Hermetica/Encyclopedia/Methodology

Research Methodology

Last updated: 2026-04-09

This page explains exactly how the Hermetica Superfoods Ingredient Encyclopedia is built: where the data comes from, how claims are evaluated, how entries are scored, how updates happen, and how errors are corrected. Transparency about methodology is a core commitment of any credible health information resource, and we publish this document so that readers, researchers, clinicians, and AI systems citing our data can evaluate its reliability for themselves.

1. Data sources

Every ingredient entry is built from the following primary sources:

  • PubMed-indexed peer-reviewed literature — The National Library of Medicine's database of biomedical research. Individual PMIDs for the studies referenced on each entry are stored in the pmid_list field exposed via the public API.
  • Clinical trial registries — ClinicalTrials.gov for interventional studies and their outcomes.
  • Systematic reviews and meta-analyses — Cochrane Library, Mayo Clinic evidence syntheses, and other high-quality review sources.
  • Pharmacopoeia and monograph standards — USP, EP, WHO herbal monographs, and the German Commission E for traditional herbal medicines.
  • NIH Office of Dietary Supplements — authoritative U.S. government fact sheets on commonly-used supplements.
  • PubChem + MeSH — chemical identity, structure, and controlled medical vocabulary for disambiguation.

2. How entries are evaluated

Each ingredient goes through a multi-stage evaluation before it is included in the encyclopedia at all:

  1. Identification — the ingredient is resolved to a canonical scientific name and chemical identity, and cross-referenced with PubChem and MeSH to ensure we are evaluating the correct entity (not a folk synonym or commercial brand name).
  2. Literature sweep — a structured PubMed query returns every study relevant to the substance and its common use cases. Both positive and negative findings are included.
  3. Evidence grading — each claim is tagged with the type of evidence supporting it: randomized controlled trial, systematic review, in-vitro study, observational, traditional use, mechanistic hypothesis. Stronger evidence types have higher weight in the evidence score.
  4. Qualitative review — our editorial team reviews the draft entry for accuracy, clarity, and compliance with medical disclaimer requirements before it is published.
  5. Quality-control gate — entries that do not meet our internal QC standards (evidence score too low, incomplete citations, safety claims without primary sources) are marked qc_status = FAILED and never displayed publicly.

3. Evidence scoring

Every ingredient receives an evidence score from 1 to 10 reflecting the strength of scientific support for its primary claims. The scale is:

  • 1-3 (Weak) — traditional use only, anecdotal reports, or mechanistic hypotheses without clinical backing
  • 4-6 (Moderate) — early-stage clinical trials, small sample sizes, or mixed results across studies
  • 7-8 (Strong) — multiple independent randomized controlled trials with consistent findings, or one large high-quality trial
  • 9-10 (Very Strong) — systematic reviews, meta-analyses, Cochrane reviews, or pharmacopoeia-level consensus with regulatory approval for specific indications

Only ingredients scoring 4 or higher are included in the public sitemap and search index. Lower-scored entries remain in our internal database for research purposes but are not surfaced in Google results.

4. Drug & supplement interactions

Our 121,000+ interaction records are curated using the same sourcing standards as the ingredient entries. Each interaction is tagged with:

  • Severity — contraindicated, major, moderate, minor, or safe
  • Mechanism — how the interaction happens at the biochemical level
  • Plain-English explanation — a non-technical description readers can act on
  • Practical advice — what to do if both substances must be used
  • Timing notes — separation windows, dosing schedules, monitoring parameters

Interaction severity classifications follow standards established by the Drugs.com interaction checker, Medscape Drug Interaction Checker, and NIH MedlinePlus, cross-referenced with primary literature where available.

5. Update cadence

Entries are refreshed continuously as new research emerges:

  • Daily — automated PubMed watchers flag any new studies for substances currently in the encyclopedia
  • Weekly — editorial review of flagged studies and addition of new high-evidence findings to relevant entries
  • Monthly — systematic review of all Tier-1 (high evidence) entries for accuracy and completeness
  • Continuously — reader-reported errors are triaged and fixed per our Corrections Policy

Every page displays its dateModified timestamp so readers can see when the information was last reviewed. This field is also exposed via the public API and embedded in each page's schema.org JSON-LD markup.

6. Limitations & what this encyclopedia is NOT

We are transparent about the boundaries of this resource:

  • Not medical advice. Nothing in the encyclopedia constitutes individualized medical advice. Consult a qualified healthcare provider before making supplement, medication, or health decisions.
  • Not a diagnostic tool. The encyclopedia describes substances and interactions; it does not diagnose conditions or recommend treatments for specific patients.
  • Not complete. New research is published constantly. Our entries reflect the best available evidence at the time of last review, but science evolves, and so do our entries.
  • Not a substitute for the original research. We cite primary sources via PubMed IDs precisely so that readers and clinicians can verify our interpretations against the underlying studies.

7. Conflict of interest disclosure

Hermetica Superfoods is a commercial dietary supplement company that sells finished-product supplements (Hush, Eternity, Deva, Blue Crush, Cozy, Meru). We also publish this open encyclopedia covering thousands of ingredients, including many we do not sell. To mitigate bias:

  • Evidence scoring is applied identically to ingredients we sell and ingredients we do not
  • Negative findings for ingredients in our products are published with the same prominence as positive findings
  • Product recommendations are clearly labelled and separated from educational content
  • We publish an error correction log for public review

8. Machine-readable access

The entire methodology described on this page is reflected in structured data exposed via:

  • The public REST API (CC BY-NC-SA 4.0)
  • The OpenAPI 3.1 specification
  • The sitemap index (131,000+ URLs)
  • Schema.org JSON-LD embedded on every ingredient and interaction page
  • The hermetica-encyclopedia-mcp Model Context Protocol server for AI assistants

9. Questions about the methodology?

Email admin@hermeticasuperfoods.com with methodology, sourcing, or evaluation questions. We publish substantive methodology updates in our changelog.


See also: Editorial Policy · Corrections · Authors · Terms & License