How the site works.

Every peptide page is produced by two passes. The drafter reads the primary literature for that compound — typically 80–150 papers, mostly via PubMed — and writes the deep-dive sections (mechanism, evidence, dose, sourcing, safety). The verifier then independently checks every cited number against the source, flags any prose that conflicts with a banned-words list, and surfaces voice-consistency issues.

Anything the verifier flags goes back to the drafter or to a manual reviewer. Pages don’t ship until the verifier signs off. The pipeline catches fabricated citations, made-up numbers, and the worst kind of confident-but-wrong claim.

What counts as a source, ranked from strongest to weakest:

1. 1. Randomized controlled trial · published, peer-reviewed, with the PMID linked
2. 2. ClinicalTrials.gov NCT registry · primary registration with verifiable status
3. 3. FDA / EMA / PMDA labeling · official prescribing information
4. 4. Systematic review or meta-analysis · if higher-quality primary trials exist, we cite those instead
5. 5. Open-label or single-arm human study · cited as such, never elevated to RCT-equivalent
6. 6. Animal study · cited only when no human data exists, with an explicit "rodent only" species tag
7. 7. Community / off-label consensus · cited as “community protocol,” never as evidence

When we say “the trial used X mg,” we mean a tier-1 source. When we say “the community uses Y mg,” we mean tier 7. The two are different things and we never pretend otherwise.

Every peptide gets a 1–5 evidence dot rating. The scale is calibrated against human trial quality, not animal data:

• 5 ●●●●● · FDA-approved with multiple phase 3 trials (Tirzepatide, Semaglutide, Liraglutide)
• 4 ●●●●○ · Phase 3 complete or approved for narrow indication (Tesamorelin, PT-141, SS-31 Barth)
• 3 ●●●○○ · Phase 2 with positive readout (Retatrutide, Cagrilintide, Survodutide)
• 2 ●●○○○ · Phase 1 / open-label / preclinical with consistent signal (MOTS-c, GHK-Cu)
• 1 ●○○○○ · Animal data only, community use only (BPC-157, TB-500, KPV, Khavinson tetrapeptides)

Animal-only peptides earn a 1, not a 0 — there’s sometimes real signal in the rodent data. But a 1 is also a warning that what you read on Reddit isn’t a substitute for what hasn’t been tested in humans.

Every dose number that appears on the site has been reviewed against published trial protocols and tagged with a species (human, rodent, extrapolated, in-vitro, unknown) and a phase (approved, phase 3, phase 2, phase 1, preclinical, none).

For peptides where the species is “rodent” or weaker — meaning no published human trial dose exists — the Stack Builder refuses to pre-fill a doseand the safety check disables the ratio comparison. The user has to enter their own number, and we tell them why we won’t guess for them.

Of the 59 compounds in the catalog, the trial-dose review has tagged each compound with its evidence species. The breakdown is visible on the library page right rail.

Every quantitative claim is sourced. Every clinical-trial reference includes the PMID, NCT, or DOI when available. Where we cite a community protocol, we say so explicitly — the words “community-extrapolated” or “Reddit consensus” appear by design, not by accident.

We don’t cite second-hand. If a podcast mentions a study, we read the study and cite the study, not the podcast.

• Daily · automated PubMed + ClinicalTrials.gov + FDA RSS scan for new content tagged with any of the 59 catalog peptides. Candidates land in an editorial queue.
• Weekly · 5–10 news items reviewed, written, and published to /news.
• Quarterly · every peptide page is re-reviewed against the latest literature. The “next review” date is on each page.
• Ad hoc · any FDA approval, CRL, or significant trial readout triggers an immediate update to the affected peptide page.

When we’re wrong, we say so publicly. Every correction is logged on /corrections with the original version preserved. We don’t silently rewrite. Quietly fixing errors is what vendor-affiliated sites do; we’re trying to be the opposite of that.

If you find an error, email corrections@yesitakepeptides.com. We respond within 48 hours.