VibeCheck is the AI companion you can actually hand your clients — because you supervise every part of it.

You write the house rules.

Tone, topics, what the companion may encourage, what it must refuse, and when it must escalate. A general-purpose AI companion follows the conversation wherever it goes. VibeCheck follows a clinician-defined role — and stays inside it.

The client enters with the house lights on.

Your client opts in with you, in session. They know what the companion does, what you see, and what it will never do. No surprise surveillance, no fine print doing the consent work a conversation should.

A brief, not a flood.

The trend, the check-ins, entries in their own words, sudden-gain flags, and anything they wanted you to know walking in — assembled before every session. You don't live inside a chat log. VibeCheck organizes the material. You make it clinical.

Urgency needs a real door.

Crisis language routes to established crisis resources and flags you immediately, on escalation paths set with you at onboarding. The companion does not promise rescue, impersonate a clinician, or create an unseen on-call obligation for the therapist. A companion that can't recognize its limits isn't safe to hand a client. This one was built by a clinician who knew that first.

An AI therapist pretends the relationship belongs to the model.

A supervised companion keeps the relationship with you. Blueprint's public petition demands that client-facing mental-health AI be "supervised by licensed professionals" (blueprint.ai, reviewed July 13, 2026). VibeCheck was built to that standard from day one: companion, never therapist; supervised, never freelancing.

Common questions.