Awareness tools for BFRBs: how a webcam app fits alongside therapy

> Important. This post is not medical advice. Body-focused repetitive behaviors (BFRBs) — trichotillomania, dermatillomania, onychophagia, and others — are clinical conditions, and the gold-standard treatment is habit reversal training (HRT) or ComB therapy with a trained clinician. If you are looking for help, the TLC Foundation for BFRBs has a clinician directory and free family resources. The tool described below is one optional adjunct, not a replacement for treatment.

What "awareness training" actually means

The first phase of any HRT-based BFRB treatment is awareness training. Most BFRB episodes happen in a state clinicians call "automatic mode" or "trance" — your hand has reached your hair, scalp, or skin before the conscious part of your brain has caught up. You can't apply a competing response to a behavior you didn't notice, so the entire treatment is gated on awareness.

In a clinical setting, awareness training looks like:

• Stimulus discrimination. Learning to identify the precursor — the urge, the body position, the typical context — that reliably precedes a pulling/picking episode for you specifically.
• Self-monitoring logs. Writing down (often in real time) episodes, contexts, and what preceded them.
• Behavioral mirrors. Sometimes, in office, the clinician will literally mirror back what they see your hands doing while you talk.

Self-monitoring logs are one of the more demanding components, because the population most affected by BFRBs is also the population least able to spot the behavior unprompted. The whole problem is that you don't notice. Asking you to notice consistently enough to fill out a log is a chicken-and-egg situation.

Where an external signal fits

This is the gap a passive awareness tool can fill. An external signal — a beep that fires when your hand approaches your face — does the noticing for you, in real time, without requiring you to remember anything. It's the same mechanism a clinician's behavioral mirror provides, just continuously and at home.

Used well (and we'll get to "used well" in a second), it can:

• Catch trance episodes, which are the hardest ones to surface in self-monitoring logs.
• Provide data, like time-of-day patterns or contextual triggers, you can bring back to your clinician.
• Cue a competing response. Once your clinician has helped you build one, the external signal becomes the training-wheels reminder until the response is automatic.

A few things it cannot do:

• It can't replace the personalized trigger-and-pattern analysis of HRT/ComB. Two people with trichotillomania can have completely different treatment plans.
• It only catches hand-to-face/head movements. If your pattern doesn't go through that path (e.g., scalp picking with a tool, picking sites that aren't on the face), it won't help.
• It can't help with the affective drivers — anxiety, boredom, sensory regulation needs — that motivate the behavior in the first place. Those need their own work.

How to use it well

Based on our own usage, conversations with users, and a small number of clinicians who've tried integrating it into their HRT protocols, here's what tends to go well and what doesn't.

Do: introduce it after the first sessions of HRT.

Most clinicians prefer that the awareness/self-monitoring phase happens with their guidance first, so you've identified your specific triggers and the tool is being deployed against a target you understand. Showing up to session 1 with two weeks of beep data is fine; replacing session 1 with the app is not.

Do: pair it with a competing response your clinician has helped you build.

Without a competing response, the beep becomes a notification that you've already done the thing you're trying not to do — which can be demoralizing. The point is to insert a behavior, not to issue verdicts.

Don't: use it as a punishment.

The beep is information. It is not a moral failing. If you find yourself getting angry at the beep, take that to your clinician — it's data about your relationship with the behavior, and it matters.

Don't: use it during high-affect or vulnerable moments.

If you've identified that your pulling/picking spikes during a specific high-stress context (a difficult conversation, after a hard work email), the beep in that moment can intensify rather than help. Pause it. The tool serves you, not the other way around.

Do: mute the audio, surface the visual count.

Some users (especially in shared spaces) prefer the visual counter without the beep. The behavior data is most of the value; the beep is a delivery mechanism that's optional.

Privacy

This is non-negotiable for BFRB tooling. Anything you'd hesitate to type into a Google search or ask your therapist about should not run on a remote server.

no_touch_face runs entirely on your computer. The webcam feed is processed locally with on-device models. Nothing is uploaded, ever. There is no account, no telemetry, no sync. The session log is a local plaintext file you can read, edit, or delete with rm.

If a future version offers cloud sync (e.g., for clinician-shared logs), it will be opt-in, end-to-end encrypted, and clearly separated. The default will always be local.

Try it, talk to your clinician

The app is free. It's intentionally bare-bones. If your clinician has questions about how it works, point them at this post (or just have them email us). And if you don't yet have a clinician — start there. The app is a useful adjunct; it isn't enough on its own, and it isn't trying to be.