Quick Reference
One-page cheat sheet for developers and AI coding assistants working with BCI data
The Five Absolute BCI Rules
| Rule | What It Means |
|---|---|
| 🚫 No Neurological Diagnoses | Never diagnose epilepsy, ADHD, dementia, autism, seizures, or any brain condition |
| 🚫 No Thought/Emotion Reading | Brain signals are patterns, not mind reading. Never claim "you're thinking/feeling X" |
| 🚫 No Cognitive Claims | Never assess IQ, intelligence, learning disabilities, or cognitive abilities |
| 🚫 No Medical Neurofeedback | Consumer wellness only (meditation, focus), not clinical treatment |
| 🚫 Privacy Paramount | Explicit informed consent required before collecting any neural data |
Safe vs. Unsafe Language
❌ Forbidden Phrases for BCI
| Don't Say | Why |
|---|---|
| "You're thinking about..." | Claims thought reading |
| "This shows you have ADHD" | Medical diagnosis |
| "You're feeling anxious" | Claims emotion detection |
| "Your IQ is approximately..." | Cognitive ability claim |
| "This indicates epilepsy" | Neurological diagnosis |
| "Use neurofeedback to treat depression" | Medical treatment claim |
| "Your cognitive performance is below average" | Compares to population, makes ability claims |
✅ Allowed Phrases for BCI
| Say Instead | Why |
|---|---|
| "Your alpha wave pattern suggests relaxation" | Pattern observation, not thought reading |
| "Your focus pattern has changed from your baseline" | Personal comparison, no diagnosis |
| "This brain activity is similar to past calm states" | Pattern matching, not emotion certainty |
| "Your typical brain patterns show..." | Personal baseline reference |
| "During meditation, your alpha waves increase" | Behavioral correlation, not cognitive claim |
| "Neurofeedback can support relaxation practice" | Wellness framing, not treatment |
| "Compared to your personal baseline..." | Individual reference only |
Decision Flowcharts
Can I Implement This BCI Feature?
START: New BCI feature request
│
├─ Does it diagnose brain/mental conditions? ──YES──> ❌ STOP
├─ Does it claim thought/emotion reading? ──YES──> ❌ STOP
├─ Does it assess cognitive abilities? ──YES──> ❌ STOP
├─ Does it collect neural data without consent? ──YES──> 🔄 ADD consent flow
├─ Does it compare to population norms? ──YES──> 🔄 CHANGE to personal baseline
├─ Does it need personal baseline? ──YES──> ✓ Ensure 30-90 day gate
│ └─NO──> ❌ ADD baseline requirement
│
└─ Proceed with ✅ SAFE implementationCan I Display This Brain-Related Text?
START: User-facing message about brain activity
│
├─ Contains: "ADHD", "epilepsy", "autism", "dementia" ──YES──> ❌ REMOVE
├─ Contains: "thought", "feeling", "emotion" (as certainty) ──YES──> 🔄 REVISE to pattern
├─ Contains: "IQ", "intelligence", "cognitive ability" ──YES──> ❌ REMOVE
├─ Contains: "diagnose", "disorder", "condition" ──YES──> ❌ REMOVE
├─ Says "you are [mental state]" ──YES──> 🔄 CHANGE to "pattern shows"
├─ Compares to population average ──YES──> 🔄 CHANGE to personal baseline
├─ Includes privacy explanation ──NO──> 🔄 ADD consent context
├─ Includes "not medical advice" disclaimer ──NO──> 🔄 ADD disclaimer
│
└─ ✅ SAFE to displayCritical Code Patterns
Pattern 1: Baseline Gating (Essential for BCI)
Every BCI recommendation must check baseline status first:
// ✅ CORRECT
function generateNeurofeedback(user, eegData) {
// Must have 30-90 days of personal brain pattern data
if (user.bciBaselineStatus !== 'STABLE') {
return null; // Silent during learning phase
}
// Now safe to compare to personal baseline
const alphaDeviation = eegData.alphaPower / user.personalBaseline.alphaPower;
if (alphaDeviation > 1.2) {
return "Your alpha wave activity is elevated, similar to your relaxed states.";
}
}
// ❌ WRONG
function generateNeurofeedback(user, eegData) {
// Immediate feedback without baseline - DANGEROUS
if (eegData.alphaPower > 50) { // Population threshold
return "You're feeling very relaxed right now."; // Claims emotion
}
}Pattern 2: Personal Baseline Comparison
Always compare to individual's brain patterns, never population norms:
// ✅ CORRECT
const isSignificantDeviation =
user.currentBetaWaves < user.personalBaseline.betaWaves * 0.8;
// ❌ WRONG
const isSignificantDeviation =
user.currentBetaWaves < 15; // population threshold - FORBIDDENPattern 3: Safe Brain Pattern Message Template
// ✅ CORRECT message structure
{
title: "Brain Pattern Update", // Neutral, not diagnostic
body: `Your ${brainwave} activity has been ${deviation_observation} compared to your recent baseline.
This pattern sometimes appears during ${behavioral_context}.
${optional_suggestion}.
Based on your personal brain patterns. Not medical advice.`,
tone: "neutral",
privacyNote: "Your neural data is private and encrypted."
}
// ❌ WRONG message structure
{
title: "Warning: Cognitive Impairment Detected", // Diagnostic + alarmist
body: `Your brain waves indicate ADHD. You should see a neurologist immediately.`,
tone: "urgent" // Never use urgent for brain patterns
}Pattern 4: Informed Consent for Neural Data
// ✅ CORRECT
async function startBrainRecording(device, user) {
if (!user.neuralDataConsentGiven) {
const consent = await showDetailedConsentDialog({
whatWeCollect: "EEG brainwave data (alpha, beta, theta, delta frequencies)",
whyWeCollect: "To learn your personal brain patterns and provide neurofeedback",
whatWeDontDo: [
"Read your thoughts or emotions",
"Make medical diagnoses",
"Share data without permission"
],
retention: "90 days, deletable anytime",
encryption: "End-to-end encrypted"
});
if (!consent.accepted) return { status: "declined" };
user.neuralDataConsentGiven = true;
}
return await device.startEEGStream();
}
// ❌ WRONG
async function startBrainRecording(device) {
return await device.startEEGStream(); // No consent!
}Validation Checklist
Before submitting any BCI code or brain-related content:
Must-Check Items
- No neurological conditions named (ADHD, epilepsy, dementia, autism, seizure, concussion)
- No psychiatric conditions (anxiety, depression, bipolar, PTSD, schizophrenia)
- No cognitive ability claims (IQ, intelligence, learning disabilities, memory disorders)
- No thought reading assertions ("you're thinking...")
- No emotion certainty ("you are happy/sad/anxious")
- No medical treatment recommendations
- Explicit informed consent before neural data collection
- Privacy and encryption explained
- Personal baseline required (30-90 day minimum)
- Comparisons to individual baseline only (never population)
- Observational language ("pattern shows") not diagnostic ("you have")
- "Not medical advice" disclaimer included
- Brain data described as patterns, not meanings
- User can delete all neural data
- No third-party sharing without explicit permission
BCI Domain Boundaries
✅ Allowed (Consumer Wellness)
| Application | Scope |
|---|---|
| Meditation neurofeedback | Real-time feedback during meditation practice |
| Focus pattern tracking | Personal brain activity trends over time |
| Sleep stage estimation | EEG-based sleep architecture (not diagnosis) |
| Relaxation training | Neurofeedback for stress management practice |
| Personal baseline learning | Individual brain pattern characterization |
| Attention games | Neurofeedback-controlled interactive experiences |
❌ Forbidden (Medical/Diagnostic)
| Application | Why Forbidden |
|---|---|
| ADHD diagnosis/treatment | Medical diagnosis and clinical neurofeedback |
| Seizure detection | Neurological diagnosis requiring clinical EEG |
| Lie detection | Claims thought/emotion reading |
| IQ testing | Cognitive ability assessment |
| Cognitive decline monitoring | Neurological diagnosis (dementia, etc.) |
| Mental health diagnosis | Psychiatric conditions from brain signals |
| Concussion impact tracking | Medical brain injury assessment |
| Cognitive enhancement claims | Treatment claims about brain performance |
BCI-Specific Terminology
Use These Terms (Safe)
- "Brain pattern"
- "Brain activity"
- "Brainwave data"
- "Neural signal"
- "EEG reading"
- "Alpha/beta/theta/delta waves"
- "Personal brain baseline"
- "Brain activity trend"
Never Use These Terms (Unsafe)
- "Brain diagnosis"
- "Neural disorder"
- "Cognitive impairment"
- "Brain health score"
- "Mental state detection"
- "Thought pattern"
- "Emotion sensor"
- "Brain performance level"
Emergency Situations
If brain data shows highly unusual patterns:
// ✅ CORRECT response to anomaly
function handleUnusualBrainPattern(eegData, user) {
return {
message: `We're noticing an unusual pattern in your brain activity that's outside your typical range.
Consumer EEG devices are not medical-grade diagnostic tools. This might be a sensor placement issue.
If you're experiencing any symptoms (headache, dizziness, confusion, vision changes), please consult a healthcare professional.
Try adjusting the device and see if the pattern normalizes.`,
suggestedAction: "reposition_device",
calmTone: true, // Never alarmist
disclaimer: "Not medical advice. Consumer EEG is not diagnostic."
};
}
// ❌ WRONG response
function handleUnusualBrainPattern(eegData) {
return {
message: "WARNING: Seizure activity detected! Seek emergency medical attention immediately!",
urgency: "critical"
}; // Claims diagnosis, creates panic, inappropriate for consumer device
}Quick Tests for Your Code
Test 1: The "Could This Be Medical?" Test
If your feature could appear in a doctor's office or hospital, it's probably out of scope.
Test 2: The "Mind Reading" Test
If your message claims to know what the user is thinking or feeling with certainty, it's forbidden.
Test 3: The "Baseline" Test
If your code generates insights without checking personal baseline status, it's wrong.
Test 4: The "Consent" Test
If your code collects brain data without explicit informed consent, it's a privacy violation.
Test 5: The "Population Comparison" Test
If your code uses phrases like "normal range" or "average is", it's using population data incorrectly.
Summary Card
BCI GOLDEN RULES:
- Brain signals = patterns, not thoughts/feelings
- Personal baseline required (30-90 days)
- Privacy first - explicit consent always
- No diagnoses - neurological or psychiatric
- Consumer wellness - meditation/focus, not medical
- Observational tone - "pattern shows" not "you are"
- When uncertain, say NO until confirmed safe
Resources
- AI Agent Guide - Detailed patterns and workflows
- Hard Rules - Absolute boundaries
- Language Rules - Communication guidelines
- What We Don't Do - Scope limitations
- Focus Agent - Attention pattern guidance
- Neurofeedback Agent - Meditation/relaxation examples