Every change to an evidence grade, every change to a commercial relationship, and every substantive correction, with its date and the source it can be checked against. Newest first.
This log starts on August 11, 2026 and does not pretend otherwise. Entries below that date are changes recorded in the site’s own build history — grading-system changes, content launches and monetisation changes — reconstructed from dated markers in the repository, each named under the entry. No past correction has been invented to make the record look longer. Read the corrections policy for what gets logged and how.
| Entries in this log | 12 |
|---|---|
| Grade change | 2 |
| Disclosure | 4 |
| Policy | 2 |
| Content | 4 |
| Corrections recorded to date | 0 — see the note above |
Magellan published its author, grading, editorial-policy and corrections pages today, and this log with them. It starts empty on purpose: no correction, retraction or grade reversal that predates this entry has been reconstructed, because reconstructing one from memory would be exactly the kind of unsourced claim this site exists to avoid. Everything below this line is a dated change recorded in the site’s own build history — content launches, grading-system changes and monetisation changes — not a corrections backlog.
From today, every evidence-grade change and every substantive correction is written here as it happens.
A Fullscript dispensary trading as “Magellan Longevity” went live, earning a 15% product margin on orders placed through it. Dispensary buttons render only on products that already hold a top-tier evidence grade — the grade gates the button, not the other way round. The relationship is listed in full on the editorial policy page.
The site’s newsletter capture went live the same day.
A Google AdSense publisher account was added and the site submitted for review. No advertising renders until the account is approved, and ad placement is programmatic: Magellan does not choose which advertisers appear, and an advertiser buying an ad slot buys nothing else.
An affiliate relationship with the testing laboratory OmegaQuant was approved (affiliate ID 1034). It renders as a direct-lab link on the blood-and-age-tests buyer’s guide. Disclosed on the editorial policy page.
The Tech Desk published 25 stories measuring consumer devices and home diagnostics against the laboratory equipment they claim to replace — wearable sleep staging against polysomnography, cuffless blood pressure, over-the-counter continuous glucose monitors, epigenetic age-test reliability, and the recovery-tech category. Several are deliberately negative. All are listed on the author page.
The site’s monetisation configuration was consolidated into a single block, with every channel disabled until its value is filled in. That is why the display-advertising entry above says “configured, pending approval” rather than “live”: the account exists and renders nothing.
Paid coaching and consultation booking were removed from the site and their routes retired. Magellan no longer sells clinical services of any kind, and reading the site does not create a doctor–patient relationship. This removes the one relationship in which the site’s author could have been paid directly for an individual recommendation.
Until this date, devices inherited a single grade from the compound or category they were filed under. That was wrong for hardware: nine wearables shared one entry, but an FDA-cleared ECG feature and an unvalidated proprietary “readiness” score are not the same claim. 26 products — wearables, home tests and recovery devices — were regraded individually against FDA clearance, validation quality and price-versus-evidence.
The clearest case is the Oura Ring Gen 4, which now carries a grade that praises its cardiac accuracy against an ECG reference and states plainly that its composite scores have no published validation against any reference standard.
The scale gained a sixth grade — Tested — did not work — for interventions with adequate human evidence pointing at no benefit, as distinct from interventions with thin evidence, which stay at Preliminary. It is used by exactly one product today: SYB Clear Blue Light Glasses, after a 2023 Cochrane review pooled 17 randomized trials and found blue-light-filtering lenses may make no difference to eye strain, probably make little or no difference to visual acuity, and have indeterminate effects on sleep. The product stays in the catalogue at that grade rather than being deleted. See the blue light monograph.
34 recipes went live, each built around a compound with human data — dietary nitrate, potassium, cocoa flavanols, spermidine, sulforaphane, urolithin A precursors — and each linking the studies behind the claim rather than asserting a benefit. Listed on the author page.
Nine long-form articles published, including the urolithin A write-up that states in its own summary that two of its randomized trials missed their primary strength and walking-distance endpoints, and the NMN-versus-nicotinamide-riboside comparison. Publishing the misses inside the article summary, rather than at the bottom, is the house style.
The first thirteen long-form articles went live across topical pain relief, biological-age testing, senolytics and recovery hardware, including the topical-analgesics comparison and the piece on why biological-age clocks disagree with each other.
Grade change means a product moved on the six-point evidence scale, or the scale itself changed. Disclosure means a commercial relationship started, changed or ended. Policy means a rule about how this site operates changed. Content means a body of work was published.
A correction — a claim that was wrong and has been fixed — is logged with the class defined in the corrections policy, and if it could have changed what a reader did, a dated notice also appears on the page itself. Found something wrong? How to report it.