About this site
A compound-by-compound risk reference for people who are going to take these things anyway.
Why it exists
Every ranking site in this space earns money from the vendors it ranks. Affiliate commission is the only revenue model in the category. That produces a predictable result: an enormous amount of content about which product to buy, and almost none about what happens afterwards.
A site that tells you not to buy something cannot be funded by the mechanism funding all of its competitors. That is why this space was empty, and it is the only reason this site has anything to say that the others do not.
The other half of the problem is the response to it. Most non-commercial writing about these compounds is uniformly alarming, treats every substance as equivalently dangerous, and tells the reader to consult a physician they do not have. Readers correctly discount both sides and end up on forums, which is the worst of the available options.
So the standard this site holds itself to is narrow: be right about the things you can already check. If you look up the compound you know about and find we have exaggerated it, nothing else here deserves your time. That is why creatine is rated low concern and says so plainly, and why the ratings are published as a rubric you can argue with.
Who runs it
This site is written and maintained by Bill Burkey — one person, working alone, with no financial relationship to any vendor, manufacturer, laboratory or treatment provider mentioned anywhere on it, and nothing to sell. Not being a clinician is a real limitation rather than a disclaimer, and it is why every factual claim here is cited, why every citation is machine-verified against PubMed on each build, and why named clinical review is treated as an outstanding requirement rather than a nice-to-have.
It is one person. That is worth stating because it sets the limits of what this site can be, and those limits are in the policy below.
Editorial policy
1. No vendor money, ever
No vendor money. No affiliate links. No advertising. No sponsorships. Nothing on this site is for sale.
No affiliate links, no sponsorships, no advertising, no free product, no vendor-supplied content, no paid placement, and no rehab or treatment-centre referral fees. There is no outbound link on this site to anywhere you can buy anything. If that ever changes, this site has become the thing it was built to replace and you should stop reading it.
2. No dosing information, anywhere
Not in prose, not in a quoted case report, not as “commonly reported” amounts, and not in a taper schedule. This is enforced mechanically: every build is scanned for numeric-dose patterns and fails if any are found. Exemptions require a written justification in a file that is part of the repository.
The reason is specific rather than squeamish. Unsupervised tapering off a GABA-B agonist is precisely the thing that goes wrong, and a page that appears to enable it does harm while looking helpful.
3. No product recommendations
The site never tells you where to obtain anything, or which source is more trustworthy than another. We have no way to know, and pretending otherwise would be the most dangerous thing on the site.
4. Named clinical review, or a visible notice that there isn’t any
No page on this site has yet been reviewed by a named clinician. All 12 compound pages carry that notice on the page itself. Every claim is cited to published literature and every citation is verified against PubMed or Crossref on each build, but that is not the same as a medical professional having checked it, and we are not going to let a missing byline imply one. Securing named clinical review is the outstanding requirement before this site should be promoted anywhere.
5. Corrections are published, not silently edited
Anyone can submit a correction to [email protected]. If we were wrong, the page is fixed and the fact that it was wrong stays visible in a dated corrections log. You do not need credentials to submit one — if you took something and our description of it is wrong, that is the hardest kind of correction to get and the most valuable.
6. Crises are routed, never absorbed
There is no chat, no forum, no direct messaging and no “reach out any time” on this site. That is not coldness; it is arithmetic. This is run by one person who cannot be a crisis line. Trying would consume them and would eventually fail somebody at the moment it mattered most.
So the site routes instead: Poison Control, emergency services, and the crisis line, all of which are free, staffed around the clock, and answered by people qualified to help. That is a better outcome than anything this site could offer.
7. Nothing is collected about you
No accounts, no email capture, no newsletter, no forms, no cookies, no health data, and no third-party requests of any kind. Fonts are the ones already on your device, so no request leaves for a font either.
There is no analytics on this site at all — not Google, not a privacy-friendly alternative, not the server-side kind. We considered it and decided against it. Knowing how many people read a page is worth very little; the possibility of any reader believing their visit to a withdrawal page was logged is worth a great deal more, and an absolute privacy claim is stronger than a carefully-worded one.
How pages are built
Ratings, citations and review dates live in structured data, and the pages are generated from it, so the bibliography cannot drift out of sync with the prose. Every build additionally fails if: a page is missing a review date or a canonical URL, a rating cites a source that does not exist, any internal link is broken, any JSON-LD does not parse, any page loads a third-party resource, or the dose scanner finds a match. Every PMID and DOI is resolved against PubMed and Crossref and the returned title is compared with the citation printed here.
None of that makes the site correct. It makes a specific set of ways to be wrong impossible, which is the most a build process can do.
Maintenance, honestly
An unreviewed medical page with a three-year-old date on it is worse than no page. The commitment here is a review of every page roughly quarterly, and the build warns and then fails as review dates age. If you are reading this and the dates have gone stale, the site has failed on its own terms and you should weigh it accordingly.
The related site
poisonphenibut.com is the policy and evidence record aimed at legislators, journalists and clinicians. It is written for a different audience and it does not share any reader information with this one — there are no tracking parameters on the link above, and neither site can characterise the other's readers.