mitr is not therapy. It cannot diagnose anything, it cannot treat anything, and it should not be the only place you turn when things are hard. What it can do is sit with you on an ordinary day, and offer one old idea, clearly explained, that might help you carry it.
You tell mitr what is on your mind, in your own words. It listens, and it answers with something from the Bhagavad Gita, in plain language. A conversation takes about five minutes.
It is meant for an ordinary day, not for a crisis. If today is not an ordinary day, please find a helpline near you.
We think a short daily practice may help some people feel steadier. That is a wellbeing claim, not a medical one.
We are not claiming that mitr treats anxiety, depression, or any other condition. We are not claiming it is clinically proven. We are not claiming it works for everyone, and we do not know yet who it works for.
[TBD] A short, accurate account of the research this rests on, described without overreach, plus what the five domains of the Peace Compass are and how they are used. Must be written from the handbook and the peace compass spec, not from memory, and must not overstate what the literature supports. No number appears here unless it can be evidenced.
Adults, eighteen and over. mitr is not built for children or teenagers, and it is not built for anyone under a doctor's care for a mental health condition. If that is you, mitr is not a substitute for the person treating you, and it is not a second opinion.
The answers are generated by an AI model. Nobody writes them in advance, and nobody reads your conversation to reply to it.
That means it can be wrong. It can misread what you meant. It can state something about the tradition that is not quite right, and it can say something about a verse that a scholar would correct. We test for this and we keep finding things to fix.
It is also not a person. It does not know you the way a friend does, and it cannot notice that your voice sounds different today.
[TBD] This section previously said mitr has no streaks and nothing designed to make stopping feel like failure. That is no longer accurate: streaks are expected in the product. What we will and will not do around habit and guilt needs restating once Shyam and Shy have settled it against Research Principle P-04, which forbids streaks and guilt mechanics. Until then this page makes no claim either way.
It will not pretend to be a person, and it will not pretend to be a substitute for one. The research on this is not ambiguous: AI can add to human care, and it does not replace it. Chatbots in particular are poor at noticing a crisis, which is why mitr does not try to be the thing you reach for in one.
Write to team@bewithmitr.com, or see our support page.