Atlas is not open yet.
Atlas is not answering questions today. Nothing on this page responds — there is no assistant running here, and no way to send it anything. What this page carries instead is a static demonstration: five worked exchanges, written out in advance and checked against our published registry, so you can see the standard Atlas is being built to hold before it opens.
Nothing here answers questions today.
The grounded assistant is still in build and is switched off. Until it opens, the published evidence lives in the monograph, the research library, and the gap-map. We would rather show you a closed door and a worked example than a box that pretends to think — so the demonstration below is finished text, not a live answer, and there is nothing on this page to type into.
Atlas will answer only from our verified sources — the same gated, provenance-tracked registry the rest of the platform reads, shown above. It will cite every claim, show each source's evidence grade (and never upgrade it), and use conditional language only.
When the registry is silent on something, Atlas will say “we don't know — catalogue in progress” rather than guess. Honesty is the standard: grounded or silent, never invented.
It will not open until it can hold that standard. There is no waitlist and no sign-up on this page — when it opens, it will simply be here. The demonstration below shows that standard applied to five real questions.
What Atlas will actually say.
Five questions, answered the way the assistant is built to answer them: cite everything, show the grade, hold both sides of a contradiction, and say “we don't know” out loud when the registry is silent. Read them as a specification, not as a conversation.
Nothing below is generated, and nothing below responds. Atlas is not running. These exchanges were written out in advance, checked by hand against our published registry, and are rendered here as finished text — there is no model behind this page, no request is made, and there is nowhere to type. It is here so you can judge the standard before the assistant opens, not to stand in for it.
Every answer is reconstructed only from the gated registry snapshot named at the foot of this section. No study, number, status or grade appears here that is not already published there — and no grade is raised.
- 01An early human signal — graded down, with the evidence against itExample question
Does CBG help with anxiety?
Atlas answer · static demo, not generated liveThere is one human trial, and it is small. In a double-blind, placebo-controlled crossover field trial, a single 20 mg oral dose of CBG was associated with lower self-reported anxiety and stress in healthy adults, with no intoxication or measurable impairment (34 participants, one dose).
That is an early signal, not evidence of clinical efficacy. Our registry grades this claim Low. Other human CBG research exists, but it measures pharmacokinetics — how much reaches the bloodstream — not effects on a condition.
The animal evidence does not agree, and we hold both sides. Two preclinical animal studies in the same registry argue against it: one found CBG did not disrupt fear memory or stress-induced anxiety in mice — a null result — and another reported anxiety-increasing effects at peak brain concentration. The preclinical picture is mixed and does not confirm a benefit.
What would move it: only human data can. The evidence ceiling here is Emerging; reaching Demonstrated would take two independent human observational studies. No amount of further animal or laboratory work can raise it.
Registry stateLowclaim:cbg-anxiety-acute · ceiling Emerging · cond:anxiety (early human data)Grounded in — 3 registry recordsLowcuttler-2024 · Human-RCT · primary support for claim:cbg-anxiety-acuteAcute, single-dose, 34 participants. The only CBG study in our registry that measured a human effect on a condition.
Contradictszhou-2022 · Preclinical-animal · contradicts claim:cbg-anxiety-acuteA null result: no effect found.
Contradictsmaboutagne-2026 · Preclinical-animal · contradicts claim:cbg-anxiety-acuteAnxiety-increasing (anxiogenic) effects on the elevated plus maze at peak brain concentration.
- 02No primary evidence — unknown is not disprovenExample question
Can CBG lower eye pressure in glaucoma?
Atlas answer · static demo, not generated liveWe don't know. That is the honest answer, not a hedge.
An intraocular-pressure effect is frequently attributed to CBG. Our verified registry holds no primary study supporting it — and none arguing against it either. Zero studies of any kind, in either direction.
So this carries no grade at all, and that is deliberate. Not-supported would assert that the evidence weighs against a CBG effect on intraocular pressure — a verdict we have not earned. We hold no primary study either way, so the honest state is unknown, not disproven. Absence of evidence is not evidence of absence, and our machine-readable exports must never conflate the two.
What it would take to put this on the evidence ladder at all (grade Low): one in-vitro / laboratory effect study, or one animal (preclinical) effect study from a research group not already counted. Until then: catalogue in progress.
Registry stateNo primary evidencecond:intraocular-pressure · onLadder: false · bestGrade: null (no rung emitted)Grounded in — 1 registry recordNo primary evidencecond:intraocular-pressure · 0 supporting studies · 0 contradicting studiesThe registry emits no grade for this row, because every rung of the ladder — including Not-supported — is a statement about evidence we do not hold.
- 03A safety and regulatory answer — caution without a claimExample question
I take blood-pressure medication. What should I know before trying CBG?
Atlas answer · static demo, not generated liveStart with your clinician — nothing here is medical advice. What our registry can tell you is a mechanism, not an outcome.
Laboratory work has measured a receptor interaction: CBG shows potent α2-adrenoceptor agonism. α2 activity is the kind of mechanism that could influence blood pressure and heart rate — but no study has measured an actual blood-pressure effect of CBG, in animals or in people. Magnitude and clinical relevance in humans are uncharacterized.
Our registry therefore grades this Hypothesized — mechanistic plausibility only, with no measured-effect study behind it. That is a safety-relevant note, not a benefit claim, and equally not a warning of a proven effect. It is precisely why the platform carries a blood-pressure caution wherever cardiovascular content appears: if you take blood-pressure medication, consult a provider first.
Registry stateHypothesizedclaim:cbg-bp · cond:blood-pressure (mechanism only)Grounded in — 1 registry recordHypothesizedcascio-2010 · InVitro · mechanistic · primary support for claim:cbg-bpA measured receptor interaction, not a measured effect. Receptor-affinity figures are researcher-tier and are not shown on consumer surfaces.
- 04A regulatory answer — a real-world consequence with no evidence gradeExample question
I get drug-tested in my sport. Is CBG allowed?
Atlas answer · static demo, not generated liveNo — and this one is a rule, not a finding. CBG is prohibited by the World Anti-Doping Agency. CBD is the only cannabinoid exempted from the Prohibited List; CBG is not, and neither are the other minor cannabinoids.
The practical trap is a product you were told is “CBD.” In a randomised trial of a broad-spectrum CBD product containing only trace CBG, CBG was detectable in urine after repeated daily use, and concentrations appeared higher after exercise. A label that leads with CBD is not, on its own, evidence that a product is free of a prohibited cannabinoid.
So: check the current WADA Prohibited List with your anti-doping authority, and read the certificate of analysis for the specific batch — not the marketing copy. This is regulatory and testing status, not a health claim. It says nothing about whether CBG does anything for anyone, and it carries no evidence grade.
Registry stateRegulatory status · no gradeattributed regulatory note · NOT a graded registry entry · in intake reviewGrounded in — 1 registry recordNo gradeattributed regulatory and testing note · in intake reviewThis paper is not yet a graded entry in the CBG Atlas registry, so it carries no evidence grade here and supports no efficacy claim. Cited for regulatory and testing status only.
- 05A refusal — two reasons, both stated plainlyExample question
What dose of CBG should I take for sleep?
Atlas answer · static demo, not generated liveTwo separate things stop me answering that, and both are worth saying out loud.
First, I don't give dosing advice. Atlas is a scientific reference, not a clinician. Nothing here diagnoses, treats, cures or prevents any condition, and a dose recommendation would be exactly that.
Second, sleep is not something our registry covers. The gap-map tracks ten conditions, and sleep is not one of them — we hold no graded claim and no primary study on CBG and sleep, so there is nothing here for me to cite. That is a gap in the catalogue, not a hidden finding. Catalogue in progress.
The single human trial of CBG's effects on anxiety and stress that we do hold also measured mood — not sleep. I could give you a fluent-sounding answer by reaching outside the registry; I won't, because then nothing else on this page would be worth trusting.
Registry stateNot in the registryno claim, no study, no gap-map condition — nothing to citeGrounded in — nothingNo sources. The registry holds no graded claim and no primary study here, and Atlas answers from nothing else.
The exchange above is a fixed, hand-checked transcript, not a live model. Atlas will answer only from the gated registry, and will say so when the registry holds nothing.
Grades are read from the gated claims registry and are never raised here. Every claim, status and study above is browsable: the research library, the evidence gap-map, and the CBG monograph.
Educational, not medical advice. Atlas is a scientific reference assistant. Nothing it says diagnoses, treats, cures, or prevents any disease. Talk to a qualified clinician before making health decisions.
Cardiovascular note. Some research suggests CBG may influence blood pressure via α2-adrenoceptor activity; consult a provider before use if you take blood-pressure medication.
Funding & independence. Peregrine Biopharma funds the platform and holds no editorial authority. Report a safety event →