts: Wearable · blood · DNA · diet.
What is: The sequence that runs with every message.
Why learn: Understanding the snapshot means understanding the system.
Key concepts: Snapshot · agent · Supabase.
What is: A full day on Telegram, from green morning to amber.
Why learn: Shows how today’s choice becomes tomorrow’s number.
Key concepts: Gating · loop · logging.
What is: How your data stays secure by default.
Why learn: Health data is sensitive.
Key concepts: Private · RLS · secrets.
What is: The most important rule + what it costs to run.
Why learn: Doctor in the loop, always.
Key concepts: Non-medical · affordable · WHOOP optional.
1.2~35 min
📈 The body's signals
What each number means: recovery, HRV, RHR, sleep, and the blood markers (HOMA-IR, ApoB, Lp(a)) the coach reads.
Progress0 of 7 · 0%
What is: A measure of how recovered your body is (0-100%).
Why learn: It’s what decides whether to pull or rest.
Key concepts: Green · amber · red.
What is: The variation in time between heartbeats.
Why learn: Indicates recovery tone.
Key concepts: Higher is usually better.
What is: Resting heart rate in beats per minute.
Why learn: Lower usually indicates better health.
Key concepts: Trend > point.
What is: The difference between the sleep you got and the sleep you need.
Why learn: Poor sleep raises cortisol and amplifies everything.
Key concepts: 7-day moving average.
What is: A fasting glucose + insulin index.
Why learn: Tracks insulin resistance.
Key concepts: Energy metabolism.
What is: Particle and risk markers beyond LDL.
Why learn: More predictive than classic cholesterol markers.
Key concepts: ApoB · Lp(a).
What is: Blood pressure and weight measured day to day.
Why learn: Salt can raise blood pressure by 10-15%.
Key concepts: Trend, not a point.
1.3~35 min
🧬 Applied genetics (SNPs)
Genes are dials, not destiny. CYP1A2 (caffeine), APOE, MTHFR, COMT, VDR, and how DNA + tests become coaching rules.
Progress0 of 7 · 0%
What is: A variation in DNA that changes how a gene works.
Why learn: Genes are dials, not destiny.
Key concepts: Variation · expression.
What is: The gene that determines whether you metabolize caffeine quickly or slowly.
Why learn: Decides when and how much coffee to have.
Key concepts: Fast/slow metabolizer.
What is: Gene linked to cholesterol management.
Why learn: Influences saturated fat and risk.
Key concepts: e4 variants.
What is: Gene involved in B vitamin metabolism.
Why learn: Affects methylation and homocysteine.
Key concepts: Folate · B12.
What is: Gene involved in neurotransmitter metabolism.
Why learn: Links stress, focus, and caffeine.
Key concepts: Dopamine.
What is: Receptor that determines your response to vitamin D.
Why learn: Adjusts the supplement dose.
Key concepts: Individual response.
What is: Rules that cross-reference gene + test to determine what to do.
Why learn: It’s how the coach becomes mechanism-aware.
Key concepts: Gene + lab → flag → action.
1.4~35 min
🧠 How AI becomes a coach
The brain (CLAUDE.md), the snapshot, semantic memory (embeddings), advice RAG, risk flags in practice, and limitations.
Progress0 of 7 · 0%
What is: The file with the coach’s profile, goals, and style.
Why learn: That’s where personalization lives.
Key concepts: Profile · style.
What is: A snapshot of the present, read at every turn.
Why learn: It’s what keeps the coach in the current context.
Key concepts: state.py.
What is: Recall by meaning, not exact word.
Why learn: Lets the coach surf old patterns.
Key concepts: Embedding · pgvector.
What is: An expert tip cross-referenced with your numbers.
Why learn: Your data always outweighs generic advice.
Key concepts: RAG · citation.
What is: The DNA+lab rules guiding supplements and alerts.
Why learn: Connects fundamentals to action.
Key concepts: Flag → stack.
What is: Where AI gets things wrong and needs verification.
Why learn: Doctor in the loop, always.
Key concepts: Hallucination · verification.
What is: How the agent stores everything securely.
Why learn: Health data requires care.
Key concepts: Service role · RLS.