Editorial Standards

Voice

Target: understated, technical, direct. The reader already lifts or runs. We give them the formula, the source, and the limits — not a motivational pep talk.

• Understated, technical, direct — lifter-to-lifter and runner-to-runner. Numbers first, no hype.
• No corporate jargon (`leverage`, `empower`, `seamless`, `solutions for`, `industry-leading`).
• No AI slop (`delve`, `tapestry`, `robust` unless fault-tolerant, `unlock`, `deep dive` as noun).
• No fitness hype (`crush your goals`, `transform your body`, `your fitness journey`, `no pain, no gain`, `trust the process`).
• No prescriptive coaching voice — this is a reference hub, not a coaching app.
• No personal PR logs, no before/after narratives, no transformation diaries.
• Body-neutral language: `energy needs` not `calories to lose weight`; ranges, not judgments.

Claims — what we can and can't say

• Never publish invented numbers as executed measurements. If a guide cites a meta-analysis, the paper is linked (PubMed/DOI).
• Never claim a cadence we cannot commit to. No 'weekly cadence', 'coming soon', 'new tools every month'.
• Never imply regulatory coverage we do not have — we do not claim EU MDR or FDA SaMD framing.
• Hedge honestly: `roughly`, `in the typical case`, `the Kouri 1995 sample of 157 men suggests`, `as of 2026`.
• Surface the limits: Mifflin-St Jeor was derived on American adults; FFMI's 25 'ceiling' is an observational claim (n=157), not biological law; Riegel loses accuracy outside 5K–marathon.

Citation hierarchy

• 1. Peer-reviewed academic

  PubMed PMID, DOI, or original journal URL — preferred for any 'research shows' claim.

• 2. Federation / governing body

  IPF, USAPL, WADA, UKA, USATF, World Athletics for sport-specific standards.

• 3. Textbook

  Named edition with ISBN or chapter URL — acceptable for settled physiology basics.

• 4. Named-authorship aggregator

  Greg Nuckols' MASS, Eric Helms' references — only where the underlying research is unsettled.

• Never

  Vendor blogs, anonymous listicles, unlinked 'studies show', myFitnessPal-tier aggregators.

Meta descriptions + titles

Meta descriptions target 140–155 characters with the primary keyword in the first 60 characters. No `no signup / no tracking` tails — they burn SERP real estate without keyword value. Article titles are ≤60 characters for Google SERP fit (aim ≤55). Enforced by scripts/audit-meta-descriptions.mjs.

Compliance

Tools compute deterministic math from user inputs. Output is educational; consult a physician before changing training load, nutrition, or medication. Injury-risk tools — direct 1RM testing, VO2-max field tests, elite-percentile strength standards — render an elevated-caution banner. No personalised output; nothing profiles the reader. No fabricated regulatory lineage — we do not claim EU MDR or FDA SaMD coverage.

Corrections

Dated and append-only at /corrections/. Never silently rewritten. When a methodology page, tool output, or article is fixed, a dated entry is added to the corrections log.

What doesn't ship

• Personal PR logs, before/after photos, transformation narratives, coaching prescriptions.
• Unverified study counts or effect-size numbers — link the paper or cut the claim.
• Sponsored content without the `Sponsored` / `Affiliate` chip rendered inline.
• Any UI claim we cannot defend on an honest audit. Boring-but-true over punchy-but-risky.
• Tools without a methodology page linked from them.

Conflict of interest

AI Fit Hub is a self-funded, independent publication. There are no active sponsors, affiliate links, or paid placements. The current state and the rules that apply if that ever changes are documented at /sponsor-disclosure/.

Contact

Factual corrections, methodology disputes, and sponsor inquiries: see the contact details at /about/.