the evidence

where the practices come from.

Every technique inside sanctuary is drawn from a studied research tradition. This page names each one, links the surface where it lives, and says in plain language what the research found. It also says what we will not claim.

what this page is, and is not.
The techniques are drawn from studied traditions. sanctuary itself is a new product and has not been independently studied. We would rather tell you that plainly than imply otherwise. When we say a practice is evidence-informed, we mean the underlying technique has a research lineage we can name, not that sanctuary's version of it has been through a clinical trial. If a therapist or a journalist is reading this looking for the fine print, this is the fine print, at the top of the page.

the practices, mapped honestly

Eight practices, eight traditions. For each one: what it is inside sanctuary, where the technique comes from, and one sentence on what the research actually found. The research studied the techniques. It did not study sanctuary.

the physiological sigh
in sanctuary
A double inhale through the nose followed by a long, slow exhale, offered in the grounding room when the body is spiking and words are not landing yet.
the tradition
Cyclic sighing research, including Balban and colleagues' 2023 study at Stanford, part of a broader literature on slow-exhale breathing and vagal activity.
what the research found
In that study, a few minutes of daily cyclic sighing improved mood and lowered resting breathing rate more than mindfulness meditation did over the same month. sanctuary applies the technique; sanctuary was not in the study.
urge-surfing
in sanctuary
The core of the protest-interrupt walk: when the urge to text, call, or check their profile hits, you ride the wave of the urge instead of acting on it or fighting it.
the tradition
Mindfulness-based relapse prevention, developed by Alan Marlatt and colleagues for addiction, applied here to contact urges after a breakup.
what the research found
In this tradition, urges are treated as waves that crest and pass on their own, usually within minutes, and practicing observing an urge without acting on it is taught as a way of letting it pass. Applying this to heartbreak is sanctuary's adaptation, not something Marlatt's trials tested.
the unsent letter
in sanctuary
An expressive-writing encounter, the letter you never send: you write everything you would say to them, and the letter goes nowhere by design.
the tradition
James Pennebaker's expressive writing paradigm, studied since the 1980s across hundreds of experiments.
what the research found
Writing about emotional upheaval for around fifteen to twenty minutes on a few consecutive days produced modest but repeatable improvements in mood and health markers in many studies. The effect sizes are modest, and we would rather say that than inflate them.
five-senses grounding
in sanctuary
Also in the grounding room: naming what you can see, hear, touch, smell, and taste, one sense at a time, to bring attention back into the present room.
the tradition
Sensory grounding as used in trauma-informed care, where clinicians teach it as a first-line tool for dissociation, flashbacks, and spiraling.
what the research found
Grounding is a widely taught clinical skill with a strong practice tradition rather than a large trial literature of its own; clinicians use it because directing attention to the senses reliably interrupts rumination in the moment. We are honest that this one rests more on clinical consensus than on trials.
resonance breathing
in sanctuary
A visual pacer in the meadow that slows breathing to roughly six breaths per minute, with nothing to achieve and nothing tracked.
the tradition
The heart rate variability biofeedback literature, including the work of Paul Lehrer and Richard Gevirtz on resonance-frequency breathing.
what the research found
Breathing near six breaths per minute maximizes heart rate variability in most people, and meta-analyses of HRV biofeedback report reductions in self-reported stress and anxiety. sanctuary borrows the pacing, not the biofeedback hardware, and does not measure your heart.
one hinge for an empty sunday
in sanctuary
tonight helps you place one small, doable action into an evening or a weekend that grief has hollowed out: a walk, a call, a meal cooked instead of skipped.
the tradition
Behavioral activation for low mood, developed through the work of Peter Lewinsohn, Neil Jacobson, and later Christopher Martell and colleagues.
what the research found
In trials for depression, scheduling small, values-linked activities performed comparably to full cognitive therapy for many people. sanctuary uses the technique at a much lighter dose than those trials did, and we do not claim the trial outcomes as our own.
naming the move, not the person
in sanctuary
Throughout the demo and the harbor, sanctuary uses four words for what a nervous system does after rupture: protest, withdraw, repair, freeze. The frame names the move you are in, never labels who you are.
the tradition
Attachment theory, from John Bowlby's work on separation and protest onward, through Mary Ainsworth and decades of adult attachment research.
what the research found
Separation from an attachment figure reliably produces a protest response, an urge to restore contact, before the system reorganizes; the pattern is well documented across the lifespan. Attachment styles describe patterns, not fixed identities, and sanctuary deliberately avoids typing people.
co-regulation and honest presence
in sanctuary
the grove is where travelers are quietly alongside each other, and dove is a steady presence in the hardest hours. One honesty note here, because it matters: dove is an AI presence and is always labeled as one. She is not a person, and sanctuary never pretends she is.
the tradition
Co-regulation and social baseline research, including James Coan's work on how the presence of another person changes how the brain handles threat.
what the research found
In Coan's handholding studies, the brain's threat response was measurably quieter when a trusted person was present; humans regulate distress more efficiently together than alone. That research is about human presence. Whether an honestly labeled AI presence carries any of the same weight is an open question, and we treat it as one.

what sanctuary will never do

Evidence about the practices is half the picture. The other half is what a product built for hurting people refuses to do. We call this the First Law: absence never costs anything. These are design commitments, not marketing copy, and a clinician can verify every one of them by simply using the product.

We hold this line because the same reward loops that make apps sticky are the loops heartbreak already has its hands on: checking, refreshing, waiting for the ping. A product for this moment in someone's life should not run on the mechanics of the problem.

the honest boundary

sanctuary is not therapy, not a medical device, and not crisis care. It is a self-directed space built on evidence-informed practices, and it works best alongside human support, not instead of it. If you are working with a therapist, sanctuary can sit beside that work; our page for therapists explains how, and our safety stack documents the guardrails in full.

If you are in crisis, sanctuary is the wrong tool, and it will tell you so. Call or text 988 (the Suicide and Crisis Lifeline, US, 24/7), text HOME to 741741 (Crisis Text Line), or find your country's line at findahelpline.com. Crisis resources are never gated behind a signup, a paywall, or any other step, anywhere in the product.

plain answers

is sanctuary therapy?
No. sanctuary is not therapy, not a medical device, and not crisis care. It is a self-directed space built on evidence-informed practices drawn from studied traditions. It can sit alongside therapy, and it can be a starting point for someone who is not ready for therapy yet, but it does not replace a licensed clinician.
has sanctuary itself been studied?
No. The techniques inside sanctuary come from research traditions that have been studied, in some cases for decades. sanctuary's specific implementation has not been independently studied, and we do not claim otherwise. If that changes, this page will say so.
is dove a person?
No. dove is an AI presence and is always labeled as one. sanctuary never presents her as a human, a therapist, or a clinician, and she will tell you what she is if you ask. Human connection inside sanctuary happens in the grove, and even there, honesty about what is human and what is not is a hard rule.
what happens if someone is in crisis?
Ordinary conversation suspends and the person is routed to the 988 Suicide and Crisis Lifeline (call or text 988 in the US), the Crisis Text Line (text HOME to 741741), and findahelpline.com for international users. Crisis resources are never gated. The full behavior is documented on the safety stack.
what does evidence-informed mean here?
It means each practice is drawn from a named, studied research tradition, and we tell you which one. It deliberately does not mean clinically proven or evidence-based treatment, because those phrases would imply sanctuary itself has been through clinical trials, and it has not.
If you are in crisis right now
988 Suicide & Crisis Lifeline
Call or text 988. Available 24/7 in the US. Free, confidential support from trained human counselors.
Crisis Text Line
Text HOME to 741741. Trained crisis counselors, any time.
International
findahelpline.com: crisis lines in your country.
sanctuary cannot help in a crisis. No practice on this page is crisis care. If you or someone you know is in danger, close this page and call one of the numbers above. We will be here when you come back.