# MYND Master Build Blueprint v0.1

**Date:** June 29, 2026  
**Status:** production-intent implementation atlas generated from the verified MYND frontier research package  
**Scope:** iPhone + Apple Watch first, with lab-lane hooks for later external hardware and research export  
**Source of truth:** this blueprint inherits the compact ingest packet, readiness audit, master synthesis, implementation ticket bridge, initial preset matrix, parameter freeze, and build-facing extraction notes.

---

## 1. Product Definition

MYND is an Apple-native personal state-space instrument.

It lets a user choose a target state, run safe and inspectable sensory/body/voice protocols, capture subjective and quality-gated sensor context, and gradually learn which parameter regions are useful for that person in that context.

The product logic is:

```text
person + context + target + route + timing/body/voice/visual parameters
-> subjective response + sensor quality + aftereffect + comparator history
-> safe next candidate
```

The product logic is not:

```text
frequency -> guaranteed state
```

MYND must feel like a serious instrument: tactile, alive, transparent, frontier-capable, and precise about uncertainty. It is not a meditation content library, a generic wellness tracker, a binaural-beat app, a medical-treatment product, or a paranormal-claim product.

Core thesis:

```text
Binaural beats alone are not the product.
Individualized phase/synchrony exploration is the product surface.
```

Source anchors: `F0405-F0418`, `F0434`, `F0436`, `F0441`, `F0447`, `F0448`, `F0580`, `SUP0002`, `SUP0003`, `SUP0025-SUP0029`, `SUP0047-SUP0056`, `SUP0093-SUP0106`, `SUP0141-SUP0143`.

---

## 2. Non-Negotiable Boundaries

MYND has two layers that must never collapse into each other.

| Layer | Meaning | Operating rule |
| --- | --- | --- |
| Claim layer | What MYND may say is established, causal, diagnostic, or effective. | Conservative, source-bound, user-specific, uncertainty-preserving. |
| Protocol layer | What MYND may safely implement, compare, personalize, and learn from. | Ambitious, frontier-expansive, controlled, logged, opt-in where needed. |

This split is what lets MYND mine Gateway, Hemi-Sync, Ganzfeld, remote viewing, presentiment, God Helmet, hypnosis, floatation, body-substrate, and vector-potential edge material without turning into a false-claim product.

The implementation must not claim:

- fixed frequencies cause fixed mental states
- consumer sensors detect brain states
- consumer sensors diagnose conditions
- MYND treats or cures any condition
- audio directly targets named brain regions
- haptics, bone conduction, or contact vibration stimulate named nerves or brain regions without validated hardware
- Gateway, Hemi-Sync, remote viewing, telepathy, presentiment, or entity contact is proven
- scalar/vector-potential body communication is an active v1 mechanism
- hidden suggestion is acceptable
- consumer telemetry measures dopamine, serotonin, ASSR, FFR, CFC, or network switching

Allowed language:

- "This timing pattern felt useful before, so MYND is testing a nearby variant."
- "This run is subjective-only because sensor quality was weak."
- "This target trial keeps misses and nulls beside hits."
- "This condition vector is worth retesting, not proven."
- "This protocol compares nearby timing and modality choices for you."

Rejected language:

- "This frequency caused calm."
- "Your brain synchronized."
- "MYND trained your dopamine."
- "You improved remote viewing."
- "The vector field carried information."
- "This treats anxiety, depression, PTSD, ADHD, insomnia, tinnitus, migraine, or any other condition."

---

## 3. Master Architecture

```mermaid
flowchart TD
  A["Onboarding + ConditionCalibration"] --> B["ConditionStateSignatureRegistry"]
  B --> C["StateRiskVector"]
  C --> D["SafetyGovernor"]
  D --> E["ProtocolPresetRegistry"]
  F["AppleRuntimeConstraintRegistry"] --> G["DeviceCapabilityRegistry"]
  G --> E
  F --> H["RuntimeDowngradeLedger"]
  I["StateTargetRegistry"] --> J["StateTargetCompiler"]
  J --> E
  J --> K["ClaimCopyVerifier"]
  E --> L["ProtocolRunner"]
  M["SensorQualityGate"] --> L
  N["AudioTimingSurfaceRegistry"] --> O["AudioTimingEngine"]
  P["BodySubstrateCouplingRegistry"] --> Q["HapticContactEngine"]
  R["VisualFieldEngine"] --> S["ActuatorGraph"]
  O --> S
  Q --> S
  R --> S
  T["VoiceCoRegulator"] --> L
  S --> L
  U["AlteredStateInductionRegistry"] --> L
  V["FrontierSandbox"] --> L
  L --> W["ExperimentLedger"]
  H --> W
  W --> X["ResponseSurfaceModel"]
  W --> Y["ImplicitReinforcementEngine"]
  Y --> X
  X --> E
```

No protocol bypasses calibration, state-risk resolution, runtime capability checks, safety gating, sensor-quality gating, trace logging, or copy verification.

Architecture invariants:

1. `StateTargetCompiler` runs before preset selection.
2. `SafetyGovernor` can block, reduce, or force research-only status at any time.
3. `AppleRuntimeConstraintRegistry` is platform truth; marketing assumptions or beta-only features do not override local API capability.
4. `SensorQualityGate` decides whether physiology context can influence copy or adaptation.
5. `ExperimentLedger` is the source of truth for every completed or stopped run.
6. `ClaimCopyVerifier` checks UI text, voice text, summaries, exports, and generated explanations.
7. `FrontierSandbox` protocols require explicit opt-in, denominator/null retention where relevant, and no-proof framing.

---

## 4. Platform Split

### iPhone Owns

- primary session UI
- target selector
- preset inspector
- calibration and settings review
- audio rendering and route management
- non-flicker visual fields
- text and voice capture
- frontier opt-in and target reveal UI
- trace review and response-surface rationale
- local ledger storage, export, and verifier fixtures

The iPhone session surface must always show:

- active target
- active protocol and preset
- current phase
- delivered actuator families
- route/capability status
- sensor-quality status
- stop/cooldown
- comparator/control arm when present
- claim level
- why this candidate was selected

### Apple Watch Owns

- immediate stop/cooldown
- quick safety flags
- semantic haptic markers
- heart-rate context when permissioned and quality-gated
- motion/stillness context
- wrist-level grounding cue
- immediate aftereffect prompt

Apple Watch does not own:

- precision vibroacoustic dosing
- diagnosis or treatment inference
- proof of entrainment
- hidden physiology optimization
- frontier trial interpretation
- intense or frontier launch from a voice-only path

### Shared Local Core

The first implementation should use a shared Swift package for:

- models
- registries
- safety gates
- protocol definitions
- preset fixtures
- trace schemas
- copy-verifier fixtures
- target compiler
- response-surface logic

The iOS app and watchOS extension consume the same shared core. Platform-specific renderers adapt protocols into iPhone audio/visual/voice surfaces and Watch haptic/quick-response surfaces.

---

## 5. Runtime Lifecycle

Every run follows this lifecycle:

```text
1. Target selected.
2. Calibration profile resolved.
3. Condition-state signatures resolved.
4. StateRiskVector emitted.
5. Device capability snapshot captured.
6. Apple runtime constraints checked.
7. StateTargetCompiler emits StateTargetResolution.
8. Preset or response-surface candidate selected.
9. SafetyGovernor applies block/reduce/research-only rules.
10. Comparator/control arm assigned where required.
11. Baseline recorded.
12. Ramp-in starts.
13. Active phase runs.
14. Subjective probes captured.
15. Ramp-out or stop.
16. Cooldown/re-entry.
17. ExperimentLedger row written.
18. Immediate aftereffect captured.
19. ResponseSurfaceModel updates within allowed claim level.
20. Delayed aftereffect scheduled where needed.
```

Stop always wins. A stop event must:

- halt actuator dispatch
- mark `safetyEvents`
- force cooldown/re-entry
- persist a stopped-run trace
- block promotion for the nearby parameter region until reviewed

---

## 6. Data Model Backbone

The first build should define these in the shared core before polished UI work.

```swift
struct SourceAnchor: Codable, Hashable {
    let id: String
    let lane: EvidenceLane
    let title: String
    let claimBoundary: String
}

struct StateRiskVector: Codable, Equatable {
    let sensoryLoadCap: LoadCap
    let arousalLoadCap: LoadCap
    let selfBoundaryLoadCap: LoadCap
    let checkingLoopLoadCap: LoadCap
    let sleepDisruptionLoadCap: LoadCap
    let vestibularLoadCap: LoadCap
    let suggestionLoadCap: LoadCap
}

struct StateTargetResolution: Codable, Equatable {
    let resolutionID: UUID
    let targetState: StateTarget
    let allowedProtocols: [ProtocolID]
    let blockedProtocols: [ProtocolID]
    let researchOnlyProtocols: [ProtocolID]
    let actuatorBundle: TargetActuatorBundle
    let outcomeContract: TargetOutcomeContract
    let riskTransform: StateRiskVector
    let copyBoundary: ClaimBoundary
    let sourceAnchors: [SourceAnchor]
}

struct TimingParameterFreezeState: Codable, Equatable {
    let timingFamily: TimingFamily
    let nominalRateHz: ClosedRange<Double>?
    let carrierHz: ClosedRange<Double>?
    let modulationDepth: ClosedRange<Double>?
    let dutyCycle: ClosedRange<Double>?
    let routeRequirement: RouteRequirement
    let rampInSeconds: Int
    let activeSeconds: Int
    let rampOutSeconds: Int
    let comparatorPolicy: ComparatorPolicy
    let blockedByDefault: Bool
}

struct ProtocolDefinition: Codable, Identifiable {
    let id: ProtocolID
    let frontierMode: FrontierMode
    let targetStates: [StateTarget]
    let actuatorFamilies: [ActuatorFamily]
    let claimBoundary: ClaimBoundary
    let sourceAnchors: [SourceAnchor]
    let safetyGates: [SafetyGate]
}

struct ProtocolPreset: Codable, Identifiable {
    let id: PresetID
    let protocolID: ProtocolID
    let targetState: StateTarget
    let buildLane: BuildLane
    let claimLevel: ClaimLevel
    let requirements: PresetRequirements
    let sessionShape: SessionShape
    let actuatorStack: [ActuatorPreset]
    let comparatorPolicy: ComparatorPolicy
    let promotionPolicy: PromotionPolicy
    let sourceAnchors: [SourceAnchor]
}

struct ProtocolRun: Codable, Identifiable {
    let id: UUID
    let startedAt: Date
    let endedAt: Date?
    let protocolID: ProtocolID
    let presetID: PresetID?
    let targetState: StateTarget
    let baseline: BaselineSnapshot
    let controlArm: ControlArm
    let actuatorEvents: [ActuatorEvent]
    let sensorQuality: SensorQualitySnapshot
    let subjectiveMarkers: SubjectiveMarkerSet
    let safetyEvents: [SafetyEvent]
    let aftereffect: AftereffectReport?
    let claimLevel: ClaimLevel
}

struct FrontierTrial: Codable, Identifiable {
    let id: UUID
    let trialType: FrontierTrialType
    let targetPoolID: String
    let lockedTargetID: String?
    let captureID: UUID
    let revealTime: Date?
    let decoyIDs: [String]
    let denominator: Int
    let nullRetained: Bool
    let noProofCopyShown: Bool
}

struct ReinforcementEvent: Codable, Identifiable {
    let id: UUID
    let lockedOutcomeID: UUID
    let sourceRunIDs: [UUID]
    let reinforcedProcessVariable: String
    let resultClass: ReinforcementResultClass
    let denominatorVisible: Bool
    let nullsRetained: Bool
    let claimBoundary: ClaimBoundary
}
```

Additional required model families:

- `ConditionCalibrationProfile`
- `ConditionStateSignatureProfile`
- `DeviceCapabilitySnapshot`
- `RuntimeDowngradeEvent`
- `AudioTimingSurface`
- `TimingClock`
- `TimingComparatorAssigner`
- `BodySubstrateCouplingSnapshot`
- `InductionPhaseGraph`
- `PhaseLoadVector`
- `SuggestionTransparencyState`
- `BlindCaptureRecord`
- `ResponseSurfaceCandidate`
- `ResponseSurfaceUpdate`
- `ClaimCopyFixture`

---

## 7. Foundation Modules

### `ResearchManifestRegistry`

Owns source anchors, evidence lanes, protocol mappings, and claim boundaries.

First build:

- load static JSON or Swift fixtures for all `F####` and `SUP####` anchors used by presets
- expose `anchors(for protocolID:)`
- expose `claimBoundary(for protocolID:)`
- fail tests when a protocol has no anchors

### `ConditionCalibration`

Captures sensory thresholds and preferences without diagnosis.

First build surfaces:

- low-stim baseline
- audio comfort sweep
- haptic semantic sweep
- breath comfort
- visual safety
- stop/cooldown preference
- sleep-fragility flag
- sound-tolerance/tinnitus/misophonia flag
- vestibular sensitivity flag
- panic/hyperarousal flag
- dissociation/agency sensitivity flag
- checking/proof-seeking loop flag

### `ConditionStateSignatureRegistry`

Converts user inputs and session history into non-diagnostic safety signatures.

Initial signatures:

- `hyperarousal_threat`
- `sensory_threshold_variability`
- `auditory_aftereffect_risk`
- `gamma_assr_caution`
- `attention_arousal_variability`
- `rumination_internal_stickiness`
- `checking_reassurance_loop`
- `agency_body_boundary_sensitivity`
- `sleep_fragility_hyperarousal`
- `vestibular_motion_load`
- `suggestion_susceptibility_load`
- `frontier_overbelief_load`

Outputs:

- active reductions
- blocked modes
- research-only modes
- aftereffect watchlists
- copy constraints
- allowed actuator families
- `StateRiskVector`

Source anchors: `SUP0031`, `SUP0074-SUP0092`, `SUP0122-SUP0140`, `F0060`, `F0083`, `F0179`, `F0322`, `F0379`, `F0385`, `F0394`, `F0414`, `F0415`, `F0420`, `F0421`, `F0453-F0472`, `F0478-F0489`, `F0499`, `F0575`, `F0577`, `F0578`.

### `StateTargetRegistry`

Defines product target grammar. These are control labels, not detected network states.

Initial targets:

- `calm`
- `detail_focus`
- `salience_switch`
- `internal_mode`
- `big_picture_semantic`
- `creative_alternation`
- `imagery`
- `sleep_edge`
- `presence`
- `frontier_capture`
- `rumination_guard`

### `StateTargetCompiler`

Resolves target + risk profile + runtime/device state into an executable target plan.

Inputs:

- selected `StateTarget`
- `ConditionCalibrationProfile`
- `StateRiskVector`
- `DeviceCapabilitySnapshot`
- `AppleRuntimeConstraintSnapshot`
- optional prior `ResponseSurfaceCandidate`

Outputs:

- allowed protocols
- blocked protocols
- research-only protocols
- actuator bundle
- outcome probes
- trace fields
- comparator expectations
- aftereffect rule
- copy boundary

Acceptance tests:

- `presence` is blocked or research-only when self-boundary load is capped.
- `sleep_edge` blocks focus/gamma-style candidates near sleep.
- `frontier_capture` requires opt-in, locked capture, denominator where relevant, and no-proof copy.
- weak sensors force subjective-only or context-only language.

### `SafetyGovernor`

Owns stop, cooldown, intensity caps, aftereffect rules, and escalation.

Universal stop/downgrade triggers:

- headache or migraine warning
- tinnitus spike or auditory irritation
- nausea or dizziness
- panic or hyperarousal spike
- dissociation, depersonalization, derealization, or loss-of-agency language
- sleep disruption
- sensory overload
- body pain, numbness, or delayed flare
- compulsive checking or repeated proof-seeking
- grandiosity, paranoia, or distress after frontier hits/misses
- route change during route-dependent trials
- any effect that requires hidden or misleading copy

### `ExperimentLedger`

Writes immutable traces for:

- completed runs
- stopped runs
- downgraded runs
- calibration sweeps
- frontier captures
- delayed aftereffects
- response-surface promotions and quarantines

Every trace must include target, preset/candidate, anchors, route, device snapshot, baseline, comparator/control arm, actuator events, subjective markers, safety events, sensor quality, aftereffect, and claim level.

---

## 8. Platform Modules

### `AppleRuntimeConstraintRegistry`

Platform truth for audio, haptics, sensors, speech, intents, permissions, and OS feature flags.

Rules:

- Do not assume an Apple feature exists because a beta or keynote names it.
- Separate installed SDK support, runtime OS support, entitlement availability, user permission, route state, and device-specific capability.
- Every protocol receives a capability snapshot before launch.
- Every route or permission change writes `RuntimeDowngradeEvent`.

First build checks:

- `AVAudioSession` category, route, headphones, Bluetooth, mono collapse risk
- `AVAudioEngine` render readiness and sample-rate consistency
- Core Haptics availability on iPhone
- Watch haptic semantic capability
- HealthKit authorization and recent sample quality
- Core Motion availability and stillness confidence
- Speech/local dictation availability if used
- App Intent/Siri launch safety class

Source anchors: `SUP0107-SUP0121`.

### `DeviceCapabilityRegistry`

Captures:

- iPhone model/runtime class
- Watch availability
- headphones/speakers/Bluetooth route
- haptic class
- motion class
- HealthKit permission state
- speech/input capability
- Focus/Do Not Disturb/sleep-context hints where permitted

### `SensorQualityGate`

Classifies sensor data as:

- `unavailable`
- `weak_context_only`
- `usable_context`
- `high_quality_context`

No consumer sensor state may be presented as brain-state detection. Weak sensors block adaptive physiology copy and force subjective/context-only explanations.

### `SafeIntentRouter`

Voice/App Intent paths can:

- start low-intensity routine sessions when pre-authorized
- stop, pause, resume, or cool down
- record subjective markers
- open a protocol preview

Voice/App Intent paths cannot:

- launch intense, frontier, altered-state, or research-only protocols by voice alone
- skip safety preview
- claim physiological certainty
- hide suggestion or target framing

---

## 9. Actuation Modules

### `AudioTimingSurfaceRegistry`

Timing families:

| Family | First-build status | Defaults and rules |
| --- | --- | --- |
| `amplitude_modulation` | v1 active | 0.05-20 Hz, low depth first, speaker or headphones, comparator required. |
| `monaural_beat` | v1 active/comparator | 1-20 Hz, low carrier 160-440 Hz first, softened envelope. |
| `binaural_beat` | optional comparator | 1-30 Hz difference, headphones required, route invalidation blocks or downgrades. |
| `gated_isochronic` | guarded/research | low duty harshness, no migraine/sensory overload flags. |
| `speech_voice_timing` | v1 support | low-density cadence, user interrupt, transparent prompts. |
| `music_rhythm` | phase 1.5 or v1 if scope allows | 40-140 BPM, low-medium beat strength, syncopation low first. |
| `breath_timing` | v1 support | user-comfort cadence, no forced breath holds. |
| `haptic_phase` | v1 support | semantic markers and gentle pulses, not neural dosing. |
| `visual_timing` | v1 support | slow, non-flicker visual fields only. |

Source anchors: `SUP0002`, `SUP0003`, `SUP0093-SUP0106`, `SUP0141-SUP0143`, `F0003`, `F0007-F0009`, `F0014`, `F0015`, `F0018`, `F0022`, `F0023`, `F0030`, `F0035`, `F0039`, `F0405-F0418`, `F0434`, `F0441`, `F0472`, `F0503`, `F0504`.

### `AudioTimingEngine`

First implementation:

- render deterministic AM fields
- render deterministic monaural comparators
- render binaural comparator only with confirmed stereo headphone route
- expose ramp in/out
- log carrier, rate, modulation depth, duty, route, route validity, level proxy, duration, comparator, and stop events
- keep gamma-style probes blocked or research-only until safety and validation harden

### `HapticContactEngine`

First implementation:

- Watch semantic haptic markers
- iPhone gentle contact pulse
- no named-nerve or vagus claims
- log contact site, posture/support, perceived intensity, vestibular flag, and aftereffect

### `VisualFieldEngine`

First implementation:

- non-flicker low-detail fields
- slow gradients, fields, dimming, and spatial framing
- no strobe-like entrainment
- disable high contrast or rhythmic pulses for photosensitivity/migraine flags

Source anchors: `F0060`, `F0394`, `SUP0023`.

### `VoiceCoRegulator`

First implementation:

- text-first interaction
- optional low-density spoken prompts
- user interrupt always available
- grounding/re-entry script library
- capture mode for first impressions, creative notes, aftereffects, and consent
- no hidden suggestion and no clinical omniscience

Voice design:

- short phrases
- low command density
- transparent intent
- visible transcript when voice influences a session
- clear distinction between guidance, capture, and reflection

Source anchors: `F0379`, `F0385`, `F0450-F0452`, `F0472`, `SUP0120`, `SUP0121`, `SUP0135-SUP0138`.

---

## 10. UI System

The UI should be instrument-like rather than therapeutic decor.

Design posture:

- quiet precision
- tactile glass/translucency where it clarifies layers
- warm but not cute
- no mystical overclaiming
- visible uncertainty
- persistent agency and stop affordances

Primary surfaces:

| Surface | Purpose | Required details |
| --- | --- | --- |
| Target selector | Choose intended state. | target label, evidence lane, risk estimate, disabled/research-only reasons. |
| Calibration panel | Tune thresholds. | audio, haptic, visual, breath, voice, spatial, stop/cooldown. |
| Preset inspector | Preview a candidate. | preset ID, source anchors, changed variables, comparator/control, expected probes, blocked reasons. |
| Safety strip | Persistent safety. | stop, cooldown, intensity, phase, sensor quality, route state. |
| Session surface | Run protocol. | phase graph, target, actuator stack, comparator, route, stop. |
| Trace review | Inspect result. | delivered pattern, subjective movement, sensor quality, aftereffect, claim level, why it matters. |
| Response surface | Explain personalization. | nearby parameter map, prior traces, candidate rationale, quarantine regions. |
| Frontier opt-in | Enter contested protocols. | no-proof boundary, null retention, denominator, risks, consent. |
| Induction ladder | Run P016. | orientation, settle, sensory field, capture, integration, re-entry, load classes. |
| Watch session | Safety companion. | stop, phase, quick flag, haptic marker, aftereffect. |

No UI surface may hide:

- whether a run is subjective-only
- whether a sensor was weak
- whether a route changed
- whether a protocol is frontier/lab-lane
- whether nulls and misses exist
- whether a result is personal and uncertain

---

## 11. Protocol Specs P001-P016

### `P014_CONDITION_SENSITIVE_CALIBRATION`

Build first. Creates non-diagnostic safety and preference profile.

First presets:

- `P014_CALIBRATION_LOW_STIM_V0`
- `P014_AUDIO_COMFORT_SWEEP_V0`
- `P014_HAPTIC_SEMANTIC_SWEEP_V0`
- `P014_BREATH_COMFORT_V0`
- `P014_VISUAL_SAFETY_V0`

Acceptance:

- emits `StateRiskVector`
- blocks unsafe visual timing for photosensitivity
- creates non-audio alternatives for sound-tolerance flags
- stores aftereffect watchlists

Anchors: `SUP0031`, `SUP0074-SUP0092`, `SUP0107-SUP0121`, `F0060`, `F0083`, `F0179`, `F0322`, `F0394`, `F0414`, `F0453-F0472`, `F0483-F0499`.

### `P001_SYNCHRONY_SEARCH`

Core individualized rhythmic field exploration.

First presets:

- `P001_SYNC_BREATH_AM_CALM_V0`
- `P001_SYNC_ALPHA_STYLE_FOCUS_V0`
- `P001_SYNC_THETA_STYLE_IMAGERY_V0`
- `P001_SYNC_BINAURAL_COMPARATOR_V0`

Rules:

- AM/monaural first
- binaural only with valid stereo headphone route
- breath-scale AM 0.05-0.2 Hz for calm/grounding hypotheses
- alpha-style 8-12 Hz and theta-style 4-8 Hz are labels for timing candidates, not claims
- gamma 35-45 Hz remains research-only or blocked by safety flags

Acceptance:

- route invalidation downgrades or stops binaural
- comparator arm logs correctly
- no fixed-frequency copy
- aftereffect captured

Anchors: `F0405-F0418`, `F0434`, `SUP0002`, `SUP0003`, `SUP0093-SUP0106`, `SUP0141-SUP0143`, `SUP0036-SUP0039`.

### `P013_VOICE_COREGULATOR`

Text/voice grounding, capture, reflection, and re-entry.

First presets:

- `P013_TEXT_FIRST_AGENCY_V0`
- `P013_VOICE_LOW_DENSITY_V0`
- `P013_GROUNDING_REENTRY_V0`

Acceptance:

- text fallback always exists
- transcript is visible where voice affects session
- interrupt always works
- no hidden suggestion

Anchors: `F0379`, `F0385`, `F0450-F0452`, `F0472`, `SUP0120`, `SUP0121`, `SUP0135-SUP0138`.

### `P003_CREATIVE_ALTERNATION`

Alternates open association and convergent selection.

First presets:

- `P003_OPEN_SHAPE_2CYCLE_V0`
- `P003_BIG_PICTURE_SEMANTIC_V0`
- `P003_FOCUS_RETURN_V0`

Session shape:

- 180s open
- 90s shape
- 30s capture buffer
- optional repeat

Acceptance:

- capture does not interrupt open phase
- rumination/agitation can stop or shorten session
- output logs novelty/usefulness without claiming creativity enhancement

Anchors: `F0438`, `F0448`, `F0580`, `SUP0049`, `SUP0052-SUP0054`.

### `P006_SENSORY_HOMOGENIZATION`

Reduces sensory complexity for calm, imagery, or frontier capture prep.

First presets:

- `P006_AUDIO_ONLY_LOW_INPUT_V0`
- `P006_VISUAL_ONLY_LOW_DETAIL_V0`
- `P006_COMBINED_LOW_GUARDED_V0`

Acceptance:

- combined mode only after component arms pass
- re-entry mandatory
- stop on dissociation, panic, claustrophobia, or sleepiness

Anchors: `F0400`, `F0401`, `F0501`, `F0379`, `F0385`, `SUP0139`, `SUP0140`.

### `P009_CONTACT_COUPLING`

Uses body-contact rhythm and grounding without precision dosing claims.

First presets:

- `P009_WATCH_GROUNDING_MARKER_V0`
- `P009_PHONE_CONTACT_PULSE_V0`

Acceptance:

- Watch haptics are semantic markers
- vestibular flags reduce or block
- contact site, posture/support, perceived load, and aftereffect log

Anchors: `F0538-F0547`, `F0576-F0579`, `SUP0057-SUP0070`, `SUP0107-SUP0112`.

### `P002_SALIENCE_SWITCHER`

Soft mode-shift cues without claiming to toggle a brain network.

First build:

- low haptic/audio cue
- planned phase boundary
- cue/no-cue comparator
- no startle default

Anchors: `F0436`, `F0441`, `F0449`, `SUP0047`, `SUP0048`, `SUP0050`, `SUP0051`.

### `P007_INTENTION_LOCK`

Timestamped first-impression capture before rational editing.

First build:

- text/sketch/voice capture
- edit lock
- confidence before reveal/compare
- user-authored phrase only when visible and consented

Anchors: `F0396`, `F0397`, `F0402`, `F0450-F0452`, `F0501`.

### `P016_ALTERED_STATE_INDUCTION_LADDER`

Decomposes Gateway-like, meditation-like, hypnosis-adjacent, sensory-reduction, body-boundary, and frontier-capture variables into phase graphs.

First preset:

- `P016_SHORT_GUARDED_LADDER_V0`

Phase graph:

```text
orientation 60s
-> settle 120s
-> sensory field 180s
-> capture 90s
-> integration 60s
-> re-entry 90s
```

Acceptance:

- logs `suggestion_load`, `self_boundary_load`, and `sensory_reduction_load`
- cannot skip re-entry
- no Gateway validation, OBE proof, metaphysical access, clinical hypnosis, or treatment claim

Anchors: `F0395`, `F0379`, `F0385`, `F0400`, `F0401`, `F0450-F0452`, `SUP0001`, `SUP0122-SUP0140`.

### `P012_HYPNAGOGIC_CUE`

Sleep-edge cue and capture without sleep harm.

First build:

- phase 2
- low-intensity cue
- no active-use mode
- morning aftereffect
- downgrade if sleep worsens

Anchors: `F0427`, `F0428`, `F0432`, `F0433`, `F0478`, `F0575`, `SUP0090`, `SUP0091`.

### `P004_PRESENCE_SANDBOX`

Opt-in self-boundary/presence perception experiment.

First build:

- lab/frontier only
- low ambiguity
- no entity/contact/body-exit framing
- no-prime, prime-only, spatial-only, and haptic-only controls

Anchors: `F0398`, `F0399`, `F0585`, `F0586`, `SUP0004`, `SUP0005`, `SUP0122-SUP0129`.

### `P005_ANOMALOUS_PROTOCOL_SANDBOX`

Blind first-impression and target-imagery experiments.

First preset:

- `P005_LOCKED_TARGET_IMAGERY_V0`

Rules:

- sealed target pool
- random locked target
- decoys
- locked capture before reveal
- confidence before reveal
- denominator visible
- misses and nulls retained
- no remote-viewing proof language

Anchors: `F0396`, `F0397`, `F0400-F0403`, `F0500`, `F0501`, `SUP0006-SUP0013`.

### `P015_IMPLICIT_SIGNAL_REINFORCEMENT`

Links locked/scored outcomes to prior safe condition vectors.

First preset:

- `P015_PROCESS_REINFORCEMENT_V0` behind gates

Rules:

- metadata hooks first
- process reinforcement before outcome reinforcement
- no streaks
- no ability/proof/dopamine/serotonin/subconscious-control copy
- misses and nulls can reinforce good process

Acceptance:

- no reinforcement without locked outcome, denominator, trace link, null retention, and copy guard

Anchors: `F0252`, `F0268`, `F0450-F0452`, `F0472`, `F0500`, `F0502`, `SUP0042-SUP0046`.

### `P008_INTERPERSONAL_SYNCHRONY`

Consented shared timing and co-regulation.

First build:

- phase 2
- both users consent
- privacy boundary visible
- no mind-reading, telepathy, or authorship claim

Anchors: `F0452`, `F0472`, `F0579`, `SUP0014-SUP0017`.

### `P010_BONE_LINE`

Body/bone-conducted sound as hardware/lab research lane.

First build:

- hardware/lab only
- vestibular calibration required
- short trials
- air-only/contact-only/combined/sham arms

Anchors: `F0538-F0547`, `SUP0021`, `SUP0067-SUP0070`.

### `P011_LOAD_PROBE`

Posture/contact/support-surface exploration.

First build:

- lab lane
- posture/support/contact metadata
- no fascia, collagen, scalar, or vector-potential proof claim

Anchors: `F0557`, `SUP0022`, `SUP0061-SUP0064`, `SUP0071-SUP0073`.

---

## 12. Personalization And Promotion

`ResponseSurfaceModel` learns personal usefulness regions without implying universal causality.

Candidate inputs:

- target state
- calibration profile
- risk vector
- recent aftereffects
- prior similar runs
- comparator history
- sensor-quality class
- device route class

Promotion to "worth retesting" requires:

1. at least three completed runs in the same broad parameter neighborhood
2. no stop events or adverse aftereffects
3. comfort plus target-state movement or usefulness
4. at least one nearby comparator/control
5. sensor-informed copy passes `SensorQualityGate`
6. claim remains personal and uncertain

Promotion to "preferred" requires:

- repeated safe traces
- comparator history
- retained null/negative runs
- visible uncertainty
- no safety cap relaxed from excitement alone

Immediate quarantine requires:

- adverse physical/sensory/agency aftereffect
- route change during route-dependent trial
- two early stops in a similar region
- effect depends on suggestive/proof-heavy copy
- frontier trial produces obsession, overbelief, distress, paranoia, or grandiosity

Source anchors: `SUP0025-SUP0029`, `SUP0042-SUP0046`, `F0252`, `F0268`, `F0470`, `F0471`, `F0500`, `F0502`.

---

## 13. Frontier Sandbox Rules

Frontier work is part of MYND's strength only if it is controlled.

All frontier protocols require:

- explicit opt-in
- no-proof framing
- visible uncertainty
- first-impression lock where relevant
- denominator/null retention where relevant
- stop/cooldown
- re-entry for altered-state or sensory-reduction loads
- copy verifier
- trace export that includes misses and nulls

Remote-viewing-like and anomalous trials are implemented as blind target/capture protocols, not as ability claims. RNG/presentiment-adjacent sources may inform locked-outcome methodology and reinforcement design, not public proof copy.

The remote-viewing self-reinforcement concept is handled as a generalized implicit-process learner:

```text
safe condition vector + locked process/outcome + delayed score
-> conservative reinforcement event
-> candidate replay or cue adjustment
```

It must never say dopamine or serotonin was measured or trained. It may say a previously logged condition pattern is being gently retested because it preceded useful process quality.

Source anchors: `F0396-F0403`, `F0500-F0502`, `SUP0006-SUP0013`, `SUP0042-SUP0046`.

---

## 14. Build Execution Sequence

### Stage A - Instrument Kernel

Build first:

1. `ResearchManifestRegistry`
2. `ConditionCalibration`
3. `ConditionStateSignatureRegistry`
4. `StateRiskVector`
5. `StateTargetRegistry`
6. `StateTargetCompiler`
7. `ProtocolPresetRegistry`
8. `SafetyGovernor`
9. `ExperimentLedger`
10. `DeviceCapabilityRegistry`
11. `AppleRuntimeConstraintRegistry`
12. `SensorQualityGate`
13. `ClaimCopyVerifier`

Exit criteria:

- calibration emits allowed/reduced/blocked surfaces
- target selection emits `StateTargetResolution`
- presets resolve through safety and capability gates
- traces persist
- stop/cooldown works
- copy fixtures pass

### Stage B - Core State Instrument

Build:

1. `AudioTimingSurfaceRegistry`
2. `AudioTimingEngine`
3. `ActuatorGraph`
4. `VisualFieldEngine`
5. `HapticContactEngine`
6. `VoiceCoRegulator`
7. `P001_SYNCHRONY_SEARCH`
8. `P003_CREATIVE_ALTERNATION`
9. `P006_SENSORY_HOMOGENIZATION`
10. `P009_CONTACT_COUPLING`
11. `P013_VOICE_COREGULATOR`

Exit criteria:

- AM/monaural/binaural comparator logic works
- route-invalid binaural downgrades or blocks
- Watch haptics are semantic/contact-classed
- no unsafe visual flicker
- weak sensors downgrade copy
- text-first voice works without speech

### Stage C - Personalization

Build:

1. `ResponseSurfaceModel`
2. `NOfOneControlScheduler`
3. `MicroRandomizedCueScheduler`
4. promotion/downgrade engine
5. response-surface rationale UI
6. aftereffect scheduler

Exit criteria:

- no preset becomes preferred from one run
- adverse aftereffects quarantine nearby regions
- user sees why a variant was selected
- null/negative runs remain visible

### Stage D - Guarded Frontier

Build:

1. `FrontierSandbox`
2. `AlteredStateInductionRegistry`
3. `P016_ALTERED_STATE_INDUCTION_LADDER`
4. `P007_INTENTION_LOCK`
5. lab-gated `P005_ANOMALOUS_PROTOCOL_SANDBOX`
6. metadata hooks for `P015_IMPLICIT_SIGNAL_REINFORCEMENT`

Exit criteria:

- induction phase graph cannot skip re-entry
- load classes are logged
- suggestion transparency is visible
- locked target capture works
- denominator/null retention works
- no proof copy appears

### Stage E - Later Expansion

Build:

1. `P004_PRESENCE_SANDBOX`
2. `P008_INTERPERSONAL_SYNCHRONY`
3. `P010_BONE_LINE`
4. `P011_LOAD_PROBE`
5. full `P015_IMPLICIT_SIGNAL_REINFORCEMENT`
6. `P012_HYPNAGOGIC_CUE`
7. external hardware registry
8. research export mode

Exit criteria:

- explicit opt-in
- hardware safety envelopes
- privacy and consent controls
- lab-lane export rules
- no body-region, vagus, scalar, vector, entity, telepathy, remote-viewing, or treatment proof claims

---

## 15. Verifier And Test Matrix

Required verifier families:

| Verifier | Must check |
| --- | --- |
| Research package verifier | anchors, stale language, claim boundaries, protocol coverage |
| Protocol metadata verifier | anchors, gates, claim boundary, evidence lane |
| Protocol preset verifier | preset anchors, trace fields, blocked defaults, comparator policy |
| Claim copy verifier | UI/voice/export copy cannot exceed evidence lane |
| Trace completeness verifier | baseline, actuator events, sensor quality, subjective report, safety state |
| Safety scenario tests | photosensitivity, vestibular, panic, dissociation, sleep, sound tolerance, checking-loop flags |
| Audio route tests | binaural and route-dependent sessions downgrade/cancel correctly |
| Haptic capability tests | Watch haptic semantics and Core Haptics capability are separate |
| Sensor quality tests | weak sensors block adaptive physiology language |
| Frontier sandbox tests | opt-in, locked target, denominator, null retention, no-proof copy |
| Altered-state tests | phase graph, load vector, stop route, re-entry, aftereffect |
| Reinforcement tests | locked outcome, no streak/proof feedback, nulls preserved |

First acceptance tests:

- photosensitivity disables flicker/strobe-like visual patterns
- vestibular sensitivity disables motion/contact modes unless explicitly re-enabled with short trial
- panic flag disables startle/presence intensity
- dissociation flag gates sensory homogenization, presence sandbox, anomalous sandbox, bone-line, and load-probe protocols
- stop action always wins over active protocol
- no frontier protocol launches without opt-in
- no induction run lacks re-entry when sensory-reduction, presence, body-boundary, frontier-capture, or suggestion load is active
- no reinforcement event exists without locked outcome, denominator, trace link, null retention, and copy guard
- no UI/voice/export copy says fixed frequencies cause fixed states
- no completed run lacks source anchors and trace fields

---

## 16. Open Research And Validation Hooks

These do not block Stage A/B implementation, but they must remain visible:

1. Paper-by-paper parameter tables for high-value phase/synchrony sources.
2. Deeper Gateway/Hemi-Sync primary-source extraction beyond the CIA seed.
3. Deeper God Helmet/Persinger/Koren/Granqvist trail.
4. Deeper condition-state extraction for misophonia, PTSD, dissociation/FND, OCD/checking, tinnitus, autism sensory processing, insomnia, migraine, and bipolar/psychosis-spectrum ASSR/gamma caution.
5. Apple Watch haptic semantics and Core Haptics validation.
6. Audio route/spatial audio/sensor quality validation.
7. Product-copy verifier once real UI copy exists.
8. Lab-lane protocol design for EEG/ear-EEG, external transducers, anomalous locked trials, and human-subjects-style exports.

The build should proceed by instrumenting safely first, then letting real traces identify which research gaps matter most.

---

## 17. Immediate Decomposition

The next build-planning pass should create epics and stories from this order:

1. Shared core package and fixture loading.
2. Source anchor registry and claim-boundary fixtures.
3. Calibration and `StateRiskVector`.
4. State target compiler.
5. Preset registry with P014/P001/P013/P003/P006/P009 fixtures.
6. Protocol runner and ledger.
7. Audio timing engine with AM/monaural/binaural route gates.
8. Watch stop/cooldown and semantic haptics.
9. Visual non-flicker engine.
10. Text-first voice/capture/re-entry.
11. Trace review and aftereffect scheduler.
12. Response-surface candidate selection.
13. Frontier opt-in shell, P016 metadata, P007 locked capture, and P005 lab gate.
14. Verifier expansion.

Definition of done for this blueprint:

```text
An engineer can now decompose MYND into build stories without guessing the
product category, safety model, audio timing posture, Watch role, voice role,
frontier boundaries, data backbone, protocol rollout, or verifier strategy.
```