Editorial policy
How we work, what we won't do, and how to verify what you read here.
This is the binding policy document behind every page on The Plug Guide. What we publish, what we refuse to publish, how we evaluate evidence, what triggers a correction, and how to catch us if we get something wrong.
Who this site is for
We write for the person searching at 2 AM for whether a pill is going to kill them when mixed with the cocktail they just had. That reader has already decided. They are not asking us for permission. They are asking for the most accurate information we have, in their language, in plain words, with the trade-offs spelled out.
Every editorial decision on this site is anchored to that reader. If a sentence would be more useful in a peer-reviewed paper than in front of someone with a pill in their hand, the sentence does not belong here.
Where the money comes from (and doesn't)
The site is funded out of our own pockets. There are no advertisers, no affiliate partners, no sponsorships, no paid placements — past, present, or planned. The hosting bill is small and we cover it ourselves so the writing stays uncompromised. If that ever changes, the change will be announced here, dated, and signed.
No display, no programmatic, no sponsored placements ever.
We don't earn anything from vendor reviews. They're independent.
No "supported by", no native ads, no influencer placements.
From idea to publication
- Research
Every page begins with a primary-source pass. For substance pages, that means peer-reviewed pharmacology plus the canonical community references (PsychonautWiki, TripSit, Erowid). For scam alerts, it means WHOIS history, archive.org diffs, payment-flow forensics, and at least two corroborating threads of evidence.
- Draft
A human writes the draft. Not an LLM, not a paraphraser, not a translation tool dressed up as a writer. The draft cites every non-obvious claim inline and tags any claim we are not certain about for the fact-check pass.
- Fact-check
A second pair of eyes — the editor who did not write the piece — re-reads the draft against the cited sources. Disputed claims either get a stronger source or get cut. Anything weakly-sourced that survives this pass gets an inline hedge so the reader can see our confidence level.
- Publish
The article ships with the date stamped, the author byline tied to a real profile page, and (where applicable) a "Sources & further reading" section that lets readers retrace our reasoning. We never silently update a published article without leaving a corrections note.
What counts as evidence
Peer-reviewed pharmacology; primary public records (WHOIS, court filings, regulatory enforcement); on-chain analysis from named providers; multiple independent first-hand reports each with order ID, timestamp and shipping label.
Community wikis (PsychonautWiki, TripSit, Erowid) read against primary sources; named consumer-protection bodies (IGJ, Watchlist Internet, AVROTROS Opgelicht!); harm-reduction collectives reporting under their own name.
Single anonymous forum reports; screenshots that cannot be independently verified; trip reports older than 12 months without a corroborating fresh datapoint; reports from competing vendors or affiliates.
Anything that originates from an LLM without a primary source behind it. Anything from an informant whose safety could be compromised by publication. Anything we cannot trace back to a verifiable artefact within seven days of receipt.
When we publish a scam alert
Vendor reputations are easy to destroy and hard to rebuild. We treat publication like an investigative journalist would treat a small-claims allegation: multiple corroborating threads of evidence, no single anonymous reports as primary basis, no rumour, no competitor-supplied claims taken at face value.
The full vendor-evaluation methodology — including what counts as strong vs weak evidence and exactly what every scam alert must include — lives at the methodology page. We update it whenever the rules change.
Read the vendor-evaluation methodologyWhen we get something wrong
Errors get corrected in the article, dated, and labelled at the foot of the page so a reader who saw the original version can see exactly what changed and when. Material errors (a wrong dose, a wrong vendor verdict, a wrong identity) are also called out at the top of the article for at least 30 days.
We never silently rewrite a published claim. If you spot something wrong, the contact page is the fastest route — include the URL, the specific claim, and a primary source if you have one. Most corrections land within a week.
AI-written content has no place here
We do not use large language models to draft articles, fill out substance pages, write product reviews, or compose translations. Every byline on this site corresponds to a human who took responsibility for what's underneath it. AI tools are used internally for tasks like spell-check or formatting passes, never for generating content that ships under an editor's name.
This rule exists because LLMs hallucinate doses, invent vendor histories, and confidently cite sources that do not exist. None of those failure modes are tolerable on a harm-reduction site. If we ever change this policy, the change will be announced on this page with a date and signed by both editors.
What we will not review and why
We do not run display ads, programmatic ads, sponsored placements, or affiliate links. We do not accept free product from vendors. We do not write reviews of vendors we have personal commercial relationships with — and we disclose any prior relationship at the top of any review where one existed.
If a vendor sends us product unsolicited, the product is destroyed and the unsolicited contact is documented in our internal log. If a vendor offers payment to influence a review, the offer becomes the lead of the next scam alert.
What happens to tips you send us
Tips are read by one of the two editors. If we publish based on a tip, we anonymise the source by default — you have to explicitly opt in to be named. The tip channels we operate are listed on the contact page; we deliberately do not advertise Signal or PGP because we do not run those channels at the moment, and listing a channel we cannot honour reliably is worse than listing none.
We do not retain raw evidence beyond what is needed to corroborate the published claim. Once a scam alert is live, the supporting WHOIS dumps, archive.org snapshots and on-chain trails are kept; identifying details about the tipster are not.
How to verify you are reading the real authors
Every byline on this site links to a profile page that names the editor, lists their beat, languages, prior work, and the contact channels we actually run (a tip email and our public Telegram channel). If a byline you see anywhere else claims to come from us but does not link back to one of those profile pages, it is not us.
The two contact channels below are the canonical ones for both editors. We do not DM readers from any other handle, we do not run Signal or PGP at the moment, and we will never ask you to send money, crypto or personal documents — anyone messaging you in our name asking for any of those things is an impersonator.
Lead investigator · scam network exposés
Tips: [email protected]
Public channel: t.me/theplugguide — we do not DM readers; messages claiming to be from us in any other channel are not us.
Lead writer · harm reduction & substance guides
Tips: [email protected]
Public channel: t.me/theplugguide — we do not DM readers; messages claiming to be from us in any other channel are not us.
Questions, tips, corrections
Anything in this document you want challenged, anything we got wrong, anything you can corroborate with a primary source — the contact page is the fastest route. Tips can also go straight to [email protected] or to our public Telegram channel.