An open-source, evidence-based, fully-cited knowledge base for maximizing muscle growth — and a free coaching app built on top of it, live at hypertrophybible.com.
Two goals: the knowledge base aims to be the gold standard for hypertrophy science; the app aims to be the gold standard for coaching that applies it. Free, no ads, no premium tier, no selling data.
Every substantive claim here is backed by a real, web-verified study (with a DOI or PMID) and
carries an A–D evidence grade. Where the science is uncertain, we say so — that honesty is the
point. See STYLE.md for the authoring rules and the design spec for the full rationale.
A mobile-first PWA: anonymous one-tap start → a single "Today" card → one exercise at a time → a
derived recap → progress inferred from your logs (per-muscle volume vs. the KB's landmarks,
estimated-1RM trends, energy balance from bodyweight — no calorie counting). Optional passwordless
email backup syncs across devices. One codebase runs on Node locally and Cloudflare Workers + D1 in
production. To run it, see app/README.md.
The knowledge base has two synchronized halves:
content/— readable prose, organized into ten pillars. New to the gym? Start at Getting Started.data/— the same knowledge as machine-readable JSON (validated against schemas indata/schemas/). This is the contract the future app consumes.
Citations live once in citations/registry.json and are referenced by key
from both prose and data. A human-readable bibliography is generated at citations/registry.md.
| # | Pillar | What it covers | Status |
|---|---|---|---|
| 00 | Foundations | Physiology & mechanisms of hypertrophy | ✅ Complete |
| 01 | Training Variables | Volume, load, frequency, failure, rest, tempo, ROM, exercise selection | ✅ Complete (exemplar) |
| 02 | Muscle Guides | Per-muscle anatomy, function, exercise selection, volume | ✅ Complete |
| 03 | Programming | Splits, periodization, progression, deloads | ✅ Complete |
| 04 | Nutrition | Energy balance, protein, timing, supplements | ✅ Complete |
| 05 | Recovery | Sleep, stress, deloads, injury management | ✅ Complete |
| 06 | Individualization | Training status, age, sex, genetics, populations | ✅ Complete |
| 07 | Tracking | Metrics, RPE/RIR, progress assessment | ✅ Complete |
| 08 | Myths | Common misconceptions, corrected | ✅ Complete |
| 09 | Getting Started | Total-beginner on-ramp: gym basics, equipment, safety, first sessions, the full-arc roadmap | ✅ Complete |
Legend: ✅ Complete · 🚧 In progress · 🔲 Outlined (table of contents + stubs only)
| Grade | Meaning |
|---|---|
| A | Strong — consistent across multiple RCTs/meta-analyses. Act on it confidently. |
| B | Moderate — supported but with some inconsistency or limited populations. |
| C | Limited — few studies, mixed results, or mostly untrained/short-term. |
| D | Mechanistic/practice-based — physiology or coaching consensus, little direct RCT support. |
Model-based numbers (e.g. volume landmarks) are estimates, graded honestly — never presented as fact.
npm install # one dependency: ajv (JSON Schema validation)
npm run validate # every data/ + examples/ file validates against its schema (+ landmark ordering)
npm run check # citation integrity + data cross-reference integrity (exercises/muscles/progressions resolve)
npm run check-refs # data cross-reference integrity only
npm run build-bib # regenerate citations/registry.md from registry.json
npm run derive # demo the derive-metrics engine over examples/
npm run test-derive # unit tests for the derive-metrics engine
npm test # everything: validate + check + derive testsThe KB is designed to power a self-learning training app. The data contract and feature-derivation layer are built and tested (the app UI/backend is downstream):
data/schemas/{onboarding-profile,workout-session,daily-checkin,body-metric}.schema.json— the data contract for what the app collects.examples/— validating sample data (an onboarding profile, a two-week workout log, daily check-ins, body metrics).tools/derive-metrics.mjs— the "derive-don't-ask" engine: computes per-muscle weekly volume (vs the KB's own MEV/MRV landmarks), estimated-1RM progression, energy balance inferred from the bodyweight trend (no calorie counting), objective proximity-to-failure, and personal-baseline readiness — all from low-burden primitives.docs/data-and-learning-spec.md— the full strategy: signal vs noise per stream, confidence tiers, the accuracy playbook, and the prior→personalize→aggregate→ feed-back-into-the-KB self-learning architecture.
Ten pillars complete, audited, expanded, and depth-reviewed. The knowledge base spans 99 content pages backed by 87 web-verified citations (each confirmed to exist via PubMed, with 20 independently cross-checked against Crossref), plus a data layer of 16 muscle files, 59 exercises, 5 program templates, 2 progression rules, and 8 supplement entries — all passing schema validation, citation integrity, and cross-reference integrity checks.
A dedicated Getting Started pillar now provides the total-beginner on-ramp the science pillars assumed away — gym basics, equipment operation, a glossary, safety (spotting, failing/bailing, red-flag symptoms), first sessions, gym anxiety, and a full-arc roadmap from "never trained" to "approaching your genetic ceiling" — plus advanced pages (long-term planning, weak-point specialization, advanced plateau-breaking, the genetic-ceiling journey) so the KB honestly spans the entire journey.
A full audit corrected one factual error (older-adult training volume), reconciled internal
inconsistencies, and closed a data-integrity gap by adding a referential-integrity checker
(npm run check-refs) and landmark-ordering validation.
A subsequent evidence-quality pass deliberately diversified the source base (no single research group exceeds ~18% of citations, across 60+ distinct lead authors), added recent syntheses (2021–2025), and refined several positions to match current evidence — the proximity-to-failure and lengthened-training debates are now presented as genuinely unsettled, and load, protein, and frequency claims are corroborated by multiple independent meta-analyses. New Foundations coverage (regional & architectural hypertrophy, connective-tissue adaptation, and a "reading the evidence" methodology page) rounds out the physiology.
Verify it yourself:
npm install && npm run validate && npm run checkThe app is built and live at hypertrophybible.com — onboarding,
KB-derived sessions, one-exercise-at-a-time logging, derived progress, offline support, and passwordless
email backup, on Cloudflare's free tier. Active work: making the coaching itself gold-standard —
generating plans directly from the KB's volume landmarks and exercise database, explaining the "why,"
and capturing effort/readiness. See app/README.md.
MIT © 2026 Nathan Doyle. Educational content — not medical advice.